THE SONY SERVICE CODES - Articles, Comments, Discoveries [Text View] - Page 4

Nitewatchman

01-21-06, 02:52 PM

First off, am more active in other threads in HDTV areas and unfortunetly don't have a lot of extra time to contribute to this one, but I have been utilizing it extensively for the past 7 months. Wanted to say thanks much to everyone for all the excellent info, as well as contribute a couple of hopefully helpful items while I have some time for it on a sat afternoon. AVS has seen a lot of great threads I've seen over the years, but I've got to rate this one near the top.

First a few thoughts on Ken's new image processing settings/posts, then a couple of XBR960 specific things relevant to MID5 columns and the relevant 2103-1/2103-2 settings ...

Having used them for almost a week now, I really like it! I am obsessing a little over SYSM=2 or 3, though, probably mostly just because I've been using 3 for everything for so long. As it turns out I do really like the unsharp-mask like effect, and I also think more of a photo-realistic look with it at 2 - given the other new P2170-3 and MID5 Column's I've set up/settings at least. Haven't had a lot of time with the new settings actually watching a lot of content however.

At first, had one of the pic modes still set up so I could compare the "old" settings to the new(well, except for the 3dcomb/2101-1 settings when applicable and checked those out a little bit when making the changes as well), and so far haven't seen anywhere I would want to keep my "old" settings. In fact, I've already seen enough that I have already since changed that one to the new settings as well, but currently pointing to a MID5 column with all 0's currently as I usually use it for experimenting, anyway. I have also left "pro" 2170p3 settings at set default for reference, BTW, except for the VM settings, 2170p3 #2~5. I did have things pretty much set up per Ken Tech's Sept. 05 updates to his image processing article/posts and/PDF, except that I found I had to do a little "more"(actually a lot more) for my particular DVD player with 480p DVD via component, via a different MID5 Column I sest up for it. Strangely enough, I'm not getting that now, and am using the same MID5 column I'm using for 720p.

Would also like to note that I agree with Ken concerning the SVM settings, and I'm using his VM settings at either "low or Medium" for just about everything - I like it anyway. I used to have the RCA 38"(F38310 - XBR960 is actually a warranty replacement for that set) and allways thought they did a great job implementing SVM on it, which I think was very much along the lines(but maybe a tad bit more/not much) of the same thing I get with the XBR and "medium" or Low with Ken's settings. Wish I could figure out how to do much the same on a couple of other CRT HD displays I have(Toshiba 34HF84 and Samsung TXN2668WHF), but doesn't look like it's going to work out, so SVM is off on those.

Also, might be interesting to note that with the new settings, I have sharpness slider at "30" for everything except from My DVD Player, and a JVC S-VHS VCR, for which I use sharpness "31". Seems that, with my eyes at least, that's right where it "needs" to be.

XBR960 specific notes concerning image processing settings :

#1). XBR960 and factory MID5 Columns -- - On my XBR(manufactured Jan 05 at the PA plant), and according to the XBR960 service codes PDF :

A). For the set defaults, MID5 Columns #53~#56 are actually assigned to P2170-3 MIDE#16. -- It's really silly I think, as for instance, Columns #53~63 are all zeros at the factory defaults, however, #53~56 are assigned for MIDE for the different Pic modes for Twin-view feature ... #53=Pro, #54=Movie,etc. So, just something for XBR owners to keep in mind when deciding which MID5 column's to take over. Right now, what I'm doing is I have MID#5 POP#60~63 still set up with values I had been using or experimenting with previously, and Column #57~59 with a few of the suggested "columns" from the Ken Tech's new findings. So far, those 3 have been all I have setup with the new settings(and so far appears that those 3 will probably be all I need), but I don't have anything assigned for #54~56 and may use those if necessarly as well. Am using #53 for the "all zeros" column when I want to use it/try it(and still for pro "twin view" left at default) and, can probably use column #60~63 with other values than what I have them at, currently if desired.

B). I don't think anything is actually assigned to Column #28~52 either, according to the PDF at least. But I haven't checked inputs/pic modes and 2170P3 MIDE#16 value for inputs I'm not using to see. Also, according to the XBR960 service codes PDF file, except for #50, those Columns aren't all zeros, so I probably wouldn't want to change those, anyway.

2). 2103-2 #6-SHAP, #7-SHFO, #8-PREO, #16-SSMD : In addition to the changes for 2103-1 #6SHAP,#7-SHFO,#8-PREO,and #16-SSMD as mentioned by Ken for the various appicable inputs -- XBR960 owners might also want set 2103-2 #6-SHAP,#7-SHFO,#8-PREO and #16-SSMD to the suggested values for the 2103-1 "versions" as well. These 2103-2 values seem to be solely for such features as "twin-view" and Index .. although, (and here's one place where it gets odd), on my set at least, the "input" specific nature of these settings don't seem to be quite exactly as lableled the in the 960 servvice codes PDF. From what I could tell they aren't using a lookup table/etc, so don't know why some of the default values+such are labeled in red lettering in the PDF.

I did confirm by visably inspecting the results that modifying values for 2103-2 #6,#7,#8 and #16 did affect "twin-view", but no other input or "feature" that I checked at first -- However, although this is probably going to sound odd -- I didn't spend a lot of time on these, so can't specifically tell you exactly WHAT all inputs/features you need to go through and modify these values for if you want to set these to new values -- I *think* I set the new values for P2103-2 #6+#7+#8 while in "Twin-View" and "RF"(using analog OTA with it specifically) -- On the RF part of it, don't ask me why since the values used for "RF" input(but not for index) are actually set in 2103-1 #6#7#8 and #16 as Ken wrote up, so would have thought I'd have had to select "index" to set them in 2103-2 for "index", but that didn't turn out to be the case .. What I did really was just go ahead and cycle through all inputs/scan rates I'm using and twin-view+Index and make sure the new settings for 2103-2 stuck everywhere(they did), even though, it seems its actually only "twin-view" and "index" features which are effected by these in 2103-2.

Hope that makes some sense, should become more apparent what I'm talking about when/if you go to change those settings - Perhaps, just setting them to the new values while you are using both "twin-view" and "index" would do the trick - that would seem to make sense, but of course I didn't quite do it that way. Since the PDF file doesn't seem to label the "inputs" correctly, you'll probably want to write down the default values and "where" you actually changed them rather than expect to be able to rely on the PDF file.

Wow, what a bunch of writing just about setting SHAP/SHFO/PREO/SSMD "index feature" and for "twin-view" ... Not that you may feel it actually necessary to change those for index+twin view features(and I certianly didn't spend much time on it), but in case you do, there it is .... I do think this thing has WAY too many settings at times ... Yikes, why couldn't it just use the 2101-1 values for these. Sure would be nice to be able to do this with a PC "hooked up" to the set rather than using the remote .... Only had the set about 7 months, and the labels on the 1, 2 and 3 buttons are already ALL the way gone. Went through that first set of batteries for it awfully quick as well ....

Nitewatchman

01-21-06, 03:01 PM

Sorry for the adjacent double posts, thought It made sense to do these seperately. Also - apologize for the length of these, unfortnetely I've never been all that good about writing clear and consise+short posts.

Actually it looks like we're both slightly off and that these controls effect 480p/1080i, but not 480i. FWIW, I did a few tests to see which inputs and signals are effected by these controls on my 34XBR800 and this was the result.INPUT SIGNAL MTRX CBGN CRGN YGN

Composite 480i No No No No
Component 480i No No No No
Component 480p Yes Yes Yes Yes
DVI 480p Yes Yes Yes Yes
DVI 1080i Yes Yes Yes YesSo it appears that these may change the color space (MTRX) and offset decoding (CBGN, CRGN, YGN) for ATSC/HD (480p/720p/1080i) versus NTSC/SD (480i).

I can't be sure about 720p since I don't have any 720p signals handy, and it also has to be "upconverted" for display like 480i. And my 34XBR800 has a slightly older version of this chip (CXA2151, rather than CXA2171), which could work a little differently. But based on the above it looks like these are probably ATSC/HD-based controls. Maybe someone else could try 720p to confirm this. (Pls recall the earlier caveat re writing down the original value for MTRX before changing it, since this control is automatically saved w/o being written to memory. 0-Enter won't work on MTRX either.)

Well, that may explain a lot. Thanks for digging this up, as it may turn out to be very beneifical for me.

More detail later, but It appears color decoding differs a bit on my set for 480p+1080i component sources and I'm pretty sure from 1080i via the internal ATSC receiver as well, this must be why. I think I've noticed this in a couple of different ways, now I know where to look to see if I can do anything about it.

I'm not sure about 720p yet, from what I've seen I'd bet either that #1). it doesn't, or #2). If it does effect 720p, then seems like something a little "different" is going on, perhaps elsewhere concerning 720p vs. 1080i sources from the internal ATSC receiver. More on this in #2 below.

Here's the 2 places where I think I've noticed this.

#1). Using AVIA/DVE Color test bars/decoder test and turning on RGB guns indiviudually, I get pretty much spt on results(AVIA color decoder test shows RGB all in the 0~+5% range, nothing above about +3% as far as I can tell, and the highest, slight bit of oversaturation would be blue - does help me to see what's going on with that if I take off my glasses!) with 480i DVD with Hue at 0, Color=31, and 2170P-4 RYR~GYB at :

13-15-5-3

However -- That doesn't work for 480p DVD. I need values of Color=34, Hue=0, and RYR~GYB = 14-14-6-3, there to get the closest results(but can't get "as close" as with 480i) using DVE or AVIA. At first, I wondered if something weird was up with the DVD player. But, I'd also used this same Zenith DVD player with DVE/AVIA to calibrate color decoder on my Toshiba 34HF84, and I noted that for instance with it, I needed the SAME exact settings for either 480p or 480i from that player with it. For that matter, the a cheap JVC 480i only DVD player needed the same settings as the Zenith. Must not be anything "wrong" with the DVD players color output, I thought. Note that 34HF84 has scan rate and in some cases SD "signal type"(s-video or composite) specific settings available for all of it's various color decoder settings it has such as G-Y or R-Y.

#2). Ok, here's the kicker concerning how I noticed something seemed a little different about color decoding at 480p and 1080i than anywhere else, and that what was going on with 480p is probably the same thing that's happening with 1080i. Might sound like I'm going off on a tangent here, but first, I have to explain something about what I did that has nothing to do with color decoding.

Having had "grey bars" for 4x3 SD signals up on the other DV-CRT HD displays I've had - It might seem very odd, but I have gotten so used to it I've even come to prefer it, and also, although I'm not really worried about it also believe having those there may help a bit with uneven phosphor wear, if one watches much 4x3 content with 4x3 aspect ratio on a 16x9 CRT.

So, I found out how to "put in" the grey bars on the XBR960(for "normal" mode, it only works with 4x3 480i/p AND *for a grey, instead of black background for "twin view" windows) by changing MID1 #24 - BCOL to 6 from it's default(black). For anyone crazy as me who likes "grey bars" Note that BCOL ranges from 0~15, with 0 being black, and 15 being white, so you can pretty much have the "shade" of grey you like. Note that also, when I adjusted this, I noticed what was "labeled" Others or "normal" in the service codes PDF didn't necessarily seemed to "jive" with "where" I had to adjust the BCOL value to where I wanted it and have it stick. And, oddly enough, I ended up having to adjust BCOL when I was using "full" screen mode to actually get it to stick for "normal" go figure - even though it doesn't even effect "full mode" in any way, shape or form. One of those things where it's probably best to cycle through Normal/full/widezoom/etc, as well as "twin view" to make sure your new value sticks everywhere it is supposed to. For my set I also set it to "6" for "twin view" as I found I preferred a grey background there as well(I know, I know I'm crazy), but, for example, one could set it for "normal" for grey sidebars, and leave "twin view" background at Black if they so desire.

I noticed RIGHT away, that when cycling through inputs in Twin view, that when I selected something that was a 1080i signal(via ATSC internal receiver/OTA is currently my only source for 1080i or 720P sources) OR 480p signal, my new "grey" background turned a slightly different color than I got from anything else - 480i via s-video or composite or component OR ATSC internal tuner, 720p via internal tuner/etc/etc. the grey background stays exactly the same, but For 480p or 1080i, the HUE for it turns slightly "pinkish", the exact same shade of "slight pinkishness" to the grey, in fact. At first, even though I'd went through and checked it/modified if necessary the 2170P-1 "CBOF"/CROF" settings for all inputs thought I might have something going on there, but that just didn't seem to be the case.

Now, while I'm not positive about this, I *think* and am somewhat certian when you are using "twin view", SM settings/values specific to the input/scan rate/etc. are used for the signal you are displaying+is active/you have selected(or might just be for the left window -- can't get to the set and check it currently). If so, and given what I'm getting with AVIA/DVE at 480p, then it looks like perhaps I need a little different settings than the factory defaults for the relevant CXA2171 settings.

Also think If I recall correctly from previous info -- it's using component video "inside the set", in this case it's likely the MTRX/CBGN/CRGN+YGN values would probably effect 480p/1080i signals fromthe internal ATSC/QAM as well.

I haven't tried looking at the Color bars for 1080i/p in the internal QM test patterns with this yet, but that's something else on my list the next time I can get a chance to get the set away from the rest of the family, or more specifically, the family away from the set for a long enough period of time I can have another "SM session" with it ...

Also, don't know about Memory stick(1080i), on this can't look at that one via Twin view, I don't think.

-------------------------------------

So what I am thinking(hoping anyway) is : Given what I'm seeing with the Hue of the "grey background" on Twin-view for 480p+1080i being exactly the same, yet slightly different for everything else : Perhaps If I can manage to get color "right" with 480p DVD using the same values I'm using for Hue/Color sliders and RYR-GYB for 480i DVD, BY adjusting the CXA2171 MTRX CBGN CRGN YGN values for 480p and AVIA/DVE Color bars/color decoder tests, then hopefully, I can somewhat expect that 1080i should be "correct" as well ?

Keep thinking about "ATSC color space" though, and wondering if it's really going to work out that way, given 480p NTSC from DVD vs. 1080i from ATSC, or if I'll have anyway of knowing for sure without a HD test pattern generator ..... wouldn't the internal QM color bars at 1080i work to check it as well, or is 2171 chip bypassed - Think I'd read it had been detirmined it wasn't but allways got ask these sorts of questions about this sort of thing, hey that's part of the fun of it :) ....

Am I sort of thinking along the right lines on this/anyone have any thoughts or suggestions about it?

I do notice that the service codes PDF file for the XBR960 shows both 1080i+720p component or ATSC under the same heading for "MTRX", which if previous experience with some other settings have been any clue may not mean much, especially as I notice there's a "others" column there as well ... certianly don't want to "mess up" 720p, though, and hope this isn't going to work out where you can have either 720p or 1080i "just right" but not both .... Will have to experiment a bit with this when I get a chance and see what happens, and to see if I get the same results as you did concerning 480p/1080i, which I'm expecting to be the case, but if I can I'll check it just to make sure -- Have no way to check DVI/HDMI at present, as well as 480p from the ATSC tuner, as I have no stations in the area sending 480p, currently.

Cleanupthat

01-21-06, 05:02 PM

Ok, i jumped into the service menu, testing different settings. I found out it seems my horizontal size and position is badly off. My vertical size and position is badly off too. (I think) I need to know how to test these settings to what they are supposed to be set at for perfect calibration. Ahhhh, i dont have any way of getting calibration discs, Can i purcase a usb thumbdrive and copy things from the internet to my 360? What would you recommend i do to get perfect (well maybe not too perfect) vertical and horizontal size and position.

DSperber

01-21-06, 07:33 PM

I need to know how to test these settings to what they are supposed to be set at for perfect calibration. Ahhhh, i dont have any way of getting calibration discs, Can i purcase a usb thumbdrive and copy things from the internet to my 360? What would you recommend i do to get perfect (well maybe not too perfect) vertical and horizontal size and position.Read this post (http://www.avsforum.com/avs-vb/showthread.php?p=6764404&&#post6764404) post, and the other posts highlighted within... as well as retrieving the OVERSCAN.ZIP 1920x1080 test pattern attached to one of those other posts. If you can display the cross-hatch JPG that's in that zip file from your 360 to your XBR960 at 1080i, you should be on your way.

ikrups

01-22-06, 04:52 AM

Jediphish

01-22-06, 09:45 AM

I bought a new 30XS955 recentely, it works okay and there is a picture szie problem.
I tested with a HIVI CAST DVD and I only have 90% picture of the patten (safty zone), for example, if we take a 1920 x 1080 picture and put it on TV then I only see 1700 x 960 something on screen.
Can this be adjusted to have more or even 98% of the picture?

thanks a lot.

Easily correctable. See post #783 above for one of the posted Serv. Manual excerpts re: geometry adjustment. Right click and open in new window to view the file. Other posts will tell you how to get into the service menu (post #1! for instance).

Also, as you'll read in other posts (on this page, I believe), you can (and should) use the "Search This Thread" function (at the top of the thread), to search for posts discussing "overscan," which is what you're trying to adjust. Just be sure to read several posts before and after any that you zero in on, to be sure that there aren't any important dissenting opinions on what is stated. Sometimes people get stuff wrong and they usually get corrected a few posts later.

Good luck with your Service Menu changes.

Jediphish

01-22-06, 11:49 AM

OK - Using DVE's Overscan and Crosshatch patterns ouput from my DVD player at 480i to Vid5, I successfully centered my raster (using 2170P1 & 2, and then correctly centered and adjusted the image using MID2 0-3. When I watch a DVD everything is great. When I watch my D* HR10-250 for an SD channel output at 480i, everything is great (the grey side panels are perfectly parallel and straight up & down). (I'm at 2.5% overscan by the way for 480i).

But when I switch the output on my HR10-250 to 1080i, I notice that the gray side bars on SD material are narrower (closer to the edge of the bezel). I think this means that I need to adjust the 1080i image using MID3, however I've had mixed results trying to do this. Since my DVD player only outputs DVE images at 480i, I can't use DVE to adjust the 1080i image or overscan. So, I tried using the PATN crosshatch pattern and a 1080i signal from my D*box, and the MID3 table adjustments. The weird thing is the "original" values that were set for MID3 0-3 (VDHP - VDVS) were different when I checked them after turning on the PATN 1 crosshatch, than the values that were there when I just check MID3 while watching a 1080i output from my HR10-250 (without the PATN crosshatch engaged).

Does anyone know why the values for MID3 for 1080i/V5/V6 would be different with PATN crosshatch turned on than when it is not turned on?

On a related issue - as I was saying, I got my geometry and overscan set perfectly for 480i V5/V6 using DVE. But now when I output SD source material at 480i, everything looks great, but the same material output/upconverted at 1080i "bows" out (as if PIN and MPIN need to be adjusted). I was under the impression that PIN and MPIN are global settings. Why would they be perfect for 480i and off at 1080i?

Last - are the first four values for MID3 completely equivalent (as far as what they control) as the first four values of MID2 - the acronyms are not the same, nor does there appear to be a Vertical Position value for MID3 (It's labeled Envelope instead).

Thanks as always.

I've balanced out PIN and MPIN so that the outwards pincushion bow on 1080i and the inwards pincushion bow on 480i (both for V5/V6) are about the same, respectively. But if I set either one to be straight, the other is concave or convex. Is this just something I'll have to live with?

Also - if anyone wants to be a friend and do me a favor, could you check your 1080i/V5/V6 MID3 values #s 0-3 with PATN crosshatch pattern engaged and see if the values are diffferent than without the PATN crosshatch engaged? Thanks!

ikrups

01-22-06, 12:09 PM

Easily correctable. See post #783 above for one of the posted Serv. Manual excerpts re: geometry adjustment. Right click and open in new window to view the file. Other posts will tell you how to get into the service menu (post #1! for instance).

Also, as you'll read in other posts (on this page, I believe), you can (and should) use the "Search This Thread" function (at the top of the thread), to search for posts discussing "overscan," which is what you're trying to adjust. Just be sure to read several posts before and after any that you zero in on, to be sure that there aren't any important dissenting opinions on what is stated. Sometimes people get stuff wrong and they usually get corrected a few posts later.

Good luck with your Service Menu changes.

Thank you Jediphish , I'm going to print all this thread and work on it. :)

Nitewatchman

01-22-06, 01:07 PM

Maybe someone else could try 720p to confirm this. (Pls recall the earlier caveat re writing down the original value for MTRX before changing it, since this control is automatically saved w/o being written to memory. 0-Enter won't work on MTRX either.)

Ok, had a chance to have a "session" with the set last night -- Update to my last post :

Checked 720p from a broadcast ATSC source, and 720p QM Color bars, and discovered that on XBR960 at leaast CXA2171 MTRX, CBGN, CRGN+YGN affect 720p ATSC as well.

So they effect 1080i ATSC, 720p ATSC, 480p ATSC(according to internal QM Color bars) and 480p NTSC(such as from a DVD player via component). I have no way to check 480p from internal tuner, but expect that would be effected as well.

Default values for my set/some notes :

MTRX = 0 - It's default was 0 for internal ATSC tuner 1080i/720p, 0 using 1080i/720p,480p, 0 for memory stick/1080i, 0 for a 480p signal from DVD hooked up to component 5. Didn't change from Zero no matter where I "went" in fact. I don't know about HDMI as I'm not using it.

CBGN = 4
CRGN = 5
YGN = 5

On my set at least, only one value appears capable of being stored for all of those. Including whether or not the signal is ATSC or NTSC for 480p - one value HAS to be used for either. Although Note that the XBR960 servicecodes PDF lists the default for MTRX V5/V6/ATSC/1080i/720p As "1", HDMI as "1" and "others" as 0. I wonder what "others" is, I certianly can't find it ...

Just in case, note that I tried changing the values to see if I could get any "different" values for 480p or 1080i or 720p/etc. to "stick" but it wouldn't work. Only 1 value was allowed.

The following is with RYR~GYB 13-15-5-3/Color=31/Hue=0, and using RGBS to turn on the RGB guns indivdually -- Which is right on the money(more or less) for 480i via component, with DVE or AVIA, as well as as far as I could tell, with QM test pattern Color bars :

I noticed, with the default values for MTRX/CBGN/CRGN/YGN, using either 1). 480p from DVD and AVIA Color tests, and/OR 2.) the internal QM color bars for 480p/720p/1080i that 480p was "closest" of 3 three scan rates to having the correct Color balance. For instance, With the internal QM color bars for 480p/720p/1080i, It was easily obvious that using "green" that all the bars that showed up were extremely close to matching each other perfectly with 480p, but this was NOT the case for 720p or 1080i. It was very also apparent that the internal QM color bars for 720p/1080i were showing the same thing - at least with my capability to compare them, since I can't have both on screen at same time. Red seemed a little closer to being similar between 480p + 720p/1080i, and seemed closer to being correctly balanced for all 3(but off enough to be noticable even with the QM Color bar patterns).

Noticed that, for BOTH 720p and 1080i, changing "MTRX" value to "1" made the "green" test perfectly balanced(as much as I could tell anyway) with the QM internal color bars for BOTH 720p+1080i, but Red, although perhaps "closer" was still noticably off a bit. HOWEVER, changing MTRX to "1" then messes up the "green" test quite significantly for 480p - either with Avia, or the internal QM test bars - Not as much difference using Red, but there was a difference for the worse there as well.

With MTRX at 0,(default) Changing CBGN to 6, CRGN to 6 and YGN=5 worked very well for 480p Color according to AVIA tests, if anything, May have improved things slightly for 720p/1080i as well.

For now, I left MTRX at default(0) and set CBGN/CRGN=6, YGN=5 for best results with 480p and set up a 2nd set of RYR~GYB values for 720p/1080i so I can select it from the User menu Color Axis setting(default or Monitor) via adjusting those the best I could using the 720p/1080i test patterns+RGBS. I came up with 13-13-12-8 for RYR~GYB for 720p/1080i internal QM test patterns, and have that one set for the "monitor" setting currently, ad 13-15-5-3 for 480i/p set for "default". Do note that you can only work with the "default" setting when using the internal QM color bars, I tried setting it to "monitor" before I went to QM, but it wouldn't "stick". Which does make me wonder if there are any SM values the QM color bars "ignores" that also may be important here.

What I really need is to catch a local broadcast station that's sending 1080i or 720p to put up color bars and check and see what really happens with the internal ATSC tuner -- BUT, I did luck out a little bit last night and PBS HD had up a program with scenes with all different varities of Greens(Trees/Grasses/etc) and, seemed like my new "monitor" color axis setting was better, and "green" push/less natural looking greens/fewer shades of green in fact. seemed evident using "default".

That's really not very ideal way to do it though, since I use 720p/1080i and 480i ATSC, as well NTSC signals via the internal tuner and would have to manually switch between "default" and "normal" color axis everytime I "switch channels" more or less ... depending on how things go, either using the QM test patterns, or even better if I catch color bars up from a 720p/1080i broadcast source, probably better to change "MTRX to 1, and adjust CBGN/CRGN/YGN accordingly for best results for 720p/1080i, and set up a different "Color axis" for 480p(if it "works" that is), which all I'm using it currently is for DVD. Just would have to remember to change "Color axis" setting for my DVD input depending upon whether I'm sending 480i, or 480p from the player ...

So,what's going on here? Am I missing something? Seems that Unless the internal QM test patterns aren't showing me something "accurate", and/or unless there are "specific" 480p vs 720p/1080i color decoder settings elsewhere I'm not aware of, it doesn't appear that, for instance -- using the same values for RYR~GYB that I can get accurate color decoding for all three scan rates ...

If I could set "MTRX" to 1 for 1080i/720p, and to "O" for 480p and adjust CBGN/CRGN/YGN seperately for 480p+ 1080i/720p I'd probably be able to at least get it pretty close without needing different values for RYR~GYB, but that doesn't seem possible, at least on my set.

Funny that I was just complaining earlier about having to many SM settings to play with, now it seems I'm asking for MORE of them ....

By the way, the slight hue shift I'm seeing in my "twin view" grey background for ONLY 480p+1080i sources as compared to everything else - including 720p(as explained in last post) isn't effected by any of this, so something else is going on there. My Twin view grey background is the only place it seems to be an issue, don't notice anything like that anywhere else. Maybe should go back and take a look at 2170P1 #2~6 (YOF/CBOF/CROF/etc) again when I get a chance.

Any thoughts/comments/corrections are appreciated, thanks.

KenTech

01-22-06, 02:21 PM

Ok, had a chance to have a "session" with the set last night -- Update to my last post : <snip>I, for one, am grateful that you are posting in such detail, although, you're right, it's fairly hard to read. But I am paying attention. I want to comment, as I can, having a 36XS955 without TwinView, etc. I am concerned about the digital colorspace vs NTSC, but I don't want to obsess too much. I will read and respond as I can, thanks. It should be clear from my last several posts that my current obsession is getting the *image* right -- meaning that I want (a) to have some settings that *reproduce accurately* what is on a DVD, good or bad, and then (b) to understand what to tweak to "fix" things I find ugly, if possible. Without knowing what the set is doing, however, one gets lost. My use of the previously posted PDF chart in #707 has helped, and the settings published are oriented toward the goal of *neutrality* in rendering image detail.

A couple of short comments:

(1) I recommend that you make some color comparisons between sources with true b/w material, not easy to come by for HD broadcast, for example: One has to have some faith that a film segment in b/w on PBS, for example, is actually color-free, without the greenish tinge that sometimes affects such segments (that disappears when you tuen down Color).

The reason to do this is to perfect the color-offset balance among the inputs, using 2170P-1: CBOF and CROF. That was never correct on my set, especially for HD, and I have fixed it nicely over time. Comparing a true b/w source with the user Color control at minimum vs maximum should yield the *same* screen color.

When I got the set, I noticed that HD broadcasts had yellow-greenish shadows fairly consistently. The columns for YOF thru CROF for MS/ATSC 720p and 1080i were originally 10-45-47 and 10-44-47, respectively. Now they are both 7-55-55, and I seem to have nothing to complain about when b/w is broadcast on HD (PBS, some commercials).

(2) One has to be careful if using an upconverting DVD player to make colorspace adjustments. *If* the player outputs 1080i and 720p with the correct colorspace for that video standard, you're fine. But, as reviewers have noted, some players don't, simply using the NTSC default. I think my Panasonic S97 does do it correctly. Further, I guess one has to assume some things about the internal pattern generators, which are all-digital. But what? That the HD patterns are HD colorspace and the SD are not? I can't say.

Even if I have a few mismatches, I am loathe to obsess over them. It's DVD playback I want to get right, since color on well-mastered DVDs has been tweaked to near-perfection. Braodcast, even HD, may not be accurate, and I am thrilled with the near-photographic rendition of the material. I guess I am willing to forgive a few color-matrix faults; but, like you, I ultimately would like to understand them and get them right.

Nitewatchman

01-22-06, 03:56 PM

I, for one, am grateful that you are posting in such detail, although, you're right, it's fairly hard to read.

First Ken, thanks much for the response and your thoughts on this.

Lack of sleep lately certianly didn't help the "clarity" of my previous two posts on this thread, sorry about that.

In additon, I have found it is difficult for me to post about some of these issues and use more precise and consise language at times. Sorry, I'm doing the best I can to not only be complete and accurate, but also make it "understandable", I will continue to strive to improve in this regard, especially where the latter is concerned.

AS just one other example of this, I think you'll find you referred to "SSMD" in a paragraph where you were referring solely to "3dcomb" settings in a post on a previous page concerning image processing settings, without mentioning SSMD is 2103-1 setting. I was able to figure out what you meant, but I'm not sure everyone will be, as I don't think you mentioned anything about SSMD elsewhere. BTW, Default was "0" for all inputs for SSMD on my set(any input/scan rate), not 3, so it was a relevant thing for me, at least to some degree.

It should be clear from my last several posts that my current obsession is getting the *image* right -- meaning that I want (a) to have some settings that *reproduce accurately* what is on a DVD, good or bad, and then (b) to understand what to tweak to "fix" things I find ugly, if possible. Without knowing what the set is doing, however, one gets lost. My use of the previously posted PDF chart in #707 has helped, and the settings published are oriented toward the goal of *neutrality* in rendering image detail.

I can't say enough concerning my appreciation for your work in this area. As I said in a previous post, I'm using the image processing settings you've found+posted recently, and previously was also using what you had found(for the most part) in your previous posts on this. In fact, I have been reading/following the posts on this thread closely and "tweaking" my set for over 7 months.

(1) I recommend that you make some color comparisons between sources with true b/w material, not easy to come by for HD broadcast, for example: One has to have some faith that a film segment in b/w on PBS, for example, is actually color-free, without the greenish tinge that sometimes affects such segments (that disappears when you tuen down Color).

The reason to do this is to perfect the color-offset balance among the inputs, using 2170P-1: CBOF and CROF. That was never correct on my set, especially for HD, and I have fixed it nicely over time. Comparing a true b/w source with the user Color control at minimum vs maximum should yield the *same* screen color.

When I got the set, I noticed that HD broadcasts had yellow-greenish shadows fairly consistently. The columns for YOF thru CROF for MS/ATSC 720p and 1080i were originally 10-45-47 and 10-44-47, respectively. Now they are both 7-55-55, and I seem to have nothing to complain about when b/w is broadcast on HD (PBS, some commercials).

Yes, I've read and looked at your previous posts and what values you're using for these and as I said, plan to revisit this again to make sure I wasn't just "seeing things".

[updated/corrected/added info following paragraph]

In the past have checked it when looking at "B&W" with Color slider at "31" or higher and "0"(latter for greyscale) and checking CBOF/CROF adjustments. Also checked various inputs/scan rates as much as possible concerning CBOF/CROF. Not seeing any "problems" I didn't spend a lot of time on it, however. Luckily, at the time I was looking into it I ran into a PBS documentary containing much WWII B&W footage that didn't seem to have that added "greenish"/etc. tint to it. Also, luckily, one of my local PBS HD affiliates transcodes PBS HD from 1080i to 720p for broadcast, the other 3 PBS HD stations I receive sends 1080i. At that time, they were running the same programming at the same time, now the one sending 720p time shifts PBS HD programming and airs it per their own schedule.

[end update]

Even changed the values in some cases to see the effects, but CBOF/CROF settings from the factory for my set seemed to not need much adjustment. No pinkish or greenish contamination/etc. or visable differences between inputs/scan rates in the "B&W" content that I noticed. But, maybe I didn't look/check this "closely" enough+will try it again.

Which is one reason why I wanted to emphacize the slight hue change to the grey background(ONLY for 480p/1080i) I have set up for "twin view" is the only place I've noticed an issue with it - But because of that, and because I had already "checked" to see if I needed different offsets for CBOF/CROF, this is why wondered if it might have been caused by differences in color decoding for 480p/1080i, which as I noted in my last post doesn't seem to be the case.

I did need to adjust Blue/Green DRV/CUT values slightly from the factory "warm" offset for good greyscale - , as the factory setting was more "greenscale"(via all inputs/scan rates/etc) than "greyscale", but I'm guessing (just guessing and "eyeballing" though) the factory "warm" offset wasn't that far off from 6500K.

So, I haven't noticed any visable differences between "B&W" between different inputs and scan rates, and, in addition to the rare times I've been able to do that with Color at 31(I certianly admit I need to try to do/find suitable content for more of that), I have also spent quite a bit of time looking at greyscale with color at "0" as well.

BTW, I have the factory "warm offset" default set up for "warm" for reference and my new values drv/cut values(I didn't change RDRV/RCUT from factory setting) set up for "neutral" - I had no use whatsoever for the factory "neutral" and "cool" values or anything near them, looked way too cool/too blue to me.

YOF~CROF 10-44-47 are my set defaults for 1080i, they seemed right on, or pretty close. I recall I had tried some of the previous values you were using for your set, but they added visable "pink/red contamination." But, will revisit this again for all inputs/scan rates again per your suggestions, next on my list.

[updated section follows]

Had to dig up my notes on this - And this is where it gets strange -- 720p defaults YOF~CROF on my set are 10-31-31 - quite different than the servicecodes listing and default on your set. 480p defaults 13-44-42, as with 1080i, same as servicecodes listing. P2170-1 CBOP/CROP defaults are 9-36, quite a bit different from servicecodes listing default of 36-37. I have them all at the defaults, currently. Remember puzzling over the significantly different(from service code file) defaults for 720p CBOF/CROF, and CBOP/CROP about 5 months ago and why I wasn't seeing any "problems" from it, but -- looks like I'm going to have to go through it all again ! LOL ....

This probably explains the bit of reddish tint I'm seeing only from *-480p+1080i sources in the Grey Twinview background I set up, I had it as black background at the time I was checking YOF~CROF to see if I needed to change things, so wouldn't have noticed it then.

* - but not 720p, it's exactly the same shade of grey as 480i or NTSC SD from say, the internal tuner, - and not surprisingly, 480i/etc. are CROF/CBOF set at 31-31, same values as in the servicecodes PDF.

One thing it doesn't explain is why I'm not seeing any sort of "contamination" of color while viewing program material, including B&W content via 720p+1080i.

Another thing it doesn't explain is why I get the exact same results(as far as I can "compare" them anyway) with the 1080i+720p Color bars from the internal QM test patterns, unless CROF/CBOF are ignored if you are using one of the test patterns in QM.

I could be wrong, and will investigate it further when I get a chance, however -- Even IF I can find "better" settings for CROF/CBOF for the various inputs/scan rates, I don't think it's going to change this color decoding "difference" between 480p and 720p/1080i to much of a degree - the difference isn't there(according to the QM test pattern color bars) between 1080i and 720p, just between 480p and 720p, and 480p and 1080i.

[end update]

(2) One has to be careful if using an upconverting DVD player to make colorspace adjustments. *If* the player outputs 1080i and 720p with the correct colorspace for that video standard, you're fine.

Don't have one of those yet to experience problems with <g>.

What I did to calibrate decoder on another set(Toshiba 34HF84) and for 1080i/720p and the particular device I'm using was to use Color bars put up by broadcast station with the output of my ATSC receiver/STB I have hooked up to it(Zenith HDV420).

I have also had the chance to check those against color bars from a different station/completely different source nothing needed adjustment/changed. But luckily, the Toshiba decoder's "roughly equivilent to RYR~GYR and "MTRX" settings"/etc. are SCAN RATE specific, and adjustable individually for 1080i, 720p/480p. Can't adjust those differently/get it just right for both both 480p NTSC output from DVD and 480p from The ATSC receiver, however - strangely enough, the "manual" that came with this ATSC receiver says it outputs "NTSC standard", maybe they're just talking about the composite/s-video outputs.

Again, should also note that as I mentioned earlier I used the same DVD player to calibrate color decoder in the toshiba for 480i/p, and unlike is the case with the XBR, were no differences in decoder settings(or tint/hue offsets/etc) required for 480i vs 480p.

Even if I have a few mismatches, I am loathe to obsess over them.

It's not that I'm all that much "worried" about it either, -- not like this is all that "BIG" and a glaring "mismatch either", although some might thing so ---

Thing is though, I've been able to, for example adjust another set I have for I think, excellent color accuracy in this regard, and given that I do find accurate color reproduction for HD broadcast sources(even though as you say it "varies" greatly, even from the source) important, the difference does show to me - especially, as it turns out in my case currently -- where greens are concerned, such as the field at Mile-high stadium(or whatever they are calling it) via CBS HD currently, or all those greens of the "forest" from a PBS HD channel documentary that was running last night ....

I guess I am willing to forgive a few color-matrix faults; but, like you, I ultimately would like to understand them and get them right.

Agree 100%.

Costas

01-22-06, 04:28 PM

The links for the pdf files don't work, is there another source for these?

Nitewatchman

01-22-06, 06:04 PM

Costas,

If it helps any, I've never had any problem with those. What I usually do (with IE 6) is right click and chose "save target as", then view them off my hard drive after I've downloaded it.

Q of BanditZ

01-22-06, 06:06 PM

Wow! Nitewatchman AND KenTech in the same thread! This is a dream team, folks! :)

Costas

01-22-06, 07:40 PM

Nitewatchman,
Thanks for the tip, I got the pdf's. Now the endless tweaking will commence....

KenTech

01-22-06, 11:55 PM

A). For the set defaults, MID5 Columns #53~#56 are actually assigned to P2170-3 MIDE#16. -- It's really silly I think, as for instance, Columns #53~63 are all zeros at the factory defaults, however, #53~56 are assigned for MIDE for the different Pic modes for Twin-view feature ... #53=Pro, #54=Movie,etc. So, just something for XBR owners to keep in mind when deciding which MID5 column's to take over.I want to understand what you are saying here. I don't get the reference to #16.

Further, you are right about Twin View MIDE pointing to MID5 #53-56. You can solve that issue easily: Defaults for MID5 cols 53-56 are all zeroes. I have designated col 63 permanently as all zeroes. So . . . simply set 2170P-3/MIDE for all Twin View picture modes to MIDE=63, and you have the exact equivalent setting -- all zeroes. Now cols 53-56 are available for other things.

There is a great deal of redundancy redundancy in the MID5 assignments. :-) A little consolidation opens up columns you can assign toy our own uses. Turns out I didn't think about this, as my set doesn't have the Twin View feature. Good that you brought that up, thanks!

ADU

01-23-06, 01:08 AM

JeffD2.

2170D-3/JUMP should be the parameter you're lookin for to expand HD to fill the screen on the 36HS510. If you need to tweak vertical sizing and/or position on the expanded HD, some folks use the MID controls. I found the various sensitivities and relationships in those controls rather complicated and easy to screw up though, so I've been using 2170D-1/ASPT for vertical size and 2170D-1/SCRL for vertical position instead. I don't really know which is preferable. 2170D-1/ASPT and SCRL are signal and aspect-ratio sensitive on my TV (which is actually 16:9 rather than 4:3), so they can store separate values for these different sensitivities. Hopefully there's some more info about these and other possible controls in one of the various spreadsheets, so you won't be flyin totally blind. If you use ASPT and SCRL jot down the original values before changing them (including also the signal and aspect ratio), so you can put them back if you need to, because there appears to be no easy way to automatically reset the data back to the TV's original OOTB condition w/o removing crucial factory tweaks. DO NOT mess with the NVM resets! And safe SM-ing.

ADU

01-23-06, 01:17 AM

Nitewatchman,

Tks for confirming that 720p is also effected by the CXA2171 controls. I'm still readin your and Ken's posts, tryin to digest it all, and will try to post back with a few thoughts soon.

Nitewatchman

01-23-06, 01:22 PM

I want to understand what you are saying here. I don't get the reference to #16.

In the section from my post you quoted, I meant 2170P3 #16 MIDE, where the MID5 columns are "assigned" - I.e. Where Twin View MIDE "pro" picture mode is "pointing to" MID5 Column #53 (all zeros). But thought that was clear? If it wasn't, this probably isn't going to help any, either - Sorry.

But, maybe, instead however you are reffering to another comment in my post you didn't quote, concerning P2103-1 #16 SSMD, and your comments about it in the last sentence of a paragraph(d) concerning 3dcomb settings in post #707 - I quote that paragraph below, for convienience :

(d) For all S-video, composite-video, and RF-tuner 480i sources, the 3D comb filter separate color from luminance, and the 3D-COMB section has significant sharpness tweaks, nos. 14 and 15: VAPG and VAPI for vertical anhancement (generally not used, which means zeros); and nos. 17 and 18: YPFT and YPFG. In this case, YPFG can act as either a detail-reducer or enhancer, with no effect at YPFG=8. (Possibly 7; still experimenting.) Note that the 3D-COMB settings have no effect on all high-quality sources, such as HDTV, HDMI, and component inputs. Recommended: Set VAPG and VAPI to zero, YPFT to 3 and YPFG to 8. Make sure #16/SSMD is set to 3.

----------------------------------

You can solve that issue easily: Defaults for MID5 cols 53-56 are all zeroes. I have designated col 63 permanently as all zeroes. So . . . simply set 2170P-3/MIDE for all Twin View picture modes to MIDE=63, and you have the exact equivalent setting -- all zeroes. Now cols 53-56 are available for other things.
!

Absolutely. I just didn't see any need to add another "all zeros" column. So, I use #53 for my "all zeros" column anywhere I want to use it, and I also have all Twin-View Pic modes set to "point towards" #53. #54~63 are thus available in my case for my own "settings" whatever I want to use them for. Thought I explained that, but perhaps not very well.

For that matter, haven't checked it to make sure, but if the servicecodes listing/PDF is correct, looks like I could use #49-* for my all zeros column(and for twin view) and then free up #53 for other uses as well.

* - Correction to previous post - I had thought the other <53 all zeros MID5 column indicated in service code file was #50, looks like I glanced at that wrong, as I now see its' #49.

There is a great deal of redundancy redundancy in the MID5 assignments. :-)

You're not kidding ... I will Never have any use for the factory settings for the "vivid"/"standard" and "movie" pic modes, or likely the MID5 columns they "point towards" -- Only left "pro" MIDE #16 at default setting for reference for all inputs/scan rates/etc, as I took over all the other pic modes+using MID5 column#53~63 everywhere for 2170P3 #16 MIDE except for "pro" picture mode.

So, If I was a real "neat freak" or something, I could REALLY reorganize the MID5 table and free up a lot of Columns for my own use ..... I don't think I'll ever have a need for that, though(at least I HOPE not)

Like you, I have "movie" set up with different P21704 Gamma setttings, I use my settings+"standard" mode primarily, for most viewing and "vivid" for experimentation. One thing I have found is that it works very well for "comparing" new things you want to try when you have the "pic modes" set up the same way, except for your new "experimental" changes. With "pro", then "vivid" than "standard" right next to each other+ using standard and pro "for reference" and Vivid for "experimenting" makes it easy to see improvements or "problems" with new experimental settings. Then when I see improvments I like, I'll move them to standard+Movie pic modes.

Which reminds me, I think it's definitely worth the effort -- but what is a bit of a "pain" about "Hijacking" the pic modes, and also modifying the "image processing settings" concerning "adding" a new device that's using a input/scan rate/etc. I hadn't used before.

For instance, I added "memory stick" for the first time/used it for the first time about a month ago. Since it had been so long since I'd "hijacked" "Vivid"/"standard"/"Movie" for my own purposes for all the inputs/scan rates/etc. I was using at the time : in addition to setting P21703 settings to how I want them for all the pic modes for "memory stick", I had to check or modify such settings as the 2170p4 "gamma" settings for the memory stick pic modes as well as check to make sure any other settings(such as the relevant 3DNR settings for example) that had needed to be changed for making the pic modes "the same" as pro, or any other settings I'd modified aren't "picture mode" and "Input/scan rate" specific or make any changes when necessary.

Although I have good notes on this and can look at the servicecodes listing(to some degree), I went ahead and checked everything "relevant" anway, just to make sure I wouldn't miss anything. But, for the next time(such as if I ever use "HDMI" which seems likely), I should make an attempt to "streamline" my notes and note exactly which items will need to be checked or changed ... Maybe next time, although I'll probably still end up checking everything just to make sure ...

On another note relating to this ... Unlike 480p V5/V6 "settings" which I have changed I haven't changed the pic mode/image processing settings/etc for 480p ATSC from the internal tuner as I don't have a local station sending 480p. So didn't *think* I had a 480p source to use to be able to modify those --- However, I noticed last night that when using the internal test patterns located in "QM" that the 480p ATSC "specific" settings I checked were at their factory defaults. Therefore I think I should be able to use say, the 480p internal test patterns from "QM" to set all those settings so if/when a 480p OTA ATSC signal "pops up" I'll be able to use my "preferred" settings with it(well, except maybe for color decoding - more on that later)... Duh! Why didn't I think of that before ... Even another item to add to my "do it list" ....

Only thing is, you can't access user menu or "choose" which Pic mode is "active" when you have a QM test pattern up - I didn't check it but hope it's whatever pic mode you're using when you turn on the QM test pattern ... That may not be the case however, as for example I've also noticed that whether or not "default" or "Monitor" is selected from User menu "Color Axis" setting before the QM test pattern is turned on, It stays, or "reverts" to "default" anyway. In other words, when you turn on a QM test pattern, seems that only the settings for RYR~GYB for "default" are accessable - even if I'd had "monitor" selected just before going to the QM test pattern.

Well, my plan had been to make this a short post, sorry it didn't turn out that way! Especially as I have another post to do concerning more I found out about the "color"/color decoding adjustments last night ...

Cleanupthat

01-23-06, 04:17 PM

WOW,
I am noticing more problems with my hs420 set that i got a week ago. Last night while playing COD2 on my 360, i began to notice a green tint on a black backround when there is a light by the dark backround, when moving. NOTE: I only notice this tint while moving. When you are stationary or not moving there is no green tint. It as if when the black or dark image comes in contact with a white image, the white or light image becomes green over the dark image. It almost seems like lag or something with these colors. Does anybody have a clue what is going on here? I have searched for my problem on AVS and have seen a couple other people with a kind of similar problem. BOTTOM LINE is I am very dissapointed in these HD CRTs. I have owned 3 since christmas. I have owned 2 sanyos both with horrible geometry (that is pretty much impossible to fix.) I then turned to the sony hs420, expecting no problems, no distortion, and well a perfect set. I have been reading that is impossible to have berfect geometry for an hd crt. Well that is BS, I need perfect geometry, or at least something reasonable. (Read my other posts in this thread on my other problems that are unfixable unless you apply magnets.) WTF I am not applying magnets EVER.

I am a gamer who plays for cash and i just thought going HD would not be this bad of an experience. I keep reading how HD CRTs are the best for gaming, and how they have the best picture. I just cant take all these problems!!! I dont want to have to lug this 150 pound monster down the stairs, then exchange it and lug another one back up, and find even more geometry distortions, and other problems.

I think i am just going to get my money back and get an lcd a couple months down the road. This is unacceptable for me. I mean my 20inch SDTV is wonderful, I think i am just going to go with that for now.

I mean am i the one in a million who gets these horrible sets with horrible geometry, that are unfixable in the service menu!!??? IT SURE SEEMS THIS WAY.

ANY WAY/ End rant. I appreciate the help you guys have given me in this thread on some of my other problems. The fact is i have been in the service menu over 30 times and cannot seem to fix my problems at all.

If anybody would know how to fix this green tint i see from a dark backround, when a light backround moves at it (slowly or fast), I will really apprieciate any help on this. THANKS...

PS( I have tried different cables, I tried the other input, I tried playing in 480p, I tried another game, And it is still noticable. The only weird thing is i never noticed this tint since yesturday, when i had this set for about a week!) Thank you.

Nitewatchman

01-23-06, 04:51 PM

Tks for confirming that 720p is also effected by the CXA2171 controls. I'm still readin your and Ken's posts, tryin to digest it all, and will try to post back with a few thoughts soon.

Looking forward to hearing your thoughts, and appreciate your posts on this so far. I'd read the discussions earlier in this thread related to CXA 21x1 settings and differences between ATSC/NTSC signals/etc, but it's been quite some time ago and I should review those again in more detail.

One thing I'm especially interested in hearing about is if other folks' are getting anything like the same sorts of results as I am on this with my set. Also, would be interested to know if other folks sets came with a MTRX default of 0 or 1, and perhaps if others are getting along the lines of what I'm getting with CBGN/CRGN/YGN as well.

I'm also wondering if I'll run into any similar issues if/when I use HDMI input and 480p or 720p/1080i sources. Looks like there are seperate settings for the values that might be important for HDMI for the most part, but I've learned to take the servicecodes listings with a grain of salt when it comes to my set and certian settings.

What I probably should also do "someday" is hook up one of my other external(STB) ATSC receivers* to XBR960 via component and see what happens, although to tell much of anything I'd need a station sending 1080i or 720p to put up color bars/test patterns to tell much of anything and I don't get the oppurtunity to come across that too often.

* - They'll output/upconvert or downconvert as necessary either 1080i/720p/480p or 480i as desired, doesn't matter what the source is sending.

Looked at it a little more last night, follows should hopefully explain it as well as summerize things. Sorry guys, I really *tried* to make this short as possible(it's not like I enjoy writing long posts or anything), but thought it better to post on this in as much detail as possible.

Part I - CXA2171 - MTRX - ATSC/NTSC or 480p vs. 720p/1080i - what's it all about?

A.) - The "evidence" -- From what I've seen it seems like the differences I'm seeing in color decoding/color balance between 480p vs 720p/1080i perhaps have more to do with 480p vs 720p/1080i than ATSC vs NTSC, and that the MTRX setting is the real "important factor" here, and that I really *need* to be able to use MTRX=0 for 480p, and MTRX=1 for 720p/1080i, but I can't do so, as only 1 value is available for all of those.

Could it be (and I do recall perhaps reading something about this) 480p is more using NTSC specs/Color space, and 720p/1080i is using ATSC specs, and perhaps that's why it looks like it should be MTRX=0 for 480p and MTRX=1 for 720p/1080i ?

Consider the following.

A#1). Note that using AVIA/DVE Color tests+RGBS -- RYR~GYB 13-15-5-3 (or 14-14-5-3 - haven't decided on that difference yet), gives me best results for color balance/accuracy with 480i NTSC, and 480i ATSC according to internal QM color bars for color balance using AVIA or DVE tests - i.e. sources that do not use CXA2171 MTRX/etc.settings. For 480i via component, it's pretty much right on with color slider=31, Hue=0, and no changes from the factory "o's" for UCOF/UHOF in P2170-3, nor was it necessary to modify SHUE/SCOL in P2170-4 from factory defaults. Note that I have RYR~GYB 13-15-5-3 set up as the "default" color axis choice from XBR's user menu.

That's pretty much the case for S-Video+composite inputs as well. For Svideo however, for best color balance with Color=31, Hue=0, the factory defaults for YV/VC are UCOF=2, UHOF=3 - and that's exactly what is needed. For composite video sources, need to change Tint to G1 for best results.

A#2). With 480p, Using the following values and either AVIA/DVE color tests from DVD OR the 480p ATSC internal QM color bars -- and this is what I think may be signifcant here -- I get the same results for color balance from both 480p DVD+AVIA or DVE, OR 480p internal Test patterns : Color balance Looks very close to "perfect"(It's as close as it will/can get anyway) with these settings, I tried to tweak them farther, just to see, but things are "right as rain" with these settings, anyway. However, color balance is equally "off" via AVIA/DVE+DVD at 480p or internal 480p test patterns If I use the values listed farther below in section A#3 :

MTRX=0 (my set's default, although I have a feeling default should have been 1)
CBGN=6
CRGN=6
YGN = 5

P2170-4

RYR=13
RYB=15
GYR=5
GYB=3

Color=31
HUE=0

A#3). With 1080i/720p + using the values from A#2 above -- Color balance is "off" quite a bit -- especially when using RGBS#2 (Green) ... And, it's off the same amount for Both 1.) the 720p/1080i QM internal color bars AND #2). Color test patterns from DVE(I think from DVE) which ken provided earlier in this thread, and which I used from Memory stick(1080i). However - Using the Following values listed below next paragraph -- Color balance is "right on the money" for both the QM 1080i/720p test patterns AND the color test from memory stick - especially note that everything is the same except for "MTRX" - MTRX is the only thing that changed.

In fact, I even tried "tweaking" both RYR~GYB and CBGN~YGN farther to see if I could find better values using the DVE Color test patterns from Memory stick(1080i), and couldn't find anything better -- even went outside Of SM and checked it via memory stick with the filters that come with DVE and using the tint/hue/sliders and it showed to be the best results(everything looks VERY good color balance wise) I could come up with using tint/hue sliders. :

MTRX=1
CBGN=6
CRGN=6
YGN = 5

P2170-4

RYR=13
RYB=15
GYR=5
GYB=3

Color=31
HUE=0

----------------------------------------------------------

B). What I'm doing about it : Maybe there is a better solution, and I'm just "missing" some 480p vs. 720p/1080i specific setting. Note : also see PartII concerning P2170 YOF~CROP notes concerning why I don't think those are an issue involved with this. Or, maybe there's something else going on and I'm just chasing up the wrong tree, and/or there's something really odd going on with "my set" on this which others aren't experiencing.

Concering this issues, The following two choices (B#1+B#2) seem to be my choices for dealing with the issue while still being cable of achieving good(and for the most part excellent) color balance/accuracy for all my input sources. For various reasons(at least currently), I think the best choice for me is #1, which is what I'm currently doing.

B#1). Set MTRX=1, for best results concerning color accuracy/balance with 720p/1080i for use with my "default" Color axis user menu choice, which points to RYR~GYB 13-15-5-3 - In other words, the values listed in section A#1. For 480p - set up alternate RYR~GYB settings of 15-13-0-0 using "monitor" -- as the 2nd "Color axis" setting selectable via user menu. Currently, This allows me to use "default" Color axis for every other source I am using except DVD player outputting 480p via component 5 input. And, since I generally watch DVD's with the player outputting 480p, most of the time it will stay "set" using the different RYR~GYB settings I have set up for it via "monitor" Color axis choice in user menu. Or, I might decide to output 480i from DVD player the vast majority of the time along with the "default" color axis setting from user menu more often for better Color accuracy - That would certianly completely take care of the issue for me -- at least for now ...

Unfortunately, however, I don't think this is going to work out very well if a desired 480p ATSC OTA source "pops up" again in my area. I could be wrong, but given I'm seeing the same thing from internal QM "ATSC 480p" test pattern as I am with NTSC 480p output from DVD player -- I expect that for best results for 480p from ATSC tuner I will probably also have to change the user menu "color axis" setting from "default" to "Monitor" for best color balance/accuracy when I'd switch to that channel, and then change it back to "default" when I switch channels elsewhere. And, there used to be a couple of those here, sending "Fox Widescreen" and 480p, but luckily it's been 720p and "FOX HD" since september 2004.

Also, unfortunetly - for 480p and "MTRX=1" I'm not getting as accurate results using AVIA/DVE or internal QM color bars/tests using RGBS=2 "green". That's even with setting up the best "RYR~GYB" settings as 15-13-0-0 for best results for this as well as also(thought I should try it and see what happens) trying to tweak CGBN~YGN for better results -- I couldn't find better values for CBGN~YGN BTW -- Still best at 6-6-5 ....

It is somewhat close howeever for 480p - For instance, If I change Color slider to "34", it's pretty much right on, although Balance using Red or Blue gun is pretty much spot on with Color=31, Hue=0. this might sound odd, but seemed like If there were negative values allowed/allowed/possible for GYR-GYB settings, I could probably have gotten it right on the money.

Also Note that at first, I was getting slightly better results for Blue/Red with Color=30 so I did a little "checking" and, given the servicecodes listing and what I'd experienced with other scan rates and experiments - was surprised to find that I could(and did) change P21704 SCOL to "31" from it's default of "32", and only effect 480p (and I THINK just via V5/V6 not ATSC but not sure) without effecting other scan rates/inputs. This way, I can leave Color slider at 31, and just change the Color Axis setting for DVD player, depending upon whether I'm sending 480i(use "default" setting= RYR~GYB 13-15-5-3) or 480p(use "monitor" setting - 15-13-0-0 - to account for the "MTRX" difference involved) from the player.

------------------------------------------------------------------------------

B#2. My other choice is to set MTRX at "0", which works perfectly for color balance for 480p and my "default" color axis setting(RYR~GYB 13-15-5-3). And, for 720p/1080i set up 2nd set of settings for "monitor" color axis setting to use with it. In this case, RYR~GYB 11-14-13-8 gave me best results with color bars via memory stick or internal 720p/1080i QM Color bars. Although not as close to "perfect" Color balance as I get for 720p/1080i with MTRX=1 and RYR~GYB 13-15-5-3, it appears to be very close to it.

I don't find this option "workable" however, because of the issue with OTA/internal tuner and for best color accuracy needing to change the "color axis" setting in user menu everytime I (or other members of family/etc) switch to/from watching a 1080i/720p source(which is a lot) to a 480i ATSC or NTSC OTA source, which is also a lot. Almost, but It also doesn't quite give me the "extremely close to perfection" color balance for 1080i/720p that using MTRX=1 seems to give. And, I did even notice the slight difference last night during HD "Smarttravels" from PBS HD. Although, on the other hand, -- the "default" color axis(13-15-5-3) in this case is a tad bit more "in the ballpark" for 720p/1080i than using Option B#1.

Also note that again -- when exploring the options for Choice B#2, again, I tried "tweaking" CBGN~YGN (and adjusting RYR~GYB where necessary) for better results, in this case with 1080/720p, and, in this case with MTRX=0 couldn't come up with anything better for 720p/1080i than RYR~GYB 11-14-13-8 and CBGN~YGN -- You guessed it ... 6-6-5 ... Also, Best results did this time occur with "color"=31 and "hue"=0.

-------------------------------------------------

Part II - YOF/CBOF/CROF(and just in case CBOP/CROP) notes :

II#1). Just to see, temporarily changed 480p/1080i P2170-1 CBOF~CROF to 31-31, and the "pinkinsh tint" with my "grey background"(set up via MID1 #24BCOL=6(default 0 - black)) went away when selecting a 480p/1080i source in Left-Twin view window -- So, that's what was causing that --

In other words, if CBOF~CROF is set the same for all inputs/scan rates/etc, the "hue" of my Grey Twinview background doesn't change when selecting among sources in the left window - obvious that sources displayed in left window is where it's looking at the SM settings for the sources being displayed, not the right twinview window it seems. Of course, I'm not going to set CBOF/CROF based on that - so, just one of those quirks I suppose. Perhaps may be one reason why they decided to have factory default for BCOL=0.

II#2) Didn't have time/didn't run into any suitable B&W content to look at -- However -- I did check YOF/CBOF/CROF(and tried CROP/CBOP just for heck of it as well) with all sorts of extensively different values using 480p/720p/1080i color bars/test patterns from AVIA/DVE at 480p, Internal QM test patterns, AND color bars from DVE but via Memory stick/1080i which were posted early in thread --

Did that using RGBS 1/2/4/6/etc and checking for changes in Color balance, and also tried looking very closely to see if I could spot differences in color balance/decoding with color bars up with RGBS=7 -- I did pretty much leave Color slider at 31, as that's where I want color decoding/balance to be accurate.

Using those test patterns, from what I could tell I did not see any appreciable, or even noticable change whatsoever in color balance/color decoding no matter where YOF/CBOF/CROF(or CBOP/CROP for that matter) is set. So, on my set It seems, I don't think YOF/CBOF/CROF/CBOP/CROP has much of anything to do with the differences in color balance/color decoding I'm seeing between 480p and 720p/1080i. If it does have anything to do with it, the effect is extremely small, so little not to be noticable by my eyes at least, nor for example does it even change the speed/etc of "flashing" in the AVIA color test bars in any noticable fashion that I can detect.

For now, since I'd already checked it previously with B&W content/etc and ended up using the defaults, I returned YOF/CBOF/CROF to defaults after I was done playing around with it. I'll continue to look for improved settings for those in the future, however.

Q of BanditZ

01-23-06, 04:53 PM

^^ Ya'll able to keep with all this? ;)

KenTech

01-23-06, 05:17 PM

In the section from my post you quoted, I meant 2170P3 #16 MIDE, where the MID5 columns are "assigned" - I.e. Where Twin View MIDE "pro" picture mode is "pointing to" MID5 Column #53 (all zeros).My bad! I am so used to associating a *setting* number with the parameter name (MIDE) that I was confused by the *index* number 16, thinking it meant a setting value. Sorry, brain scramble!
. . . concerning P2103-1 #16 SSMD, and your comments about it in the last sentence of a paragraph(d) concerning 3dcomb settings in post #707I said this in error, associating it with 3D-COMB. You are right; SSMD *is* in 2103-1 -- and it should still be set to 3.Well, my plan had been to make this a short post, sorry it didn't turn out that way!I've thought about saying that several times in my own posts. And then thought "Nah." Service-menu issues are not generally "short" issues, and so why should they be covered in short posts, eh?

KenTech

01-23-06, 05:33 PM

Color test patterns from DVE(I think from DVE) which ken provided earlier in this thread, and which I used from Memory stick(1080i).Yikes, I just read this, and I have to warn you: The use of memory-stick *color-calibration* patterns is unpredictable! I found that, if I set up the decoding matrix (RYR-GYB) with a known-accurate MS pattern in service mode, the pattern was displayed differently once I left service mode. I can't characterize the change, as I have avoided that method ever since. (I confirmed this change by viewing with the color filters instead of isolating the color guns, which can be done only in service mode.)

Instead I made a high-resolution (max bitrate) DVD of the same computer-generated color-bar pattern I once used from the MS, and used that through my DVD player. It jives almost exactly with the patterns on DVE and AVIA, so I am confident that the TV is set up correctly at this point. And a similar pattern displayed from MS I think is correct -- until you put the TV into service mode, and then something changes. (Or the reverse -- it's been a while . . .)

Further, although it seems apparent that the MS image follows the same path in the published block diagram as HD digital video, I don't know just how much is common. For one thing, it is absolutely certain that there is a 1080 X 1440 frame buffer for those MS images, determined by empirical testing way earlier in this thread. And I can't say what the overall color-matrix is or what it should be, NTSC or HD. Maybe it is neither. It's not a big issue with me, since I don't take JPEG-image display very seriously at this time. I just play with it or use computer-generated grayscale patters for focusing and geometry, which are fine for this. (Again, see previous posts.)

KenTech

01-23-06, 06:04 PM

So, If I was a real "neat freak" or something, I could REALLY reorganize the MID5 table and free up a lot of Columns for my own use ..... I don't think I'll ever have a need for that, though(at least I HOPE not)Each of us who has this urge can skin this cat in different ways. Here was my reasoning:

I was tired of the confusing redundancy in the MID5 columns, and I have not seen any need for having "reserved sections" of the MID5 chart for specific input/scan rates. I do, however, think in terms of what is being sharpened and by how much. So I set up a hierarchical sharpening scheme for the top 14 columns or so, working backward from #63 (POP=63).

So minimum effect is col 63, all zeroes. Then: Very fine sharpening (of very fine detail) increases if MHYL is set to 3 (no clipping) and MHYE is stepped from 1 to 7. The settings for MHCL and MHCE just echo the same for color -- although I think this is nearly undetectable. That accounts for cols 63-56. Then, beginning with 55 I return to MHYE = 1 and set MHYO to 1, lowering the video frequencies affected by the sharpening and making it coarser. Then from 55-50 MHYE steps from 1 to 6, keeping MYHO at 1. Thus from 63 to 50 there is a steady increase in sharpening and coarseness, and that makes experimentation easy: Put up a test image, and then, starting with POP=63 (or MIDE=63 in 2170P-3), decrement the value from 63 downwards until the desired effect is achieved, and lock in that value for MIDE. So, for example, col 60 (MHYE=3) is used for several different signal conditions and modes. And I know what I am getting every time I choose col 60 -- conservative, very fine sharpening. Those top 14 columns are not going to change!

Below 50 I have a couple of custom columns (49, 48) for experimentation with slight vertical sharpening: I duplicated col 60 but added MYVL=3 plus either MYVE=0 or MYVE=1. I use them for the HDMI 480p from my Panasonic S97 DVD player, to a small advantage. But this is on probation; easy to switch back to col 60 if I want zero vertical sharpening.

If I come up with other uses, I can add columns from 47 downwards. Since the basic sharpening settings (cols 63 > 50) follow a *system,* it's easy to experiment without getting confused or waylaid by duplicate columns. Why from the top down? I just super-cautiously wanted to leave in-place as many of the original settings as I could, although I now see that this has virtually no value in the future. I have the charts, after all, and I dispise most of the settings for Vivid, Standard, etc. That's all. I could have started from col 0 and gone *up* from there. The chart in post #707 reflects the above system.

jon_rhees

01-23-06, 07:24 PM

Does anyone know the Button sequence to enter the *factory* mode (not *service* mode) on the SXRD sets (the RPTV's, not the projectors). I am trying to get access to the 3D Gamma matrix on these sets (which should be similar to the Ruby and Q004).

Thanks,

Jon

KrammuelSTEog

01-23-06, 07:25 PM

I would first like to say that this forum for TV setup is the BEST.
Also, I have done this on the 34XBR800 and knowing that settings are different, I have had great success in adjusting mine pretty close to my satisfaction.
But, I do have a problem now.........I had moved recently and the TV had reset somewhat to a earlier state??? Is this something that certain settings revert back to factory or was it a screwup with the save (I know I had saved my settings, because it was fine for about 2 month before it was moved and power had gone out in that time and didn't affect it).
Any quick comments would be appreciated. TY

Nitewatchman

01-23-06, 07:57 PM

Yikes, I just read this, and I have to warn you: The use of memory-stick *color-calibration* patterns is unpredictable! I found that, if I set up the decoding matrix (RYR-GYB) with a known-accurate MS pattern in service mode, the pattern was displayed differently once I left service mode. ........<snip>

.......... I can't characterize the change, as I have avoided that method ever since. (I confirmed this change by viewing with the color filters instead of isolating the color guns, which can be done only in service mode.)
.)

Yes recall reading about that, from earlier in thread. However, along with internal QM test patterns/color bars it's all I currently have available to use with 1080i/720p unless I catch a station sending 1080i or 720p puting up color bars.

Also can't grasp how "errors" due to memory stick/DVE filters/in and out of service mode/different signal paths/etc/etc. would explain why just changing ONE value (MTRX=0 for 480p, MTRX=1 for 1080i or 720p) makes for pretty much perfect color balance per the test patterns from several different sources at either 480p or 720p/1080i, in EVERY case, while at the same time, MTRX=0 makes things quite a bit off with those different test patterns from different sources in the exact same way for 720p/1080i, same thing for MTRX=1 and 480p.

Changing the MTRX values from 0~1 is going to be exactly the same as errors introduced by DVE filters Outside of SM that is going to exactly match(or more or less from what I could tell) the "errors" introduced by using the MS inside SM? or, something with the signal path involving the QM 720p/1080i color bars? I didn't realize that would be anywhere near likely, if even possible, would like to learn more, however.

Note that in my case, again, for my ability to compare them, I want to stress that I got the same results from the Color bars from memory stick as I did from the internal QM 1080i/720p color bars. I just looked at it in SM with the DVE fitlers VERY quickly(I'll check it again, don't know If I'll get a chance to do that tonight), but seemed to have gotten, For whatever reason pretty much the same results with the DVE filters in and out of service mode - Now, I wasn't looking for any really "slight" differences here however.

I guess I must not be explaining myself very well about this. The differences I'm seeing in Color decoding 480p+720p/1080i with the same MTRX and RYR~GYB settings are perhaps not "huge" difference, but it's not "slight" issues either. Just look at the different values I'm needing for RYR-GYB for instance with the same MTRX value), and the differences are evident with the Internal QM test patterns, not just the MS, which is something I used additionally.

For instance, when balancing with Green gun, with the QM test bars at 1080i/72op or 480p ( and again I think this is important to note as well that avia/DVE color tests via DVD+480p match perfectly what I'm getting with the QM 480p Test bars on this, and the same thing is true with 480i QM color bars+AVIA/DVE via 480i DVD) It's VERY different "shades" of Green for the color bars that "show up" with the green only gun, that match almost PERFECTLY by changing the MTRX value to "0" for 480p test patterns, and "1" value for 720p/1080i, using 13-15-5-3 for RYR~GYB. Same thing if you look at RGBS #6 and the Y-G-R with color bars up from any source .......

The "difference" is also noticeable with program material as well, especially as I noted earlier when it comes to "natrual looking" greens, as well as via comparisons of program material with another set I know(pretty sure anyway - see previous info on this ) Is set up for proper color decoding.

First I thought maybe my DVD player is doing something strange, but again, what is happening with the test patterns match with 480p "ATSC" QM patterns+Avia/DVE Color patterns from DVD outputting 480p, and note that on Color decoder adjustments using AVIA/DVE test patterns on another set but with the SAME DVD player, what is required for it's various color decoder settings match perfectly for 480i and 480p - exactly the same.

On that particular set, it's setting similar to "MTRX"(It's called DABL specifically) can be set completely differently per scan rate(different settings also available for Svideo/composite indpedant from internal tuner/etc), just as all the other color decoder settings can be, including different values for the RYR~GYR "rough" equivlents. Again on that set/receiver, calibrated it for 1080i/720p with color bars from a broadcast ATSC source(checked it with another), and 1080i/720p/480p ATSC requires DABL=1EH, ANY NTSC source requires DABL=1CH, including 480P from DVD - which actually makes perfect sense where ATSC vs. NTSC is concerned, and it actually came from the factory with 1EH setup for 1080i/720p+480p via component(although that didn't work for me for 480p via DVD), and "1CH" for everything else.

It's problem is, "somewhat" similar<but different> in that it needs "DABL=1EH" for ATSC output form my OTA ATSC receiver, for 480p, but it won't let you set a different value for 480p from the DVD player, which requires the DABL=1CH.

[ Following section updated with corrections and add'l clarifications ]

What's seemingly different about that as compared to this issue I'm getting with the sony, is I'm needing different results from ATSC/NTSC 480p sources with the toshiba --- or, instead of ATSC vs. NTSC perhaps it might must be more because of different output of the STB Vs the DVD player? although the factory defaults of DABL=1EH for 480p/720p/1080i via component seem to suggest it might be more likely a NTSC vs ATSC colorspace thing .....

However, with the sony XBR960 - the 480p from that SAME DVD player+DVE/AVIA color tests perfectly "match" the pattern labeled "ATSC" 480p from the QM test patterns, BUT the same value "MTRX"(very simiar to "DABL" on the Toshiba) won't work "right" for 480p AND 720p/1080i. MTRX=0 works perfect for 480p, RYR~GYB 13-15-5-3, CBGN~YGN 6-5-5 -- either via DVD+AVIA or DVE tests OR the internal QM test pattern, MTRX=1 "messes that up" but seems to work perfect for 720p/1080i, Same RYR~GYB, Same CBGN~YGN -- Just looking for the "obvious" might indicate/ Might explain why the Servicecodes file says it uses MTRX=1 for specifically -- 1080i/720p via ATSC or V5/V6(no mention of 480p), and why there is also a listed column for "others" and "HDMI 1080i" ... and yet, I can't set different values for 480p vs. 720p/1080i for my set ....

Anyway -- I'll hopefully be able to catch some color bars up someday from a station when I can check the XBR so I can confirm more about what is going on with this to a greater degree --

The chart in post #707 reflects the above system.

And I very much like+appreciate the chart -- printed it out a week or so ago after the update(well and before the update as well) ---- I was able to easily understand from it what you were talking about the incremetal "sharpening" steps with the different Columns .... - and a great idea -- I'm using a few of the "most suggested" columns currently -- For instance, here's one of them from my "XBR960 service code changes.doc file :

Column #57 (Ken Tech's 1/15/06 suggestion for 720p and 480i/480p Component sources ) :

#0 POP - 57(0) - not saved temp setting - set to whatever column of table you are working in)
#1 MHLY - 0 (0)
#2 MHLC - 0 (0)
#3 MVLY - 0 (0)
#4 MVLC - 0 (0)
#5 MHYR - 0 (0)
#6 MHYL - 3 (0) - *
#7 MHYE - 3 (0) - *
#8 MHYO - 0 (0)
#9 MHCR - 0 (0)
#10 MHCL - 3 (0) - *
#11 MHCE - 3 (0) - *
#12 MHCO - 0 (0)
#13 MVYR - 0 (0)
#14 MVYL - 0 (0) - *
#15 MVYE - 0 (0)
#16 MVCR - 0 (0)
#17 MVCL - 0 (0) - *
#18 MVCE - 0 (0)

Currently using the (more or less) middle of the road values for columns per suggestions from another of your recent posts -- #59 for SD, #58 for 1080i(differs from above that MC, #57 for 720p+480p MHYE/MHCE=2 instead of 3) --- and when I get a chance I'll probably clear out everything from #54~63 and do that, or something simliar/along those lines .. but right now, I have some columns set up #60~63 I was using+experimenting with previously I'm not quite sure I'm completely done with yet ....

ADU

01-23-06, 08:46 PM

Looking forward to hearing your thoughts, and appreciate your posts on this so far. I'd read the discussions earlier in this thread related to CXA 21x1 settings and differences between ATSC/NTSC signals/etc, but it's been quite some time ago and I should review those again in more detail.Don't put an excessive amount of stock in my earlier remarks. It sounds like those controls might work a tad differently than I originally speculated.

Still readin... and still trying to catch up...

JeffD2.

01-23-06, 09:51 PM

THANKS ADU! I have a stack of papers ½" thick of the pdfs and "how to" sheets. I noticed you were a participant in that thread. Again, thanks for responding. I'll let you know how it works out. :D

KenTech

01-23-06, 10:39 PM

. . . while at the same time, MTRX=0 makes things quite a bit off with those different test patterns from different sources in the exact same way for 720p/1080i, same thing for MTRX=1 and 480p. <big snip>Just a note to say that I haven't yet wrapped my mind around the relationship between the "color matrix" settings of RYR~GYB and the definitions of the HDTV vs NTSC color standards as switched (I believe) by the MTRX value. I think it works like this: HDTV is based on entirely different coordinates for the RGB primaries in a global colorspace, and, for example, a "pure red" on HDTV is not quite the same as the same numerically-defined color displayed according to the NTSC standard. It's a slightly different red! But we have only one CRT in our TVs, and so to accurately display both, there are some remappings of color from one color definition to the other. This choice is made in MTRX, as I seem to have read before, and everything is Kosher *if* the HD video is supplied according to the real HDTV standard (as broadcast is). (I wonder what color standard is applied to digital SD broadcasts! Stations in the Portland, OR, area are starting to broadcast multiple subchannels, some with SD material.)

I don't believe that one standard can be converted to the other by fiddling with the RYR-GYB settings. It's a *different* matrix, where the actual primary colors and their max intensities (the gamut) are different. The RYR-GYB "matrix" adjusts the cross-color rendering for accuracy: GYB intensifies green and steals from blue; GYR intensifies green, stealing from red; RYR simply intensifies red, stealing from nowhere; and RYB increases red at the expense of blue. All of these adjust a "balance" in the derivation of intermediate colors as the eye sees them, given specific drive voltages. That's why you are equalizing the intensities of those color bars against each other and absolute gray *as seen by you.*

I ask myself frequently, Why is the color-rendition of this particular movie/program different between the SD broadcast and the HD simulcast of same? Both usually appear to be satisfactory, but frequently the shadow-area colors are handled better by HD and the reds are a little cooler. Sometimes the grays are different even though they are identical when Color is reduced to Minimum. Is the difference in the TV? Is there a "matrix mismatch"? Is the station at fault? I don't agonize over this, really, but I feel like there's way more going on that I (we) have control over!

I have not tinkered experimentally with the MTRX settings because it looks like it's working okay for my set. I haven't yet digested your overall experiments and their conclusions or your implied questions. That may still take a while, forgive me!

Jediphish

01-23-06, 10:41 PM

^^ Ya'll able to keep with all this? ;)

This is why I asked for a Wikipedia article on the subject, or a dedicated webpage. I'm gone for a few hours and there is so much new information that I cannot possibly keep up.

[sighs]

KenTech

01-23-06, 10:51 PM

I'm gone for a few hours and there is so much new information that I cannot possibly keep up.It's called binge-posting, a special quality of AROC posters that should be cherished! (AROC = Anal-Retentive Obsessive-Compulsive.)

Nightwatchman -- Since you *are* tinkering with image-rendition, I hope you are restarting your TV after 15-30 min warmup, as suggested in post #717. It makes a serious difference in some DA-4-based sets, really! I can really see the difference in, for example, the Matrix DVD via component-480i. I don't know where else it might be important, as in the XBR800 sets.

Nitewatchman

01-23-06, 11:25 PM

[QUOTE=KenTech]Just a note to say that I haven't yet wrapped my mind around the relationship between the "color matrix" settings of RYR~GYB and the definitions of the HDTV vs NTSC color standards as switched (I believe) by the MTRX value .....

.... I have not tinkered experimentally with the MTRX settings because it looks like it's working okay for my set. I haven't yet digested your overall experiments and their conclusions or your implied questions. That may still take a while, forgive me!

Forgive me for going out of order on quotes from your posts -- If I'm understanding you correctly -- Are you saying on your set, MTRX is set for "1" for 720p/1080i via ATSC receiver, and "changes" to "0" for 480p via say DVD?

If so, wonder why mine wasn't like that from the factory(it was "0" default for EVERYTHING I could "switch to") -- such as if say it's getting the values from "somewhere else" --- And also, - I wonder why I can't/couldn't set it for "0" for 480p, and "1" for 1080i/etc ? Tried it more than once, it changed "everywhere" to the new value when I changed the value from the default of "0" to "1", for instance ...

And No problem - I do want to emphacize that I have not really reached any "conclusions" of any sort about this really --- Just reporting on what I've found+observed via any means available to me via "poking around a bit", as more than anything else it just seems to me something isn't "right" on my set with that MTRX value of "0" that was the set default being used for "everything". 720p/1080 HD and 480p DVD.

Agree 100% with your entire post, looks like we're thinking pretty much the same thing. Just from what I've seen seems like to me I need to be able to use different MTRX values for 480p DVD/NTSC than is the case with 720p/1080i, but it isn't letting me "do that" with 2171CXA "MTRX".

What I'm really wondering about as well(and is a big part of why I've been posting about this) is if there is another "setting" or lookup table somewhere else involving this I can't find/don't know about which may be set incorrectly that might explain why you, and/or others aren't experiencing at least some sort of indication of issues with this but I am at least noticing that something *may* be off about it ...

(I wonder what color standard is applied to digital SD broadcasts! Stations in the Portland, OR, area are starting to broadcast multiple subchannels, some with SD material.)

Have wondered about that as well - According to one of their engineers One station(a statewide PBS Network in KY - KET) here which varies between 6SD services and 1HD+2SD services(subchannels) depending upon time of day/program schedule/etc -- distributes it's ATSC complaint transport stream from centrally located MPEG encoders(just one for each service/subchannel) to 16 stations statewide via Microwave links with an OTA backup ...

At the transmitter sites --- for the digital station unique PSIP VCT(virtual channel table) info for that station is added, but otherwise, more or less the datastream is just "passed on" to the RF transmission plant without a decoding/reencoding process --- For the Analog station however at each site, there is a decoder at the transmitter site which decodes the stream and outputs NTSC which is sent to the analog transmitter, and one of those services(called KET1) is used for the analog station. So, "theoretically" the "color"(or anything else, basically) one sees via the analog station should be "pretty much" the same(including fleshtones/etc) as from the KET1 subchannel from the digital station -- And, it is on the XBR960, allways has been .. After calibration of color decoder/etc. it is on my Toshiba set as well ... Actually, one of the stations the color bars I calibrated the Toshiba's decoder with which I mentioned earlier were from one of these SD subchannels, but upconverted from 480i and output at 1080i from my receiver.

I don't believe that one standard can be converted to the other by fiddling with the RYR-GYB settings.

Absolutely, I agree 100%, including your comments on this I didn't quote. I was just "trying it" for say 1080i when using "MTRX=0" to fudge it and see what happened - MTRX=0 was after all the set default on my set for instance both 480p(NTSC) and 1080i ..... [update - at least that's what it was "showing" while in I'm in SM] ... Again, this is why I don't understand how I could have had a "MTRX" value of "0" for default on my set that is supposed to work "right" with 720p+1080i via ATSC and 480p via DVD, and I can't make it "1" for 720p+1080i, and "0" for 480p.

Nitewatchman

01-23-06, 11:56 PM

Nightwatchman I hope you are restarting your TV after 15-30 min warmup, as suggested in post #717. .

Yes, doing the restart thing. It's been quite a while since I did too much tinkering with it or making many comparisons, and I've been happy with the settings you've found that I haven't done much to mess around with them.

So far with the new recommended settings, I pretty much just plugged them in, did some quick "comparisons" to the old settings+pro defaults and that's about it ...

It might be a while before I get around to more closely looking at things with the new settings, but I will look at them a little more "closely" ...

Nitewatchman

01-24-06, 01:57 AM

Folks -- don't spend time trying to deciper my previous posts on this ....

In short - It appears the color matrix is "automatically" switching to the right matrix for 480p NTSC, and also to the right matrix for ATSC 720p/1080i depending upon if source is 480p or 720p/1080i, and it is working and switching automatically On my set(at least right now) ... But, it's only doing that auto "switch" when you are NOT in SM.

It is not making that "switch" to using the "right" color matrix for HD vs NTSC when I am in SM(at least that's how it's working on my set). Also noticed that Even though I set MTRX=1 (I even used write/enter) last time I was in SM last night --- Noticed it defaulted back to "0" when I went into SM for the first time since then.

It looks like when I am in SM, unless I change it, it's "staying" with the Matrix for NTSC no matter what, even if I switch to a 1080i source. Looks like you simply have to "do it" yourself and set MTRX=1 if you want to check from within SM Color balance with the QM internal 1080i color bars look like or, a test pattern with memory stick(with the necessary "cautions" for using those as ken mentioned of course), Or, while in SM, set it/make sure its set at MTRX=0 if you want to check, say 480p NTSC from DVD.

So, seems to me what changing the MTRX setting is "useful for" is so you can manually set it for looking at color test patterns for 480p from say NTSC DVD (MTRX=0) or 1080i/720p(MTRX=1) while in SM -- at least that's how it's working on my set, since it is only making the switch to the right color matrix for ATSC HD/NTSC 480p "automatically" Outside of SM.

On a whim, tonight turned on the set(it had been off for a couple of hours), powering up in SM -- first thing I noticed was that MTRX was showing "0" again, everywhere, even though I set it to "1" last night.

Here's part of what I did to check this out :

For all of the below, RYR~GYB 13-15-5-3, CBGN~YGN= 6-5-5, Color slider 31, Hue=0.

From within SM -- Left MTRX=0 and checked AVIA color tests output from DVD/AVIA at 480p- Pretty much perfect with using RGBS 1/2/4/6, with DVE filter, I mostly see error in the DVE filter with "green" - Saturation "flahsing box" flashes, but Hue doesn't.

Then, leaving MTRX=0, looked at 1080i QM internal test pattern, and Memory stick+looked at "DVE-Color-Chart-X4.jpg, posted earlier in thread.(I know it's not the best thing to use, but I think it worked for my purposes). As I've experienced before with this, it was very evident that Color balance was "way off" for both with MTRX=0 but, setting MTRX=1 from within SM indicated pretty much perfect Color balance from those 1080i test patterns. Note that this time, I just left it at MTRX=0, however.

Then, I powered down the set. Turned it back on, not going into SM. Went to Memory stick, checked out DVE-Color-Chart-X4.jpg with DVE filters. Pretty much close to "perfect" - or at least very close - VERY different than with "MTRX=0" in SM. Went to the DVD at 480p, pretty much perfect, except again I see the same thing with "green" DVE filter I did when in SM.

Boy, do I feel stupid. Oh, well -- OTOH, glad I stuck with it until I figured out what was going on. Part of the reason why it took me so long to figure this out I think is because I wasn't trusting that pattern from memory stick "enough", and I was "thinking" my changing "MTRX" values was "sticking" in a static way after I left Service mode. Don't know how it turned out that I hadn't noticed it wasn't, until now.

One thing good that did come out of all this for me is, I was able to improve color accuracy/decoding for 480p+720p/1080i by changing CBGN~YGN from 4-5-5 default to 6-6-5.

I guess the only question I have about this now Is about 480p ATSC vs. NTSC - Just as with 480p from DVD, "MTRX=0" appears to be the correct setting for the internal QM color bars for 480p ATSC, as opposed to "MTRX=1" being correct for 720p/1080i internal QM color bars. Wonder if this would perhaps be a good indication that 480p ATSC uses/needs NTSC color matrix? OTOH, that has not been my "experience" with the color decoder settings on my Toshiba set+output from Zenith HDV420 receiver and its differences for example between 480p from the ATSC receiver, and 480p from a DVD Player ...

When I bring up the 480p QM internal test pattern, it says "480p ATSC" on screen, and it also does seem to be the case it would(or should) be the "same"(more or less) as 480p from the ATSC tuner, as all the P21703 settings/etc. remain at factory default when I'm looking at that test pattern, and in fact I haven't changed those for 480p ATSC - that isn't the case when I bring up 1080i test pattern for instance -- my changed P2170-3/etc. settings "show up".

---------------------------------------------------------------------------------

[Update]

Thought I should add --- thing is though --- At one point during my poking around on this, I had checked 480p DVD with AVIA outside of SM with DVE filters after just being in SM and changing MTRX from "0" to "1". And, With RYR~GYB 13-15-5-3, CBGN~YBGN 6-6-5, Hue=0 Color=31-- If I recall correctly(which I may not be), I Thought it was more "messed up" than for instance what I noticed with the error I noticed for DVE Green filter tonight, when I "knew" that it was using the correct color matrix for 480p DVD NTSC(equivelent to MTRX=0 if you are working in SM) ...

That, BTW led me at the time to dismiss the possibility it was "switching automatically" outside of service mode .... Probably not -- but, wonder if it is perhaps a possibility that under some circumstances/etc(such as if in some cases you had very recently set MTRX=1 in SM and immediately powered down set+powered back up), the automatic switching to the right color matrix for 480p/720p/1080i outside of SM doesn't allways occur, 100% of the time?
[end update]

ADU

01-24-06, 06:08 AM

Stranger and stranger... Based on some other posts I've been reading, it looks like this parameter could work differently depending on whether you have the CXA2151 or CXA2171 chip. [EDIT: ...or not. See later posts below which seem to suggest that both chips behave similarly with respect to MTRX.]

FWIW, whatever value I last leave CXA2151/MTRX at on my 34XBR800, that's what comes up next time I power on. This works regardless of whether the new value is written to memory or not. The CXA2171 may work differently though. [EDIT: Again see posts below, as this isn't really the whole story. And the value will "revert" to it's proper setting if you don't directly re-enter service mode.]

It's possible your TV could be storing two different values... one for 480i/480p, and one for 720p/1080i, for example. Maybe you can change/store different values for 720p/1080i, but 480i/480p always defaults back to 0. You may also need to write the new value for 720p/1080i to memory (while displaying a bona-fide 1080i signal), unlike on my TV. [Edit: again see below, as this guess was also off the mark.]

On the otherhand, if it seems to be working correctly, maybe you should just leave well enough alone.

Nitewatchman

01-24-06, 12:47 PM

ADU,

I'm just glad I'm not the only one that has seemed to notice some "odd" behavior with this, at least to some extent.

It's possible your TV could be storing two different values... one for 480i/480p, and one for 720p/1080i, for example.

Seems to be that's what is happening on my set -- but not when you're in SM. I still have to change it to "1" to get proper matrix for 720p/1080i when I'm in SM, and the value that "shows" up doesn't change when you are in SM when you switch between say, a 480p and 1080i source.

Or - and just some wild guessing about the "possibilities" here, but perhaps with my set at least it may be more something along the lines that the "MTRX" value from the SM is not even generally used by CXA2171 chip when you are outside of SM, and it's done via hardware/etc. instead, and when you enter SM, something "switches" that feature off on the chip, which has it then "look" the MTRX value and whatever you change it to(when you're in SM). And maybe, just maybe that doesn't allways "stick" perfectly when you enter/exit service mode, and sometimes in certian cases it may still be looking to the "MTRX" value in SM when you're outside of SM. Perhaps, this, or something along these lines might also explain why we don't seem to (generally) have to use Mute-enter to get it to "stick" ....

BTW, I'm pretty sure I checked it from 480i ATSC source(both broadcast+internal 480i ATSC test pattern), and don't think 480i, even 480i ATSC is effected by the CXA2171 SM settings, including "MTRX". Only 480p,720p and 1080i seem effected.

Update: Just confirmed with 480i broadcast source+internal QM 480i color bars that 480i isn't effected by "MTRX", at least on my set by changing the value while in SM. :end update

You may also need to write the new value for 720p/1080i to memory (while displaying a bona-fide 1080i signal), unlike on my TV.

Could be, but I more suspect it was already "set up" that way on my set, and was already switching between "1" for 720p/1080i and "0" for 480p outside of SM -- I just "didn't know it" for sure as I hadn't checked any patterns from MS(the only 720p/1080i source I have other than broadcast stations) outside of SM Until very recently and was looking at test patterns from within SM, and that's where I still have to set MTRX at either 0 or 1 "manually".

FWIW, whatever value I last leave CXA2151/MTRX at on my 34XBR800, that's what comes up next time I power on. This works regardless of whether the new value is written to memory or not. The CXA2171 may work differently though.

What's funny is, maybe I'm just losing my mind, but -- I could have swore that I checked that - more than once in fact, and noticed it working the same way as it is with your set - But that is NOT how it happened the last time. Instead, had set it at "1", but it had defaulted to showing "0" next time I powered up in SM - But, the set had been off since then, for several hours at a time. That wasn't the case on earlier occasions when it "looked like" it was sticking - on those occasions, the set wasn't off for more than a few seconds.

On the otherhand, if it seems to be working correctly, maybe you should just leave well enough alone.

I agree. Thing is though, in my case for 720p/1080i - the best way I have to check color balance is to go into SM and look at the Internal Test patterns/Color bars located in the QM PATN section of SM for 720p/1080. And, to do that properly(and to be able to use RGBS) from within SM, I have to set MTRX=1 manually to get the proper color matrix for HD.

I may be remembering it wrong -- but seems like at one point I found it's possible to put up a test pattern from the internal QM test patterns - then hit "mute/enter", and it will "stick" when you go outside of SM. I might try this(but unfortunetly just with ther RGB filters that came with AVIA/DVE) with the 720p/1080i/etc. color bars and see what I get, as I've come to pretty much "trust" those internal test patterns.

Update: Just confirmed that I can get the QM patterns to come up outside of SM by the above method. I just tried this with the Color bars for 1080i, 720p, 480p, 480i and checking Color and to see if it switched to the right MTRX with DVE filters outside of SM - Every time, it switched to right matrix, even at one point while I was in SM looking at the 480p color bars, I switched MTRX to "1" in SM and left it there for set power off. In this case the "Correct" MTRX was chosen outside of SM, and afterwards when I reentered SM, MTRX had returned to "0".

When I 1080i via this method - didn't change MTRX from "0" in SM, even though I had 1080i Color bars from internal test patterns up, and noted it was using the "wrong" color matrix while in SM, as usual given MTRX value of "0" --- Then, powered off set, turned back on outside of SM checked it, it chose the right color matrix -- turned off the set, went back into SM, and, as expected MTRX still said "0"(and of course with the "wrong" color matrix)..

HOWEVER -- And this seems strange - I did the exact same process with 720p Color bars as I did/explained above for 1080i - HOWEVER - When I powered the set back up for the "last" time in SM -- it was showing MTRX=1 with the 720p color bars!!!! Even though, it was showing "0"(and with the wrong color matrix) for the first part of the test, before I went outside of SM and checked it.

So, I do the process again first with 480p color bars ... This time, when I power back up into SM for the final part of the "test", it shows "MTRX=0" as it should ...

Thinking I should check 1080i again, I do so just to see what happens this time and if it was different than the first time --- Low and behold it WAS different this time! This time, it showed up with "MTRX=1" when I reentered SM for the final part of the test, whearas the first time, I did the same test, it showed up as "0", even though "1" (or the correct color matrix) was used when I was outside of SM .... VERY strange ...

BTW, Another thing I tried the first time I entered SM this afternoon was to make sure I had the internal tuner on a 1080i station before entering SM. When I powered up in SM, as expected, it showed MTRX=0". But, tried that again the last time I entered SM, and THIS time it said MTRX=1 ! How very strange.

Of course, just to make sure I wasn't losing my mind, I switched between 480i/480p/720p/1080i sources(including color bars) while within service mode, and as usual, it's still not switching "MTRX" while you are IN service mode.

end update

KenTech

01-24-06, 02:37 PM

ADU and Nightwatchman: Regarding the CXA2171 chip in the DA-4 chassis, my full-color parameter chart printout indicates the following, as I read it. (I hope this isn't too redundant with what you've already concluded.)

(1) The MTRX parameter is explicitly listed as NOT memorized after you make changes, i.e. you can't WRITE it. Its value is determined on-the-fly by the video mode plus digital signal flags.

(2) For 720 or 1080 from either broadcast (ATSC tuner) or component inputs (V5, V6), its value is 1 = HD standard.

(3) For "Others" it is 0 = NTSC standard. Looks like this includes SD digital, but I suppose one might confirm this by watching it change as you change inputs. Unless it's "frozen" in service mode, which is what Nightwatchman concluded in post #833.

(4) For HDMI it depends on a flag in the signal, and it can be 1 or 0. Here is where your favorite DVD player via HDMI can flip the color matrix for HD, if it's designed to do so. If the flag ("uTiny/$6D_00, byte1/bit6") is 0, MTRX is 0 (NTSC); if it's 1, MTRX is 1 (HD).

It seems quite clear that, for these sets anyway, MTRX is not under your control except *temporarily* in service mode.

ADU

01-24-06, 03:00 PM

Nitewatchman,

Be careful with all the powering on and off in such a short space of time. I think there could be a risk of overheating your degaussing coil.

ADU

01-24-06, 03:06 PM

Ken,

Your post makes alot of sense. I too am now noticing the same "reverting" behavior on my 34XBR800 (with the older CXA2151 chip) that Nitewatchman is seeing on his 34XBR960 (with the newer CXA2171 chip).

If I switch CXA2151/MTRX to a different setting in the service mode on my 34XBR800, then power back on again into service mode, the change remains intact (because it's like I never left the service mode to begin with).

But if I power back on without going into the service mode, MTRX reverts to its original value.

Easiest way for me to see this is by setting MTRX to 3, which has a distinct green cast.

Nitewatchman

01-24-06, 03:42 PM

ADU and Nightwatchman: Regarding the CXA2171 chip in the DA-4 chassis, my full-color parameter chart printout indicates the following, as I read it. (I hope this isn't too redundant with what you've already concluded.)

Not redundant at all, I think your post is an excellent summary concerning how it works with CXA2171. And, that's the way I read the chart as well, just wasn't sure it really "worked" that way with my set, previously -- now(finally) I am sure.

At this point -- The odd thing about it I'm getting with my set is -- such as noted in the tests I did per update of my last post : Sometimes even though outside of SM it seems to be allways using the correct MTRX value for any given source - such as "1" for 1080i HD source from ATSC tuner and "0" for 480p via component V5 -- When you power up in Service mode while that source is still selected, MTRX value *may* or *May not* be using the correct value for that same source -- I've seen it occur both ways, even with the same source+checking it twice over a time span of about 15 minutes.

Couple of other thoughts or clarifications :

3) For "Others" it is 0 = NTSC standard. Looks like this includes SD digital, but I suppose one might confirm this by watching it change as you change inputs. Unless it's "frozen" in service mode, which is what Nightwatchman concluded in post #833.

I think SD digital (OTA ATSC, or even 480i via component) must be using NTSC standards[update] For color. While that seems strange since we know OTA SD 480i digital is using ATSC standard for the most part(such as 8VSB for transmission/channel coding, as well as PSIP/etc) --- It's also using MPEG2(which is also "part of" ATSC standard but can be used w/o rest of ATSC as well), and for instance -- what is SD digital/MPEG2 stream via my Dish network SD receiver is NTSC, in this case output via my receiver as analog NTSC via s-video or composite video outputs. [end update]

From what I can tell, looks like 480p (via DVD or perhaps even ATSC -depending upon what the source/including broadcast is using, perhaps) looks to be at least be "mosty" NTSC color as well. For instance, 480p DVD is apparently using and provides accurate color balance with "MTRX=0", and 480p internal QM color bars(which is showing up as "480p ATSC input in SM") is using/needs "MTRX=0" for accurate color balance.

What I concluded is that 480i ATSC sources -- broadcast or the internal 480i color bars(or anything else except perhaps 480p that "fits" in that "others column" for that matter - including 480i via component) is not effected by changing "MTRX" setting from within SM, or by changing CBGN/CYGN/YGN for that matter. Perhaps it could be affected outside of SM, perhaps depending upon flags/metadata sent via the MPEG2 stream/etc. by the broadcaster.

What I also concluded is :

a). From within SM, 480p/720p and 1080i are effected by changing "MTRX" value.

b). In SM, or outside SM, CBGN/CRGN/YGN effects 480p/720p and 1080i.

Also -- note that color accuracy with AVIA Color tests via 480p DVD outside of SM improved via my slight modification of CBGN/CRGN/YGN, which also didn't effect color balance with 480i Component video via DVD/AVIA color tests.

It seems quite clear that, for these sets anyway, MTRX is not under your control except *temporarily* in service mode.

Yes, absolutely. And, I would add(at least on my set) it also seems quite clear that within service mode, it's a necessary control to have access to as :

#1). It is apparent(at least on my set per my observations) that unlike is the case outside of SM -- when you are working in SM, it is not properly detirming the correct value and switching "on the fly" to that correct value depending upon source, or digital signal flags/etc - at least for the internal QM test patterns, Memory stick, OR 480p via DVD. You have to do that/make sure it's set properly for the source you are using(per the codes in the service code listing), Manually.

#2) WHY It's useful to be able to modify MTRX in SM for the "correct" value , per #1), above is -- You need to be able to modify MTRX value in SM while checking or modifying CBGN/CRGN/YGN for improved Color balance for 480p/720p/1080i sources in order to select the proper MTRX value which corresponds to the source you are using -- For instance, "0" for 480p, or "1" for 1080i HD color bars --- In order that you are using the correct color matrix (ATSC or NTSC/etc.) for the given source. From what I've seen, the "correct" matrix value for instance is "1" for 1080i/720p, "0" for 480p via DVD or any other 480p source of Color bars/tests.

Nitewatchman

01-24-06, 03:51 PM

Be careful with all the powering on and off in such a short space of time. I think there could be a risk of overheating your degaussing coil.

Yes, thanks for the reminder. Probably should have thought more about that instead of all the actual "work" I have to do today -- before doing some of that so quickly ... And if I'd had the time should have waited longer than that, or even better waited until I had the time -- perhaps 15 minutes between power cycles ...

Generally it was probably at least 5 minutes or so between most power cycles to check the SM values, when in SM change RGBS/etc, but I did do it a few times much quicker than that, especially after going out of SM, powering up/checking pattern with DVE filter than powering down+powering back up In SM ..

KenTech

01-24-06, 05:27 PM

Be careful with all the powering on and off in such a short space of time. I think there could be a risk of overheating your degaussing coil.No chance of this with Sony DA-4 chassis and likely many other Sony models. A built-in timer (or an old-fashioned PTC resistor in series) prevents degaussing for quite a while after it first happens. You can hear when it's active (BZ-z-z-t). Even my ancient 1978 RCA monitor was no problem. If you turned off the set and turned it on within 5 minutes, no degaussing.

Biochemlab

01-24-06, 05:30 PM

No chance of this with Sony DA-4 chassis and likely many other Sony models. A built-in timer (or an old-fashioned PTC resistor in series) prevents degaussing for quite a while after it happens. You can hear when it's active (BZ-z-z-t). Even my ancient 1978 RCA monitor was no problem. If you turned off the set and turned it on within 5 minutes, no degaussing.

I noticed that with mine... I have had to turn it off and on a couple times within 1 - 2 minutes, and it only degaussed itself once.

KenTech

01-24-06, 05:48 PM

#2) WHY It's useful to be able to modify MTRX in SM for the "correct" value , per #1), above is -- You need to be able to modify MTRX value in SM while checking or modifying CBGN/CRGN/YGN for improved Color balance for 480p/720p/1080i sourcesWell put! Are you suggesting that a good way to set up CBGN/CRGN/YGN balance is to put up the internal color bars for (e.g.) HD, make sure MTRX=1, and then touch up those settings with the one-color guns, leaving RYR-GYB alone? If this doesn't make sense according to your recent experience (thank you!), what do you suggest as a procedure? I haven't yet tried to fiddle with these values. (Rain is predicted here for Saturday. [Rubs hands together])

JeffD2.

01-24-06, 07:35 PM

JeffD2.

2170D-3/JUMP should be the parameter you're lookin for to expand HD to fill the screen on the 36HS510. If you need to tweak vertical sizing and/or position on the expanded HD, some folks use the MID controls. I found the various sensitivities and relationships in those controls rather complicated and easy to screw up though, so I've been using 2170D-1/ASPT for vertical size and 2170D-1/SCRL for vertical position instead.................

PERFECT! I can't thank you enough for this tip. I can now pass pure 1080i from the STB directly to the TV without the cable box (SA3250HD) do any scaling (it would down convert to 480p when IT did the scaling). I just recently got an upconverting DVD player which I can NOW take full advantage of. Also I didn't have to fudge with any of vertical size or position settings. It really couldn't have gone any smoother.

A cold beer and a toast to ADU! THANK YOU!

PS- I and a few others around the internet have a problem with the memory stick icon and "access" flashing in the bottom right corner of the screen. Not all the time, and not in any given time pattern. Anyone have a fix or disable for that?

Nitewatchman

01-24-06, 08:57 PM

Well put! Are you suggesting that a good way to set up CBGN/CRGN/YGN balance is to put up the internal color bars for (e.g.) HD, make sure MTRX=1, and then touch up those settings with the one-color guns, leaving RYR-GYB alone?

I'm guessing this is probably going go to go more quickly than you might be thinking Ken -- Careful though as that's a <jinx alert> if I ever heard one ;)

I'd think that should work -- and makes sense -- I also do think it's a good idea to check what you have with your CBGN~YGN settings currently, as well as the new settings you come up with at 480p via DVD and/or internal color bars/MTRX=0, and both 720p+1080i via internal color bars/MTRX=1.

From what I'm getting, I think with the "best" CBGN~YGN values you should see the same thing for all those scan rates which CBGN~YGN effects(480p/720p/1080i) - and, with best settings for CBGN~YGN, and the same RYR~GYB settings for everything -- what you are getting color balance wise for 480p/720p/1080i should also "match" what you are getting from those scan rates/inputs that aren't effected by CBGN~YGN settings -- Of course the goal here being "perfect" or as near perfect Color balance for all sources/inputs/scan rates/etc.

And, perhaps the best procedure to use might not necessarily be the same for everyone for various reasons(I'll get into these later). Perhaps for some it might be better to use 1080i test pattern and MTRX=1 to tweak CBGN~YGN, or might be better to use 480p DVD and MTRX=0.

What I actually did, in this order was :

#1). To "tweak" CBGN~YGN -- Leave the RYR~GYB(13-15-5-3) alone and use AVIA(DVE would work as well of course) color tests from 480p DVD via component input and RGBS=1/2/4/6 with MTRX=0. My factory defaults for CBGN/CRGN/YGN were : 4-5-5. My new values are 6-6-5.

Note that unlike with RYR~GYB and using AVIA or DVE with one-color guns -- changing ANY value of CBGN/CRGN/YGN effects what you get to some degree(in some cases to more of a degree than others) with the tests with other two "one color guns" -- So, you'll have probably have to go back through looking at RGBS=1/2/4 several times while "touching up" all 3 settings ...

#2). Then, I went to the 1080i internal color bars, Changed MTRX=1 and using RGBS in the same fashion tried to see if I could Improve the settings for CBGN/CRGN/YGN that I got for 480p DVD. I couldn't Improve them - that may be more because I think I was able to do it more precisely with AVIA than I could with the internal color bars --- especially with red gun, but what I'm getting for 1080i(MTRX=1) very much seems to "match" what I'm getting with 480p(MTRX=0).

#3). Tried the 720P internal color bars in same way as #2, still wasn't able to improve anything -- Not that it needs improvement - From what I can tell, it's really right on the money.

Notes:

So, It looks like(from what I'm getting with this at least) Color balance using the one gun approach for #1 and #2, or #1 and #3 should produce "Matching results", for 480p vs 720p, or 720p vs 1080i or 720p vs. 1080i. In other words, given you have accurate color set up with say 480i DVD component with your chosen RYR~GYB settings --- with those same RYR~GYB settings, and finding the "best" CBGN~YGN settings as well -- I think the color balance should also be accurate for all 3 guns used 1 at a time, and look the same via color bars for all three scan rates 480p(with MTRX=0), or 720p(with MTRX=1), or 1080i(with MTRX=1).

Note that on my set -- I was getting pretty much perfect results with DVE or AVIA Color tests via DVD 480i component input with RYR~GYB 13-15-5-3. But, with 480p DVD via AVIA/DVE, and the factory CBGN~YGN settings, It was noticably a bit off from that(even outside of SM) Mainly was evident with tests with "blue" only or "red" only gun - It's now pretty much identical for 480i/480p given the new values I'm using for CBGN~YGN.

So, you might not even want to change CBGN/CRGN/YGN from the factory values If your 480p DVD via component is "perfect", as changing those for 1080i will effect 480p as well -- In other words - If your 480p is "perfect" now, 720p and 1080i should be as well -- But, You might must just want to check 720p/1080i with internal QM Color bars with RGBS 1/2/4/6 to make sure.

Besides CBGN~YGN values being used for 480p and not 480i concerning any differences there might be between 480i vs 480p DVD -- the other things I can think of that might be involved if they don't match color wise with the same RYR~GYB setting might be if DVD player is outputting slightly different color for 480p than it does for 480i, or perhaps something involving 480p specific HUE/TINT offsets in user controls or 2170P3 UCOF/UHOF. Also, note that I noticed on my set - even though the factory default was the same for 480i/480p/720p/1080i via component but noticed even SHUE/SCOL in 2170P4 has a 480p component "specific" value that doesn't apply to say 480i via component or 720p via internal tuner. Even though the service code listing seems to not necessarily indicate those P21704 values have such "scan rate" specific values.

In my case I think I was also able to detirmine for sure DVD player I'm using was outputting same color for 480i and 480p per adjustment with it and use of AVIA/DVE for adjusting color decoder in another set which happens to have scan rate specific color decoder settings. As, for example, on that set for both 480i and 480p from this DVD player the Color decoder settings needed to be exactly the same, and produced matching results with AVIA/DVE color tests for both 480i and 480p. And, where The sony is a bit different in that 480p also is "effected" by CBGN~YGN, whearas 480i from DVD at least, doesn't appear to be.

KenTech

01-24-06, 10:40 PM

#1). To "tweak" CBGN~YGN -- Leave the RYR~GYB(13-15-5-3) alone and use AVIA(DVE would work as well of course) color tests from 480p DVD via component input and RGBS=1/2/4/6 with MTRX=0. My factory defaults for CBGN/CRGN/YGN were : 4-5-5. My new values are 6-6-5.In my case, the output from my Panasonic S97 component jacks for 480p is dreadful! Much cruder than for the same at 480i and for HDMI at 480p. So it's out of the picture! But I'm curious what I get for the internal HD pattern and maybe HDMI/480p. I'll figure something out. The rest of the method (1080i, etc.) seems exactly right.

BTW, I believe these three parameters, CBGN~YGN, are gain ("GN") controls for the three components of the affected video, already known as Y, Cb, and Cr. Y is luminance alone, and Cb and Cr completely determine color hue and saturation along a blue-yellow and red-green axis, respectively. So their adjustment will not track with RGB adjustments elsewhere, as they are a totally different way of specifying an image. These TVs seem to process everything as YCrCb right up to the last stage before the CRT, where, of course, it must be converted for the three R, G, and B guns. Another name for component video is "YUV." Accordingly, the chip we're discussing, the CXA2171 or equivalent, is designated as "YUVSW," the YUV switch, I presume. It takes inputs from just about everywhere in the block diagram. From there, signals eventually find their way to the CXA2170 chip and thence to the CRT driver board.

Implication of above: If you are satisfied with image brightness (intensity of white), you should probably leave YGN alone. That overall parameter is best adjusted elsewhere, e.g. 2170P-4/SPIC (sub-Picture). You can further differentiate RF from CV/YC in 2103-1/SCON.

Some definitions hold that YPbPr is for analog signals and YCbCr indicates digital versions of same. The Sony DA-4 block diagram confuses YCrCb and YPrPb, showing the former as the designation at the component inputs but also after A/D conversion. But YPrPb appears at the output of the TASC tuner. Huh? Assume nothing!

Nowhere can I find any reference in the block diagram to the internal pattern generator, no chip, no module. So I don't know where it inserts its signal.

As one who has meddled with Lab color settings for images in Photoshop, a form of YPrPb, I can say that the adjustment of Pr and Pb is not nearly as intuitive as is RGB manipulation. You have been warned!

Nitewatchman

01-25-06, 01:19 AM

BTW, I believe these three parameters, CBGN~YGN, are gain ("GN") controls for the three components of the affected video, already known as Y, Cb, and Cr.

I think you'll believe that even more if you use them.

Y is luminance alone, and Cb and Cr completely determine color hue and saturation along a blue-yellow and red-green axis, respectively. So their adjustment will not track with RGB adjustments elsewhere, as they are a totally different way of specifying an image.

Absolutely, that's component video for you. What I noticed with my "differences" between 480i and 480p AVIA/DVE tests using RGB guns individually was mainly that with 480p with the "Blue" gun only, and "Red" only balance was off(but neither by a whole lot), With the "green" gun it was much closer to being accurate, but still was off a bit.

Just to see, I tried to change RYR~GYB RYR~RYB most specifically just to see how it would work for an "improvement", but that didn't really quite "work" to my satisfaction. And, adjusting Blue saturation(with Color slider or P21703 UCOF) didn't quite work as well as I wanted either - "Hue" for Blue was still pretty much right on with user control set to "0" anyway.

keep in mind that this same player via calibration of color decoder settings on another set calibrated perfectly "the same" for both 480i/p via component connection.

So, I thought what OTHER possible controls would let me perhaps improve things with 480p that might actually work and do something "different" than hue/color saturation sliders or 2170P3 UCOF/UHOF, etc. ? For me, Slight changes to CBGN/CRGN seemed to have done the trick ... Not only for 480p, but 720p/1080i as well. If it HAD of couse "messed up" 720p or 1080i from the internal tuner, I would have had to have tried something else -- instead, though, it *equally* improved things for 480p AND 720p/1080i. In fact, I noticed right off the more "natural" look to Color on 720p/1080i broadcast sources.

Now, I don't expect that will necessarily be the case for everyone, that's for sure.

That overall parameter is best adjusted elsewhere, e.g. 2170P-4/SPIC (sub-Picture). You can further differentiate RF from CV/YC in 2103-1/SCON.

Factory setting for "SPIC" on my set was probably the worst single "adjustment" from the factory of all the settings I've come across. First thing I adjusted.

So, imagine if the set defaults could be so off for "SPIC" how they could also be off, if perhaps only just a little bit for CBGN/CRGN as well ... maybe even YGN, although I did not change/need to change that one ...

Some definitions hold that YPbPr is for analog signals and YCbCr indicates digital versions of same. The Sony DA-4 block diagram confuses YCrCb and YPrPb, showing the former as the designation at the component inputs but also after A/D conversion. But YPrPb appears at the output of the TASC tuner. Huh? Assume nothing!

Nowhere can I find any reference in the block diagram to the internal pattern generator, no chip, no module. So I don't know where it inserts its signal.

Yes, I was reading and following along with the discussion and looking at the block diagram/trying to decipher it on this earlier in thread at the time you guys were talking about it. Not sure there really is much, if any any real difference between YpbPr/YcbCr - at least where pbPR+cbCr and "color" is concerned, but who knows ...

As one who has meddled with Lab color settings for images in Photoshop, a form of YPrPb, I can say that the adjustment of Pr and Pb is not nearly as intuitive as is RGB manipulation. You have been warned!

I think I tried to say a little bit about how it's not very "intuitive" above in my last post -- but, at least for me, it wasn't an extremely difficult thing either, as the values didn't need to change much. Wanted to try to say more about that, but would have been very difficult to explain, and I didn't really take extremely specific notes concerning what exactly "happened" either ...

For instance -- You can't just adjust CRGN with say "Red" gun without effecting what "happens" if you were looking at "Green" gun AND even the "blue gun" (if I recall to a lesser degree) as well, ... You just have to keep "moving" between them (RGBS 1/2/4 AND I found looking at RGBS #6 helped me as well) and checking to see with all 3 guns on indivdually to get a good idea of all the "effects" caused by a change in value + or -1 for CBGN/CRGN or YGN(and again, I did not have to change YGN). Then again, I didn't find it THAT difficult -- Honestly, it took me maybe 7~10 minutes to find the values I'm using for CBGN~YGN = 6-6-5 given the factory defaults of : 4-5-5, and only a few more mintues on each scan rate trying to "improve" those just to see if I could ... and again, I couldn't, and there really wasn't a need to ...

Using the AVIA color tests, for the most part (and I'd think DVE would have worked just as well) I think did help so I could be more "precise" about it, however, as just looking at "standard" Color bars and doing this, as maybe it's just me, but at least I find to be more difficult to be as precise -- especially with "red Only" gun+trying to match as closely as possible the "hue" as well as brightness of the L+R "side" of the pattern that shows up ....

I don't know why this is given various "errors" that may have been introduced which you've mentioned, but also, those DVE Color bars on memory stick/1080i also seemed to work awfully well for me, with MTRX=1, inside SM as well as checking it with DVE filters outside of SM. Using those also indicated that I could find no better setting for CBGN/CRGN than what I found at 480p via DVD with AVIA,(or - not even if I moved the Hue or Color sliders 1 value to Left or right) and also seemed to indicate the same thing I was seeing with the internal QM 1080i+720p color bars.

Anyway -- As they say "YMMV"(your mileage may vary), and certianly it may turn out to be completely different for you ---- It's really no big deal, if you don't get an improvement with it, they probably shouldn't be changed from the defaults anyway.

KenTech

01-25-06, 02:45 AM

Factory setting for "SPIC" on my set was probably the worst single "adjustment" from the factory of all the settings I've come across. First thing I adjusted.

So, imagine if the set defaults could be so off for "SPIC" how they could also be off, if perhaps only just a little bit for CBGN/CRGN as well ... maybe even YGN, although I did not change/need to change that one ... When my set was new, I had to set up the black level for Brightness=31 by boosting 2170P-1/SBRT from the idiotic 16 as-purchased to nearly 38, and then I redefined the three _DRV settings for more appropriate white levels for Picture=31. Black crush was gross out of the box! Yet several other settings were and are still perfect -- Hue and Color, for example, for most inputs.

I believe different groups of default settings were specified by different committees at Sony. They didn't talk to each other. They're in different cities, even.

My greenish-yellow shadows for HD had to be tamed by large changes in 2170P-1/CBOF and CROF. What were they thinking? And I have used SPIC to balance the apparent "brightness" of the picture among the inputs. HD was much too bright, and 480i thru HDMI too dim. Sheesh! All water over the bridge, now . . .

KenTech

01-25-06, 01:53 PM

So What Are Coring and Limiting?

In various descriptions of the service mode parameters provided by Sony, we see the words “coring” and also “limiting” or “clipping.” As an electronics engineer, I understand limiting/clipping: Pass a signal up to a certain amplitude, then no more. If it's instantaneous, it's clipping. In the context of image manipulation, it means: place a limit on how much of a specific effect will be applied. I have elected NOT use limiting, preferring to see the results of experiments where effects are applied in a *linear* fashion.

Examples of where it is implemented: 2170P-3/VMLM (limits amount of velocity modulation) and in the MID5 table where the code ends in “L,” e.g. MHYL.

But “coring”? Its effect is subtle, but I have wanted to understand it. Than I stumbled across a definition at videohelp.com in their glossary:

“For noise reduction:
“Coring is used to remove fine detail information that does not contribute significantly to the detail of the picture but which adds noise to the image. Imagine the detail information viewed on a scope. About the baseline you'd see primarily the noise information, with the detail extending beyond that. Now imagine that you sliced (or cored) this signal so that only the information above the noise on the baseline came through. You would be left with most of the detail information intact but with much of the noise information removed. The coring adjustment determines how far from the baseline the detail information removed. You want to use just enough coring to reduce the noise in the picture but not so much that the fine detail in the image is affected.”

Right! I think Sony means “threshold,” below which no effect is applied, like inverse-limiting at the *low* end of the intensity scale.

So, imagine the following example: a video picture of an ancient castle wall of rectangular stones, with a bit of gray sky at the top. (Think Rudy Maxa's “Travels in Europe.”) Let's use sharpening as an example of an effect.

The detail in the stones (sandy, grainy texture) has a brightness range of, say 5%, the cracks between the stones 20%, and the edge between the top of the wall and the sky 75%. If I LIMIT detail enhancement, a certain amount is applied to the stone texture, 4 times as much to the cracks, but NOT 16 times as much to the sky boundary - that might cause outlining or overshoot big-time. So I limit it to, say 10 times as much as the stone texture. The example can be extrapolated to other enhancement parameters; you get the idea: Apply up to this much, but no more.

It appears that “coring” might mean: Don't apply any effect until a threshold is *exceeded.* Example from the stone wall: Set coring so the brightness range of detail has to exceed 10% before it is sharpened. So the stone texture (5%) isn't changed at all, but the cracks (20%) are sharpened, and so is the sky boundary at 75%. Of course limiting can also be applied, making the effect quite complex: Start enhancing at 10% but stop at 50%, for example.

To keep the effects I have been experimenting with linear, I have tried my best to set any coring and limiting to zero, i.e. no effect. Some charts I have for earlier TVs (e.g. XBR2 series) show OFF = 0 with increasing amount as you go from 1 to 3 (e.g. VMCR). But I have no specific descriptions for the MID5 coring parameters; I have presumed that 0 = no effect, but I've been fooled before.

Bottom line: For these nonlinear effects, one cannot *assume* that 0 = no effect. In the case of limiting, 3 = no effect. For some other nonlinear parameters (2170P-3/SHF0) 1 seems to inhibit the effect, and 0 enables it. Go figure. Same for some of the bandwidth or low-pass filters: 1 or 3 is the “inhibit” setting, not 0.

So I will run a few experiments using test patterns of varying contrast to see if I can confirm the “zero” settings for these “coring” settings, and I'll report back.

Arguments for applying these nonlinear parameters:

(1) Limiting: CRT spot is finite in size and is a mushy little blob. Therefore enhance sharpness of fine textures, which are at a disadvantage, but lay off obvious bright edges and boundaries.

(2) Coring: Constant video noise that is generated in the signal chain somewhere will be enhanced by sharpening low-level texture, so sharpening the picture will just make it grainier-noisier-scratchier. Suppress sharpening until there is some real video detail that needs it.

Comment: Trouble is, (2) usually looks really *false,* sort of "pasty," and modern "3D" dynamic noise reduction is much better, averaging grain-noise over several frames to smooth it out (it's random) while retaining real image texture (unchanging from frame-to-frame). Some of this is already built-in to these Sony sets. I see coring as superfluous, so far.

williamtassone

01-25-06, 02:31 PM

^^ Ya'll able to keep with all this? ;)

Is it just me who's shaking my head- this thread gives new meaning to the term complexity

KenTech

01-26-06, 03:53 PM

After I wrote the previous article, I ran some controlled experiments to confirm the effects of the “coring” parameters and what settings disable them.

I put up a still from Title 17 on DVE (city office buildings and sky) and temporarily adjusted the MID5 column that was active to MHYL=3, MHYE=7, and MHYO=1, to maximize sharpening of the image. I noted that, as I varied the “coring” parameter, MHYR from 0 to 3 and back, sharpening of the fine grain in the sky was suppressed at MYHR=3 and increased as I backed off to 0. Sharpening of the high-contrast detail on the buildings was not affected. This is evidence that the coring is at zero at MHYR=0, and that is the setting I recommend in my published chart. Nice to finally understand it! It seems to work just like I described in the article — sort of a pasty smoothness I dislike.

Next I put up a horizontally-stepped grayscale pattern (vertical bars), reverted to the normal settings for sharpness, set ClearEdge VM at Maximum, and watched what happened as I varied 2170P-3/VMCR from 0 to 3 and back. I went from VM=Max to VM=Off and back to make sure I saw the effect of VM. Result: VM has no effect on the low-contrast boundaries between the grayscale steps at VMCR=3, and the effect increases as you lower VMCR to 0. VM on high-contrast boundaries (another pattern) was not affected. Conclusion: VM “coring” is off at VMCR=0, the setting I have been using and recommending. But now we know for sure that it’s Off!

The effects of sharpening and/or VM seem most natural when applied evenly across the contrast range. Although I can see a point in, perhaps, *limiting* these effects for high-contrast detail, I can see no justification for doing so on low-level detail, and so I will continue to recommend “coring” parameters be set to 0 (= Off).

I note, now that I’ve been sensitized to the subject, that there are “coring” parameters in the 3D-COMB group: DYCO, DYCG, DCCO, DCGA, YHCO, and YHCG. (O = coring amount, G = “gain.”) But I’m not going to spend any time with these any time soon, as they apply only to SD analog broadcast, CV (composite), and YC (S-video) inputs and don’t affect the high-quality sources I use. Further, some preliminary tinkering reveals no big changes to images when some of the parameters are wildly varied. So they’re low priority for now.

Silimar, easier tests confirmed that the "limiting" parameters have minimum effect when set to 3. That, too, is what I had believed but hadn't rigorously confirmed, but it's easy to see in simple tests.

KenTech

01-26-06, 04:04 PM

Is it just me who's shaking my head- this thread gives new meaning to the term complexitySorry, but this *is* the deep end of the pool! Lifejackets can be rented at the concession stand. Mention "Sony service mode" to the guy behind the counter, and he'll give you a big sympathy discount. :-)

Nitewatchman

01-26-06, 06:38 PM

Lots of thoughts, but little time to post about all of them, so I'll just comment on a couple of them.

KenTech,

Thanks for the info+for doing the tests on "coring"+ it's effects with VMCR/etc.. I am wondering if there is anything that may be useful for reducing sharpening effects of "high contrast" boundaries.

Also, If I ever get around to it, I might check out DYCO, DYCG, DCCO, DCGA, YHCO, and YHCG a bit, as I do still use analog OTA via RF and SD via Dish network receiver+S-VHS deck via svideo. Just in case I might find some small changes beneficial.

My greenish-yellow shadows for HD had to be tamed by large changes in 2170P-1/CBOF and CROF. What were they thinking?

Isn't it interesting the differences we get from these sets in some cases? I've not had any sort of issue with anything like that, with any input/scan rate/etc - even when I've went and "looked for it". As mentioned earlier, only thing I've noticed any sort of issue concerning my CBOF/CROF set defaults involve the very slight change in hue in my Grey background I have set up for "Twin view" when I select 480p or 1080i sources in the left window.

So, don't know what they were thinking but perhaps with one exception(720p, not 480p or 1080i, interestingly enough) the set's defaults for CBOF/CROF for my set seem to be either close, or right on the mark. Shadows/blacks are 'black" not contaminated with other colors. Except for the seemingly odd factory default for my set of CBOF/CROF=31/31 for 720p -- My set defaults for everything else for YOF~CROF are exactly as in service code listing.

Although I didn't actually "check it" in any specific way, when I was adjusting 2171CXA CBGN~YGN to vastly different values in order to quickly detirmine which scan rates/inputs those controls effect --- I noticed that I could imagine it's possible that "vastly different settings" for CBGN~YGN might significantly effect what you'd need for CBOF/CROF. Again, just speculating as that might not be the case since I didn't specifically check it, and only was looking to see what inputs/scan rates were effected by CGBN~YGN.

I can say the small changes I actually made to CBGN/CRGN for actual use did not require different CBOF/CROF settings. Just speculating a bit, but it seems like "different" greyscale, and/or perhaps even a different adjustment of HV control on flyback transformer for different sets could perhaps also be other factors involved that might explain why different folks may find they might need "different" CBOF/CROF settings.

Last night I went back for about the third time(or was it 5th?) and checked CBOF/CROF settings for several scan rates. I mostly checked it by setting color slider at maximum and slightly increasing brightness slider so I could more easily see what was going on. With 480i/480p from DVD I used dark, or mostly "dark" test patterns, such as THX Optimizer "brightness" test screen. For 1080i, luckily I have a couple of stations here which send black(along with a small ID bug) on their 1080i HD subchannels when they aren't sending programming on it and are in 4-channel SD multicast "mode".

I found - as has been the case previously -- by looking at this by eye, I couldn't improve on the factory CBOF/CROF settings for 1080i or 480p. Certianly -- looking at "black" with color all the way up, I could see the difference if I lowered CROF from the factory default of "47" for 1080i to say "44" - Even though I thought "44" looked like the best value, I couldn't really tell for sure by "eyeballing it" whether "44" or "47" was the better setting, so I left it at default(47).

Also note that even if I set CBOF/CROF to 31/31 for 1080i - I wouldn't say I am really seeing much color contamination -- but, those at 31/31 with use of a "pure"(hopefully anyway) black screen there is a very, very slight amount of it, and it is on the "green" side of things.

I did end up changing my 720p CBOF/CROF from default 31/31 to match the defaults in the servicecode listing. My (admittedly "flawed" reasoning) reasoning for this based more on #1). "faith" in the service codes listings matching the set defaults for everything else on my set, #2).the slight effects on the "hue" in my "twin view" grey background - and matching what is happening with 1080i+480p in that regard, and #3) - Per #2 as well - what I saw with "black screen+ID bug" from the PBS affilates at 1080i when they aren't sending programming.

I didn't have a good way to "look" at it with 720p other than looking at the black pillar bars during upconverted SD programming from local stations sending 720p. I did try that as well and did manage to catch "black" being sent for a couple of seconds from 720p stations on a few occasions. From what little I could tell with that, as well as pertaining to my "reasoning"(especially #2+3 mentioned above) the service code listing default(CBOF/CROF = 45/47) did, nevertheless appear to be a better choice than the odd CBOF/CROF=31/31 default setting. Also, after changing it to CBOF/CROF = 45/47 I noticed what I'm seeing during program material(including the extremely slight change in flesh tones/etc) from 720p stations seems to better "match" what I'm seeing with 1080i stations - although it's such a slight, slight difference it's hard to say if that's more because I'm just "seeing things" I want to see. It may be.

I doubt If they'll need to be changed much, but, 'll revisit CBOF/CROF in the future as possible if/when I happen to run across B&W programming, and I might unhook one of my external ATSC receivers from another set and hook it up to the Sony via component connection to see what happens -- I can send whatever scan rate I want from it(1080i/720p/480p/etc), and It also actually sends "video black"" if there is "no signal". Would also be interesting to see if factory CBOF/CROF setting for 720p from V5/V6 component differs from the 31/31 I had for ATSC 720p. CBOF/CROF values for 480p ATSC+480p Component are slightly different - per service codes listings, and as I confirmed via 480p via V5 and internal 480p ATSC test pattern.

I have used that STB/ATSC receiver to set similar R+B offsets for 720p/1080i on another display(Samsung TXN2668WHF) I was using it with. Although I don't necessarily expect on the XBR960 the CBOF/CROF settings will necessarily be the same via component V6 as what internal ATSC needs. I do recall that It did work very well to adjust the R/B offsets on that other set, and that very different values from its factory defaults were quite necessary.

---------------------------------------------------------

Oh --- BTW - I'd mentioned something about this earlier -- I found it does work to put up 480p ATSC internal QM test pattern to adjust your image processing settings and P2170-4 GAMM/BLK/etc. per pic mode. While you can't change pic modes with the QM test pattern up, it uses whatever pic mode you were using before you went into QM and put up the test pattern. Since I don't have a 480p ATSC signal available at current, I went through all the pic modes for 480p ATSC and set up the relevant settings to do my "hijacking" of vivid/standard/movie for 480p, so if/when a 480p ATSC OTA signal does pop up, I won't have to do it, then - besides perhaps making small adjustments for MID5 column/etc.

I'll probably do the same thing for 720p/1080i via component by hooking up one of my "STB" ATSC receivers, as all I've used 720p/1080i with on the sony are internal tuner and memory stick. The 720p/1080i V5/V6 specific settings are therefore likely still at factory defaults. If there were internal test patterns that use HDMI input, perhaps I could do the same thing for settings specific to "HDMI" as well. As someday, it's likely I'll be using something via HDMI I'd think, and I think it would just be nice to have as much "work" as possible already done.

KenTech

01-26-06, 08:12 PM

I am wondering if there is anything that may be useful for reducing sharpening effects of "high contrast" boundaries.

Oh, yes! As I mentioned, that is what the "limiting" parameters are all about. Just set up sharpening effects for good, low-contrast textures, and then try stepping the appropriate limiter parameter down from 3.

Also, If I ever get around to it, I might check out DYCO, DYCG, DCCO, DCGA, YHCO, and YHCG a bit, as I do still use analog OTA via RF and SD via Dish network receiver+S-VHS deck via svideo. Just in case I might find some small changes beneficial.I use them, too, but demand a lot less from those sources. I will be interested to hear about anything you discover. (In fact, if we keep this up, I fear there will be only 3-4 of us ever reading this thread! Oh, well . . . after we're dead and gone, the written legacy of our exploration and sacrifice will live on on some storage device. ;-) )

KenTech

01-26-06, 08:19 PM

I noticed that I could imagine it's possible that "vastly different settings" for CBGN~YGN might significantly effect what you'd need for CBOF/CROF. . . . except that the "GN" parameters are specifically *gain* or level controls, affecting colors in the whole brightness range, and the "OF" parameters are for offset, affecting mainly the darkest tones. Color offset misadjustment really screws up grayscale!

KenTech

01-26-06, 08:29 PM

Certianly -- looking at "black" with color all the way up, I could see the difference if I lowered CROF from the factory default of "47" for 1080i to say "44" - Even though I thought "44" looked like the best value, I couldn't really tell for sure by "eyeballing it" whether "44" or "47" was the better setting, so I left it at default(47).

Also note that even if I set CBOF/CROF to 31/31 for 1080i - I wouldn't say I am really seeing much color contamination -- but, those at 31/31 with use of a "pure"(hopefully anyway) black screen there is a very, very slight amount of it, and it is on the "green" side of things.I once thought that the "black" the station broadcast between program and commercial -- or what surrounded a 4:3 commercial, or what I saw when I had punched in a digital station but it hadn't given me a picture yet -- was a ture black, but now I'm not so sure. Sony's service instructions are to input a grayscale from a HD signal generator (yah, like we all have one of those!), and make sure the grayscale is without color shift throughout the brightness range. When I finally saw PBS broadcasting some really neutral B/W video on HD, I immediately invoked service mode and cranked Color to max to see if I could improve on previous settings for 2170P-1/CBOF and CROF.

I could and did! The B/W video I now encounter routinely in HD commercials and programming is truly neutral, and -- guess what -- some of those "black" backgrounds are slightly tinted. Hmpf! So much for assumptions!

KenTech

01-26-06, 08:53 PM

I'd mentioned something about this earlier -- I found it does work to put up 480p ATSC internal QM test pattern to adjust your image processing settings and P2170-4 GAMM/BLK/etc. per pic mode. While you can't change pic modes with the QM test pattern up, it uses whatever pic mode you were using before you went into QM and put up the test pattern.This is very good advice. I recently completely recalibrated my set's color for all signals by using these patterns, in partucular revisiting the settings for 2170P-4/RYR~GYB and reestablishing correct color levels with 2170P-4/SCOL.

Of particular value is pattern #6 in each series (06, 26, 46, etc.). When viewed in blue-only (2170P-2/RGBS = 1), the top and bottom halves of this pattern should be exactly the same brightness so you can hardly see the boundary between them. Since one or two fo the SCOL settings are commpn to inputs that may differ in color intensity (HD, memory stick, HDMI@480p, for example), those inputs have to be balanced against each other elsewhere, say in 2170P-3/UCOL. I also used calibration DVDs to confirm some inputs.

Result: Better color balance among the inputs, my alleged "blue push" is gone (I sorta miss it...), and some broadcast SD less garish. Amazing what a little tinkering will do. Some other chap has complained about a "green push" and "too-saturated" color even after calibration. I wonder if thorough attention to those internal patterns to "tune" the set internally may help. Small errors in GYR and GYB setup add up to noticable green push.

So now RYR~GYB are 14-15-6-4.

In the CXA2171 group, I could not find anything wrong with my CBGN~YGN controls as they were -- but they're not set to the chart defaults, either. More like 6-6-6 for HDMI instead of 3-3-3. I'll bet that as long as these three parameters are identical, you won't see any change except to the whole picture's contrast (white-brightness).

KenTech

01-26-06, 09:04 PM

Correction: In my post #707 I mentioned changing 2103-1 #16/SSMD to 3 from the default of 0.

I now think this change is superfluous. This timing parameter is called "slice level" in some documents, and I thought I had once seen an improvement in picture quality (AVIA 200TVL resolution test pattern) when it was changed from 0 to 3. I now can't see any difference, though 1 and 2 are horrible. Setting 0 is the "auto" setting, and I think I'll leave it there.

So ... regarding changing SSMD to 3: Never mind!

Nitewatchman

01-26-06, 11:24 PM

Oh, yes! As I mentioned, that is what the "limiting" parameters are all about. Just set up sharpening effects for good, low-contrast textures, and then try stepping the appropriate limiter parameter down from 3.

Thanks, that brought it together for me. I think I may need to experiment a bit more with some slightly different values for the other MID5 settings we are using(such as MHYL~MHYO) first before looking elsewhere.

Probably have to do that "clear out the MID5 column thing", to make a few more columns available for that first. I'm trying to get "everything else" (which all mostly involve minor "little" things, or things of little "current" importnance there is just a lot of them I want to get "out of the way") taken care of before trying much "fine tuning" of your latest image processing recommendations, or trying further experimentation with it. So, it might be a while.

I think I do recall messing with MHYR (and everything else in MID5) a little bit many months ago when experiementing with various programming from 480p DVD, which in my case required a little more "sharpening" than your "old" MID5 column #60 or #61 column recommendations. That experiment Included adjusting most other MID5 settings for maximum sharpening effect(but not doing much in 2170P-3 and with SYSM=3), then "backing off on things" I didn't like as much as possible. I ended up also not liking anything but "0" from MHYR, but maybe I should try revisiting it.

What I seem to want to try improving doesn't seem to involve VM, as, this little "thing" I may be seeing is there with VM turned Off as well. It's probably more just of me seeing things after having gotten used to the "old" settings, and/or something along the lines of needing to fine tune MYHE/MYHO and MVYL/MVYE a bit for different sources. As again, the more I look at the set with the "new" image processing settings, the more I like them -- in every way.

Hard to explain, but what I think I want to do is - given use of your "new" (Post #707/etc.) image processing setting recommendations --- along the lines as you put it for "limiting" -- I want to keep the "enhanced sharpness of fine textures, just as they are, but lay off a little more obvious bright edges and boundaries". As, everywhere else it's just beautiful, just in certian circumstances - not necessarily in a "input/scan rate/source specific sense---- I seem to be seeing a bit of something I don't like concerning "bright edges" and boundries ... Instead of going into more detail on it now, probably best need to wait to see what I find out with fine tuning of MID5 columns/etc, as I'm thinking that might do the trick.

. . . except that the "GN" parameters are specifically *gain* or level controls, affecting colors in the whole brightness range, and the "OF" parameters are for offset, affecting mainly the darkest tones. Color offset misadjustment really screws up grayscale!

I'm certianly not disagreeing with that. Just reporting my observations and experiences.

In my case I did do quite a bit of cycling through inputs with color at "0" after changing CBGN/CRGN from default of 4-5 to 6-6 - and I'm still getting pure "Black+White" so to speak. If there were any differences in greyscale between different sources/etc/ before, (which if any - MAYBE involved the very, very slightest bit of pink/red contamination at 480p/1080i - but if that was there before it was probably just more my imaginiation), there's less difference now. and, It seems very easy to spot little differences with greyscale with color at "0" when changing between a analog station from RF input on channel 12, to the PSIP remapped channel 12.1 with the same programming, but upconverted.

Maybe I'm wrong Ken, but I think slightly modifying CBGN/CRGN from "4-5" to "6-6"was a good thing for my set. I've certianly noticed the improvement, but have yet to see any problems with it. Maybe I'll find something later on though that indicates "4-5" is better, If so I'll post about it. And Again, I don't expect that will be the case for everyone, perhaps especially if your factory defaults for those values(and perhaps YGN as well) are the same.

once thought that the "black" the station broadcast between program and commercial -- or what surrounded a 4:3 commercial, or what I saw when I had punched in a digital station but it hadn't given me a picture yet -- was a ture black, but now I'm not so sure. Sony's service instructions are to input a grayscale from a HD signal generator (yah, like we all have one of those!), and make sure the grayscale is without color shift throughout the brightness range. When I finally saw PBS broadcasting some really neutral B/W video on HD, I immediately invoked service mode and cranked Color to max to see if I could improve on previous settings for 2170P-1/CBOF and CROF.

I could and did! The B/W video I now encounter routinely in HD commercials and programming is truly neutral, and -- guess what -- some of those "black" backgrounds are slightly tinted. Hmpf! So much for assumptions!

Here's another one. I receive Local ATSC/ HD stations from 2 different markets. One CBS HD station has a very slight, greenish tint to it - most noticable in shadows - such as say a dark scene involving actor wearing a dark suit. It's the only station out of 15 local ATSC stations I receive that seems to have this sort of issue. It's especially noticable when I compare it to CBS HD station from another market, where that greenish tint isn't there, even in the same scene. It's there on the Sony, on all other displays I own, it's there if I output 16x9 NTSC video at 480i+record to SVHS - especially noticable if I record portions of the same program from both stations. Now, since it's a very very slight thing -- seems that either I'm seeing things, Or noone else who posts on my local thread has noticed it yet ... LOL ...

Given the 1080i assumed "black screen" from the station I'm referring to, and the default values, Neutral grey is exactly what I get throughout the entire brightness range, even with color turned all the way up.

So, what are the chances I'd get that neutral grey throughout brightness range if it WEREN'T black, and CBOF/CROF weren't set correctly(or close)? In other words, How does it "just so happen" that if it isn't "black", It nevertheless, Just so happens to perfectly match my CBOF/CROF factory defaults? And, as another example I'm seeing the same thing you are for B/W video in HD commercials.

I agree with you about "assumptions" - we all remember that old saying -- but sometimes I think one can make good, reasonably educated "assumptions", "bets" or guesses, and sometimes, with this sort of thing it is all one has to go on, and we just do our best.

Anyhow, I will certianly recheck CBOF/CROF as possible when "good" B&W content is available(and I have had the oppurtunity to check it with B&W from PBS HD channel) ... but, at this point If there is any "color contamination" of blacks or shadows/etc. going on because of CBOF/CROF settings, I can't see it. And that's what is important to me, anyway.

Also, to "start off" with - at least with my ability to "eyeball it" I've got the same "B&W" with greyscale with color at "0"(as best my eyes can tell anyway) among all inputs/sources. And again, I have also run into the SAME issues you have had with this, and adjusting it's R+B offsets(they also don't effect Greyscale/btw) worked perfectly -- just on another set/model from a different manufacturer, Not this one.

I'll bet that as long as these three parameters are identical, you won't see any change except to the whole picture's contrast (white-brightness).

Didn't try it that way to see what happened, but that makes sense to me.

Of particular value is pattern #6 in each series (06, 26, 46, etc.). When viewed in blue-only (2170P-2/RGBS = 1), the top and bottom halves of this pattern should be exactly the same brightness so you can hardly see the boundary between them.

Never thought of that! Thanks for pointing that one out -- You're right, that is an excellent pattern for Blue only gun. And I thought I was Done with "color" ...

So ... regarding changing SSMD to 3: Never mind!

Thanks for the update. FWIW, I didn't notice a difference between 0+3 when I changed it, except If I recall correctly the pic shifted a bit to the right(I think) when I was moving from "0 to 3".

Nitewatchman

01-27-06, 04:10 PM

I'll bet that as long as these three parameters are identical, you won't see any change except to the whole picture's contrast (white-brightness).

A update on this. I've confirmed the above appears to be the case - albeit via a bit of an "odd" happenstance.

One of the "little" things on my list to do was to go back and center the "Picture" slider at 31. Although I had done that long ago for the user menu slider for "brightness=31" slider with SBRT+ 2170P-3 UBOF for the various inputs/scan rates /etc - --- and had also long ago already "balanced out" Contrast from different sources(to best of my ability to do so anyway -- getting the "RF" input "just right" seems especially difficult, for instance, and the way it was, SPIC for 720p/1080i really needed to be about "-1") using P2170-4 SPIC, and for RF+CV/YC, 2103-1 SCON(for twinview/index its in 2103-2) :

Back When looking at Greyscale and setting white balance/etc., I'd just ended up with a Picture=24 (or 25) setting, and left it at that, for one of those "little things" to go back and "center" Picture control, later - probably with P2170-4 "SPIO".

Ok, so last night I went to do that So I could get Picture slider up to 31. And notice my factory defaults for "SPIO/SCLO/SHUO are : 10-6-7. My SPIC settings ranged from "0 to 4". From the factory it was funny as the scan rates inputs that actually need the LOWER SPIC settings were set for HIGHER settings instead! for instance, factory SPIC was "7" for 1080i, and "4" for RF .... THAT definitely seemed a little backwards, to say the least.

So, I end up with "SPIO=7" instead of factory default of 10. Then, putting DVE color bars up via 480i component/DVD, I notice, strangely enough that that caused my "color" settings to change a bit, leaving blue a bit undersaturated ... This was easily fixed by adjusting "SCLO=7", bumping it up from the default of "6". Now I have SPIO~SHUO = "7-7-7", and SPIO still may need to be lowered a bit more - not only to give me the "right" White balance with Picture=31, but also depending upon what it ends up doing to color, might be better to lower SPIO more so I can raise SPIC a bit, since I have some of those values at "0", and I really might need a "-1" for 720p/1080i HD. Notice it does indicate SPIO at "4" in service code listing, and perhaps SPIO at "4" might be a good spot for me.

HOWEVER -- now here comes the mildly interesting part. As I noted, Lowering SPIO from factory default of "10" to "7" also effected Color balance slightly. I really wasn't expecting that. But, no big deal, raised "SCLO" one notch and that took care of that for everything EXCEPT 480p/720p/1080i --- Hmm seems familiar ... SO, I revisit CBGN~YGN again. What do you know -- CBGN~YGN values of "5-5-5" work best with SPIO=7, instead of, with factory default of SPIO=10 and CBGN~YGN 6-6-5(factory values 4-5-5) for achieving color balance for 480p/720p/1080i that "matches" what I get with the same RYR~GYB values from other inputs ....

So, considering that, now that I have the values for CBGN~YGN being all the same, and considering that "SPIC=0" for 1080i/720p really needs to be more like "SPIC=-1" in order to balance contrast among sources ..... I think - well -- why not Lower CBGN~YGN to 4-4-4 , and see what I need to do for "SPIC" for 480p and 1080i? --- At least without spending a lot of time looking at it -- This seems to be working quite well, and this is what I ended up with for the new SPIC settings(note only needed to change the 480p and 720p/1080i settings) : 720p/1080i(and everything else except 480p) - SPIC = 4. 480p SPIC = 9 ......

KenTech

01-29-06, 06:31 PM

Ok, so last night I went to do that So I could get Picture slider up to 31. And notice my factory defaults for "SPIO/SCLO/SHUO are : 10-6-7.Interesting how we have skinned this cat in different ways. I was very disappointed early-on in my contrast and black-level settings. I fixed black level (Brightness = 31) with a combination of SBRT (now about 30; was 16!) and the three 2170P-1/_CUT settings. But back then I hadn't figured out the sub-settings of SPIO-SHUO, and so I merely upped the _DRV settings until I had a picture that, at Picture = 31, has the contrast I expected. It has been that way since last spring, and so I haven't been tempted to mess with the offsets you have tinkered with.

Just goes to show how many of these settings have the sane effect in the video chain. They are distinguished mostly by scope: In 2170P-3, they're arranged by input class *and* video; 2170P-4/SPIO~SHUO are global (one setting for all); 2170P-4/SPIC~SHUE are per internal signal-path; 2170-1/YOF~CROF are by video and input class. And so forth. Hard to keep straight for non-wireheads!

I am thinking of creating a chart or two showing *related* settings in their various locations so one can record changes easily instead of notating one of Sony's charts. This is what I did with the image-processing settings in the chart attached to #707. But the gross parameters of color, black level, and grayscale could use the same treatment. I'm weary of scribbling tiny little numbers and dates on the Sony charts!

OmniBeast

01-29-06, 06:50 PM

Jediphish

01-29-06, 07:49 PM

Yesterday while my D* was tuned to one of the XM Channels, which display either a black or grey solid screen (depending your user settings - mine are at grey so I can easily distinguish the sidebars created by my HR10-250 from those transmitted by an HD channel, which are black 99% of the time) . . . Anyway, while the grey screen was displayed on my 30xs955, I notice three very faint, single pixel height, horizontal black lines running from edge to edge. The first one is about 1/8 - 1/4 from the top. The next one is about 2/3 from the top, and the last one is about 1/4 from the bottom. I have to stand at couple of feet from the screen to see them. I changed to a channel displaying an actual image and noticed the lines were still there, although not as detectable on a non-uniform image.

These could have been there since the day I turned my set on the first time and I just never noticed. Anyone know what this is about? Is it the aperture grill/mask? Is it a defect? Is it something I did while changing user menus? It's not a big deal, but now that I know it's there, I think I'm going to fixate on it.

Nitewatchman

01-29-06, 09:02 PM

so I merely upped the _DRV settings until I had a picture that, at Picture = 31, has the contrast I expected.

With my set, and the factory default of "SPIO-10", I would have had to lower the DRV settings for Picture slider=31, and they're already at midrange(Both RDRV/RCUT at 32, and that was the factory value) - That's even if I set "SPIC" to 0 for the source(s) with highest contrast.

Here are the default DRV~CUT values which correspond to the default values for "warm" offset on my set(the actual warm offset values were same as in service code listing) :

32-17-9-32-17-5

Using that, and from looking at "clouds"/etc(sorry I can't offer anything more definitive than that+"eyeballing it"), after balancing contrast with SPIC+2103-1 SCON(for RF+CV/YC)/etc -- a picture slider setting of "24~25" ended up being just right, given the factory "SPIO=10" value. Also, Higher than 25(and even 25 sometimes), and my eyes easily get fatigued after a bit of viewing.

So, Factory "warm Offset" seemed awfully close to 6500K, except that it was more "greenscale" than greyscale. Only really worked if I was looking at clouds that had a "green" hue to them(and that does happen sometimes) .... So, I went about getting rid of that bit of greenish tint to the greyscale, and these are the values I now have set up for RDRV~BUT for Neutral - 32-16-7-32-17-9, which is what I use. I have modified the "warm offset" values so "warm" is as it was from the factory, for reference. I haven't really found much of a actual use for "cool" yet, although I did use it while finding the best values for "neutral".

I have SBRT=28, which was the factory default. But, the only reason why I don't have it at say "31" is because my DVD player needs SBRT=28 with P2170-3 UBOF=0. Everything else I have at UBOF=3~5.

So, in my case it looks like lowering SPIO=4 is going to be my "answer" for getting picture slider up to 31. Interestingly enough, SPIO=4 is the default listing in the service codes. Also, interestingly enough, I haven't checked this thoroughly yet, but from what I've seen so far it looks like that with SPIO=4, many other values(such as SCLO, SHUO, CBGN~YGN, SPIC and 2103-1 SCON for RF+CV+YC) will also end up needing to be very close, if not exactly the same as is shown in the service code listing .....

Given that you did need to redefine higher _DRV values, and I think I'd read previoulsy that your "SPIO=7" --- Could our CRT's really be that "different", or perhaps is it more likely could it be the HV adjustment on the flyback transformer on your set is set a little differently than mine(which perhaps could also explain why we are getting different results for CBOF/CROF from the factory defaults) ?

Is it the aperture grill/mask?

It sounds like you are seeing the wires that hold the aperture grill in place. We all can see them from a very close distance(within a foot or two) with these AG sets, most noticable perhaps with uniform grey or white screen.

Is it a defect?

No, If you are seeing what I think you are, that's just the way it is. You shouldn't see it from more than a couple feet or so away.

skersh

01-29-06, 10:59 PM

Ken,

Thanks as well for all of your posts and very helpful (and detailed) information. I have a question. I have a 960 that has a picture that seems to be shifted too far to the left. Here's what I mean. While watching the U2 concert tonight on HDNet, the only part of the HDNet logo that I could see on my screen was "HDNe"
What service codes would work in shifting the picture to the right? I have read roughly 10 pages of this thread and am either too stupid to discern which codes would work or I might have missed it. Could you help me?

P.S. I have already downloaded the 960 service manual that you provided. Thanks!

KenTech

01-30-06, 12:12 PM

This is defenitely a newb question but, all these different settings and numbers that are being talked about here, is this what an isf trained service tech. would adjust in order to make the TV look better?I believe the answer is "It depends on the tech." From what I have read, it appears that some techs are informed and motivated enough to add some of the adjustments mentioned here to their normal work. (I recall a fellow from Ohio posted in this forum that he was quite impressed in this way with the guy he hired.)

However, I can't imagine every ISF tech having the depth of knowledge about all of the different brand's and model's service-mode settings that they can just waltz into your home and dive right in, as we discuss here, especially in balancing the myriad settings against each other. Further, such practitioners will differ widely in their intelligence, diligence, and discipline. (Best get a referral from a pleased, discriminating customer.) Investing $10,000+ in exotic instruments does not guarantee competence or diligence, ISF "training" notwithstanding.

One basic argument for learning about your own TV and doing it yourself is that you are learning about only *one* TV, in depth, depending on how much of a wirehead you are. When, say, the color adjustments drift off a bit, as they are likely to do over a year, you can go back in and tweak it for a service charge of $0! Further, nothing I have read indicates that ISF techs get into the nitty-gritty of image processing, as we have in this thread. (See post #707.)

The personal-satisfaction issue is best discussed in another thread, but I don't what there is to debate. One can either be proud of their own accomplishment in calibrating their own TV (I am!), or one can hire someone else to do it for them and miss out. That goes for a lot of different tasks. I am a fearless, curious wirehead, and so this course of action has been natural for me and will be for many others who read this thread and are willing to rev up their courage a bit.

KenTech

01-30-06, 12:41 PM

What service codes would work in shifting the picture to the right? I have read roughly 10 pages of this thread and am either too stupid to discern which codes would work or I might have missed it.Don't beat up on yourself! It takes heroic persistence to find specific instructions in this thread.

The *raster* is the entire pattern scanned by the electron beam, including the garbage at the edges. Be sure to write down your original settings as you make changes.

The strategy is to (1) center the raster on the tube. See manual section 2-8. Set 2170D-3/HBLK to 0 to disable blanking of the edges. Set 2170D-2/AGNG = 2 to temporarily shrink the raster so you can see the edges and lower HSIZ if it needs more shrinking. Then adjust HCNT to center the raster horizontally on the tube. Restore AGNG, HBLK, and HSIZ to their former values.

Now (2) use 2170D-2/HPOS to center the *picture.* Use a standard test pattern from one of the DVDs, like AVIA or DVE. Finally you can use 2170D-2/HSIZ to reduce the amount lost at the edges (overscan). The illustrations in the manual, section 2-9.1 show the different adjustments.

2170D-1/VPOS and VSIZ center and adjust height of the pattern. There may be some subtleties in adjusting your 16:9 set that I can't address, having a 4:3 set, but these are the basics.

PLEASE READ the early first couple of articles in this thread for the methods and cautions about using service mode!

skersh

01-30-06, 01:19 PM

Ken,

Will do. Thanks for the advice and the help. My TV is still under warranty and I've contacted our local Sony warranty electronics store about fixing it, but the tech I've talked to has the personality of "Oscar the Grouch" from Sesame Street! He set up a time to come and pick up the set, then never showed up! Everytime I've spoken to the guy, he sounds like I'm bothering the living crap out of him and that he'd rather not be working at all. Unfortunately, Sony has told me that they don't have another warranty service technician in Amarillo. So, I found this thread and have decided to make the adjustment myself, if possible. I've also had my analog sound go out on the TV (via the analog out jacks in the back) and am using the "monitor sound out" jacks instead. I was going to have the tech work on those too. The tech did mention that the analog sound card may have gone out on those jacks and needs to be replaced. The digital sound (via coax and optical) works just fine...and obviously, so does the analog monitor out jacks. I didn't have this many problems with my old KD-34XBR2, but I guess every set is different and I sure wouldn't trade this 960 for anything. I absolutely love the picture.

Thanks again!

Steve Kersh

Nitewatchman

01-30-06, 02:27 PM

............... I recently completely recalibrated my set's color for all signals by using these patterns, in partucular revisiting the settings for 2170P-4/RYR~GYB and reestablishing correct color levels with 2170P-4/SCOL.

Of particular value is pattern #6 in each series (06, 26, 46, etc.). When viewed in blue-only (2170P-2/RGBS = 1), the top and bottom halves of this pattern should be exactly the same brightness so you can hardly see the boundary between them. Since one or two fo the SCOL settings are commpn to inputs that may differ in color intensity (HD, memory stick, HDMI@480p, for example), those inputs have to be balanced against each other elsewhere, say in 2170P-3/UCOL. I also used calibration DVDs to confirm some inputs.

Result: Better color balance among the inputs, my alleged "blue push" is gone (I sorta miss it...), and some broadcast SD less garish. Amazing what a little tinkering will do. Some other chap has complained about a "green push" and "too-saturated" color even after calibration. I wonder if thorough attention to those internal patterns to "tune" the set internally may help. Small errors in GYR and GYB setup add up to noticable green push.

So now RYR~GYB are 14-15-6-4.

In the CXA2171 group, I could not find anything wrong with my CBGN~YGN controls as they were -- but they're not set to the chart defaults, either. More like 6-6-6 for HDMI instead of 3-3-3. I'll bet that as long as these three parameters are identical, you won't see any change except to the whole picture's contrast (white-brightness).

I went through this procedure while utilizing the #6 series of QM pattern with Blue gun last night/etc., and came up with the same, or nearly the same RYR~GYB as you did - at least for best results with 720p or 1080i -- either RYR~GYB = 14-15-6-4 or 14-14-6-4. Was Hard to say with any certianity which one was better.

On my set, at 720p/1080i+adjusting SCOL as necessary with CBGN~YGN at 5-4-4, and RYR~GYB at 14-15-6-4 resulted in as close to perfect color balance/color decoding as I could imagine using the QM #4+#6 series of patterns and turning on the guns individually, as well as color bars at 1080i from memory stick. However, that didn't quite work as perfectly for 480p ATSC internal pattern - Which unfortunetly in my case needs a SCOL value that is -1 from what 720p/1080i needs for the same results and, SCOL doesn't appear to allow different values for 480p ATSC vs 720p/1080i ATSC.

So, If I ever get a 480p signal of importance from ATSC tuner, and If want as perfect as possible color for both 480p and 720p/1080i ATSC, I'd have to reduce SCOL for ATSC 480p/720p/1080i, and increase P2170-3 UCOF value for 720p/1080i. Luckily, 480p via component V5/V6 allows a seperate SCOL value, and according to color test patterns via DVD SCOL also needs to be -1 as compared to 480i via component. Have a good feeling however that I'll also probably need to balance V5/V6 with UCOF If I ever hook anything up to the component input that uses 720p/1080i. It would probably be better to go ahead+do the "UCOF thing" now for ATSC tuner 480p vs 720p/1080i .. I've just liked the convienience of having UCOF/UHFO=0 for everything .... oh well ....

Also, noticed that with test patterns via DVD at 480i or 480p via component or S-Video - after using pattern #6 at 480i(which, along with everything else except 480p needs the same SCOL value as 1080i) -- I still need RYR~GYB at 13-15-5-3 for best results.

Right now, after trying numerous slightly different "settings" my compromise is to set CBGN~YGN at 4-4-4, increase SCOL for 480p/720p/1080i ATSC an additional +1, and use RYR~GYB 13-15-6-4 for input from the internal tuner or MS, while using RYR~GYB 13-15-5-3 for everything else. Except for 480p ATSC, This is still "very" close to being "right on the money", and I have to really, really look closely to see the difference between RYR~GYB 13-15-5-3 and 13-15-6-4 where I'm using it. If I ever get a 480p ATSC signal however, this "compromise" does not favor accuracy for 480p ATSC color.

The other compromise - which appears to be the best "all around" solution is to use CBGN~YGN at 4-4-4, but not increase SCOL for 480/720p/1080i ATSC the additional +1 as required for my other "compromise" option mentioned above. This also happens to be the best solution for best results for both 480p and 720p/1080i ATSC AND also produces the best results of anything I can find for using 720p/1080i with RYR~GYB 13-15-5-3 - Which makes it appear to be the best "all around" compromise, and still seems to offer very good color balance(just not quite as good as the other two "options" I've found) for 720p/1080i.

I tried other various "combinations" of values, including for instance CBGN~YGN at the set default (4-5-5) as well as even trying slightly different values with SCOL/SHUE -- But, in my case the 3 different possible options described above worked better than anything else I could find. At this point, I haven't decided which of the 3 above described options I'm going to use "permanently".

BTW - I did manage to forget a couple of times to turn "MTRX=0" for 480p and MTRX=1 to 1080i/720p when checking this and making slight adjustjments, which resulted in a bit of cursing on my part, LOL -- Another interesting thing -- The "right"(but not the left) "Twin view" window -- which is allways going to be either from from RF NTSC tuner or from YV/VC --- is ALSO effected by "MTRX" strangely enough -- must be because when you use "twin-view" you're actually using 1080i full mode ....

...............
PLEASE READ the early first couple of articles in this thread for the methods and cautions about using service mode!

Speaking of which -- couple of hopefully helpful add'l items about working in SM I've noticed :

#1). When switching inputs or picture modes and checking values for specific settings -- at least on my set - I notice that sometimes the info isn't allways updated "quickly enough" to reflect the actual value which is there. For instance ... I have "twin-view" "vivid" pic mode P2170-3 MIDE=59, for "twin-view" pro pic mode, MIDE-53. But, sometimes, when I'm looking at the value for MIDE, switching Pic modes it may still "show" "53" when I select "vivid" picture mode for a short amount of time. So, Sometimes, when switching pic modes/etc -- you might even have to use the #1 and #4 buttons on the remote to go down or up then back to see what the "actual" value you are interested actually is.

#2). I do wonder if this one has happened to anyone else --- On my set -- Inexplicably -- without any adjusting on my part as if there was a "ghost" in my set -- Value for 2170D-1 "ASPW" for one of the various screen modes have inexplicably "changed" on their own while I was in SM and "changing channels" or changing the screen mode. This has occured even if I wasn't changing any values and was doing nothing besides "looking", and also even if I wasn't looking at any of the values in 2170-D1 at the time. One time it happened I noted I was switching channels when a value from 2170P1 was showing on screen.

It's happened about 3 or 4 times, the first 2 occurances, I thought I was losing my mind as I didn't notice it until a day or so after I had last been in SM. Then, the next time, I happened to "see it" happen when it occured while I was switching channels or screen modes. Now, after having experienced this, after finishing a "session", I allways power set back up and make sure aspect ratio is correct on all of the various screen modes.(HDzoom, "zoom", Normal, Full,/etc).

KenTech

01-30-06, 05:53 PM

#1). When switching inputs or picture modes and checking values for specific settings -- at least on my set - I notice that sometimes the info isn't allways updated "quickly enough" to reflect the actual value which is there.I have had the same experience. Further, exactly when one has to WRITE the settings varies depending on the particular code. In MID5, for example, if you make changes to a particular column, then change columns for even a moment, the settings revert. I have gotten into the habit of writing settings very soon after I change them. I am sure that some of the settings that have mysteriously "reverted" to their old settings were just not written in time.
#2). I do wonder if this one has happened to anyone else --- On my set -- Inexplicably -- without any adjusting on my part as if there was a "ghost" in my set -- Value for 2170D-1 "ASPW" for one of the various screen modes have inexplicably "changed" on their own while I was in SM and "changing channels" or changing the screen mode.I have not experienced this. ASPW is critical to the height of 16:9 HD and 16:9 960i display of anamorphic 480i DVDs on my 4:3 set, and I think I would have noticed something. Can't say what might be causing this.

KenTech

01-30-06, 06:06 PM

PLEASE READ the early first couple of articles in this thread for the methods and cautions about using service mode!I have to add here that I think the biggest fast-fingers goof one can make is to forget that you cannot change channels with the number-keypad in service mode, but are stuck with the channel up/down buttons. Fortunately, it takes a three-key combination to really screw things up. I have absent-mindedly pushed number buttons, trying to tune a cable channel with the keypad, only to look up and see the dreaded word RESET on-screen. At that point, one must do nothing, and the mode will revert by itself to SERVICE in a few seconds. Lesson: Avoid at all costs impatiently pressing number buttons for any reason. Look up at the screen. If you ever see anything besides SERVICE, WRITE, or READ, you've screwed up and have to wait for it to go back to SERVICE before proceeding. It's also easy for the thumb to accidentally drift down to the third row of buttons when you have been rapidly cycling up and down thru the codes, groups, and values.

You must avoid the drastic RESET sequences of 7 - 9 - ENTER and 7 - JUMP - ENTER, unless you understand the huge consequences and really intend this!

Y'know, one could make a case for fashioning a little clip-on or tape-on cover for that bottom row of buttons, as insurance when fooling around in service mode. Even if all it did was make them *feel different* from the top two rows, it would alert you to an out-of-position thumb.

KenTech

01-30-06, 06:29 PM

My TV is still under warranty and I've contacted our local Sony warranty electronics store about fixing it, but the tech I've talked to has the personality of "Oscar the Grouch" from Sesame Street!For the posh XBR960, you should expect that a tech will come to your house and fix a few settings under warranty. They do not have to drag the thing into the shop, sheesh!

I have read here that the least successful method is to call the Sony Service 800-number for a referral. Rather, go to a reputable dealer for your set (maybe the one you bought from) to get an "inside" recommendation as to who does good repair on these sets. Then call them and ask them to make a couple of screen adjustments under warranty -- they're paid for this, after all. Worked for me! Just before the 3-month service warranty ran out, I called a highly recommended authorized service shop quite a ways away, and they sent a guy out who made some important adjustments with magnets I wouldn't have attempted -- and with a smile.

JeffD2.

01-30-06, 08:31 PM

As a noob to service menu tweaks I want to thank all the SM experts here. I've lurked here for over a year or a so posting occasionally but never had the courage to attempt tweaks for fear of disaster. ADU's recent advice about disabling 16:9 auto detect finally made me do it.

Now I'm correcting geometry/position issues (BTW not related to the JUMP setting) and plan on doing much more, but SLOOWWWLY and carefully. Now that I've actually seen the service menu, a lot of the posts in this thread don't seem so cryptic.

I have a writtten/digital record of every change made.

To all the other noobs- READ THE PRECAUTIONS AND INSTRUCTIONS ON PAGE ONE OF THIS THREAD. It can be done, take your time, and know a way out.

Thanks again to the experts!

KenTech

01-31-06, 02:13 AM

To all the other noobs- READ THE PRECAUTIONS AND INSTRUCTIONS ON PAGE ONE OF THIS THREAD. It can be done, take your time, and know a way out.Well, *that's* worth repeating:

READ THE PRECAUTIONS AND INSTRUCTIONS ON PAGE ONE OF THIS THREAD. It can be done, take your time, and know a way out. [I like that "know a way out" part!]

ADU

01-31-06, 04:02 AM

JeffD2.

Glad the JUMP tip worked for you. The 2170D-1/APSW parameter that Nitewatchman brought up a couple posts ago might also possibly be worth a look in your situation. Not sure how this works on the 4:3 TVs, but if APSW is set to 1 for your expanded HD, changing it to 0 might reduce some vertical overscan on the picture. Be advised though that APSW is also signal and aspect ratio sensitive on my TV, like the other 2170D-1 controls mentioned in my previous post. And if it's already set to 0 for the expanded HD, then there's probably no point complicating things by messing with it.

Nitewatchman

01-31-06, 02:12 PM

The 2170D-1/APSW parameter that Nitewatchman brought up a couple posts ago might also possibly be worth a look in your situation.

An update/correction on this ... until I ran into an occurance of it again last night, but this time occuring with ASPT and "Widezoom" instead of "ASPW", I'd forgotten that APSW is not the only control where I've seen this odd behavior occur. This particular time it effected "Widezoom", and the value changed(I certianly didn't do it) for ASPT somehow to "52" from what it was set at "22" -- all on it's own, while I was "looking at" values in 2170P-4 controls. Now, note that ASPT=52 IS a value for settings I changed -- but only for "proper" aspect ratio to result for "Zoom" and "HD Zoom", which had ASPT=43 default values. And, ASPT=52 for "Zoom" and "HD Zoom" are the ONLY changes I made to the set defaults for 2170P-4 ASPT/ASPW or SCRL or Zoom.

If I recall correctly, I think I may have also had this odd behavior occur once with 2170D-1 "Zoom".

So, it perhaps appears that somehow, in certian circumstances when I'm switching channels or screen modes/etc from within SM, for each occurance effecting only one "screen mode" its perhaps reading the value for 2170D-1 ASPT/APSW or ZOOM from a column for screen mode/etc. right NEXT to it, and somehow using it instead --and making it stick until I change it back to the correct value ....

Again, This odd behavior involving certian Aspect ratio related settings "changing" on their own while I'm working in SM - Seems to be occuring when I'm switching channels(in my case among NTSC/ATSC off air sources - which are a mix of NTSC, 480i or 720p/1080i ), and/or screen modes with the "screen mode" button. I'm Switching channels with channel up/down button. Btw, another "roundablout way you can "switch" channels while in SM is via the user mode "hide/show" channel option. Although there have been occasions I've seen the change in Aspect ratio right when the change occured, I think I've also noticed that this odd behavior is not allways directly "noticable" when it occurs as apparently the issue could end up effecting "Full" for ASPW and 1080i even though you were using say, "Normal" screen mode and changing channels with the up/down buttons at the time of the occurance.

It's really very very odd, and I have no idea what would cause this to occur, as it wouldn't even *seem* like it should be possible. And, it has never happened "on it's own" outside of SM, thank goodness. Because of it, I have checked numerous values elsewhere several times and have never seen this occur elsewhere -- at this point at least I don't *think* anything besides a single value for only ONE of the columns for either 2170D-1 "Zoom", "ASPW" or "ASPT" is ever effected by this for each instance of its occurance - In other words, when it does happen, for each instance, only ONE of those controls, and only ONE screen mode(such as 480 normal, or 1080 full or "widezoom"/etc) is effected ... It just could be any ONE of them.

I have also noticed that at times when in SM, NTSC analog channels(although strong RF signals are actually present on those channels ) show up as just "snow" - especially if I'm just quickly changing channels --- that never happens outside of SM, and never happens inside SM with the ATSC tuner/signals ... And, I also noticed a couple of times(but not allways) when using the "QM" patterns, that afterwards - with ATSC off air channels only, the Screen turns "pink" until "something" gets "reset" - if nothing else, cycling power fixes this when it has occured.

I don't know if it is clear whether or not the post at following URL (and/or other posts from same poster from around this time) might indicate that this fellow may have experienced the same issue, but reading some of his posts from around that time, it seems at least posssible that this issue, "unknowingly" was effecting his set as well :

http://www.avsforum.com/avs-vb/showthread.php?p=6068189&&#post6068189

Nitewatchman

01-31-06, 03:04 PM

An update to previous posts concerning one of the "little things" on my list involving centering "Picture slider". Well, as it turns out, this turned out to be a "big thing"(and not really necessary, as I don't really care if the "picture" slider says "22" or "31", all I really care about is how it looks), and I've ended up going back to just using Picture slider at a lower value instead.

Upon further observation I found adjusting P2170P-4 : SPIO didn't really work for me -- From what I can tell, It doesn't appear to "map directly" to picture slider, as is the case with SPIC, and it appears to affect 'contrast'(and maybe gamma) in a different way than "SPIC" or moving the picture slider. Since I can't detirmine for sure exactly what SPIO "does do", I've decided to return it to set default value of 10.

I also tried lowering P2170P-1 "DRV" values. While this almost "worked" well enough, it didn't work "enough" because(even with SPIC at "0" in some cases) I would have needed a BDRV value at just slightly less than "0", to get "picture=31" which, of course isn't possible.

So, unless there is another control(besides "SPIC") in SM that allows for more "crude" adjustments to contrast than SPIC, and that maps directly to "picture slider" I'm not aware of, looks like I'd have to adjust the HV adjustment on Flyback transformer in order to be able to center picture at 31 properly -- and, I don't want to mess with that, as for just a couple of examples : 1). then, for example, I'd have to completely redefine the DRV_cut values, 2). I'm guessing I'd probably have to mess around more with the YOF/CBOF/CROF offsets, and 3) - most importantly -- there's really no reason to, since lowering Picture slider does the trick.

I did end up settling on values of picture slider of 21~23 rather than 24~25 I'd used previously, as I noticed 24~25 was still just a tad too fatiguing on the eyes. Also, I ended up going through balancing contrast from different sources/inputs again using SPIC/etc(the highest contrast source is now set with SPIC=0) - and I'm glad I did as I was never really satisfied with what I'd done on this previously.

One problem I have is that I don't spend a large amount of time watching this set, as it's more the main set the entire family uses. Therefore, it is more difficult for me to notice any small things that may be apparent when watching programming material that need to be changed. So, The past couple of nights I've worked in SM, I've been trying to concentrate more on also "watching" quite a bit of programming from various sources, and that helped me to realize I still needed to bring picture slider down a couple more notches, as well as do a better job balancing Contrast from different sources.

KenTech

01-31-06, 07:07 PM

. . . looks like I'd have to adjust the HV adjustment on Flyback transformer in order to be able to center picture at 31 properly -- and, I don't want to mess with thatThe most important reason is that particular (actually Screen or G2) voltage is set with a voltmeter to a 170V specification, not by any visual criterion. See section 2-5 of the service manual. Thus, except to rescue a monitor that is dimming from old age, it's not a setting that should be "tweaked."

From my previous experience tweaking with this adjustment on computer monitors, it significantly changes the threshold at which the electron guns begin to activate, i.e. it would interact most strongly with _CUT and SBRT settings and would likely affect all three guns nearly alike.

ADU

01-31-06, 07:22 PM

Updating some of my earlier remarks re gamma and the CXA2151/2171 controls on pages 4-7 to try to better reflect the most recent info we have (hopefully in such a way that it reduces rather than adds to the confusion, while still maintaining some level of continuity in those earlier posts.)

Nitewatchman

01-31-06, 07:56 PM

The most important reason is that particular (actually Screen or G2) voltage is set with a voltmeter to a 170V specification, not by any visual criterion. See section 2-5 of the service manual. Thus, except to rescue a monitor that is dimming from old age, it's not a setting that should be "tweaked."

Absolutely Ken. And I do have a VOM and the ability to do so, and have done this adjustment on other CRT's. Usually the control is even "labeled" "screen" ....

[update]: your right -- It should pretty much effect all 3 guns evenly, so - if say, it's perchance the case the current adjustment from the factory isn't correct and If Iwent about adjusting it to 170V, as I believe I mentioned I'd think chances are very good I'd likely need to redefine Greyscale (DRV/CUT) -- with lower or higher values, as well as you mention -- change SBRT value. Perhaps I didn't say that right .... I am likely very wrong about any possible effects on the CBOF/CROF part of it, don't know what I was thinking about, there. Just trying to find some sort of possibilities to look at for possible "answers" for differences we seem to be seeing in these sets.

[end update]

All I was trying to do, is, I was just trying to "get at" what might explain why my set seems "so bright". My purpose was certianly not to suggest the screen voltage control could be adjusted properly based on visual means, or that it should be "tweaked" so to speak.

I don't have the KD34XBR960's Service manual (just the XBR2 manual as posted), and since I didn't know the necessary voltage spec or anything that may be "model specific" about it, that's the main reason I decided not to post anything about actually properly performing the adjustment(or check).

It might be interesting to check it to see what it is set at. Perhaps if I need to take off the back of the set for another reason, I'll check it -- although, that is not as easy as it sounds given where the set is sitting/moving it to a safe place/etc. Even drilling the hole to access the focus pot with a screwdriver+checking/adjusting focus involved a bit of a contortionist act. BTW, Hard to say in an "exact way" - but, on my set if I did make any improvments adjusting the focus control it was extremely minor -- I also couldn't improve on the SM dynamic focus adjustments from the set defaults.

Anyway, if screen voltage IS set properly on my set, it just seems like to me my CRT just must be "especially bright" for some reason, or that there must be another "contrast" control in SM I'm not aware of which could be lowered.

In any case, perhaps needing picture slider in the 25% range actually does make sense for a "calibrated set", anyway ;)

rbc009

01-31-06, 08:21 PM

I have a quick question for you guys. I owned a Panasonic 34" (forget the exact model) but it exploded on me twice. I traded it in for the Sony XS955 34" and just set it up this past week. I'm using DVI->HDMI from my HD receiver.

I find the picture quality is far worst on standard digital and some high-def programming. It seems to be more grainy in high-def and a tad.. blurry.. in standard channels.. (not sure if blurry is the right word..)

How could this be? I thought the Sony would have been much better than the Panasonic. Is it a matter of tweaking or am I doing something wrong?

Someone mentioned I should try tweaking the DRC on the set, but I dont believe the XS955 allows for manual tweaking of DRC.

Any input would be appreciated. Thanks!

Rob

ADU

01-31-06, 08:28 PM

it just seems like to me my CRT just must be "especially bright" for some reason, or that there must be another "contrast" control in SM I'm not aware of which could be lowered.You would bring up "brightness" just as I'm trying to clean-up some of my earlier and perhaps somewhat ill-advised remarks on gamma, wouldn't you, Nitewatchman. ;)

If the picture looks a bit overly bright and lacking in depth though, then gamma might be an issue to look into rather than the gain controls.

Nitewatchman

01-31-06, 08:38 PM

You would bring up "brightness" just as I'm trying to clean-up some of my earlier and perhaps somewhat ill-advised remarks on gamma, wouldn't you, Nitewatchman. ;)

If the picture looks a bit overly bright and lacking in depth though, then gamma might (unfortunately) be the issue to look into rather than the gain controls.

Update: oops think I misread some of your comments a bit at first -- sorry!

Update2 : Oh -- BTW -- even before your current editing, back at the time, I actually found ALL of the discussions/posts from yourself and other posters quite useful in one way or another ...

Note that when I was talking about "brightness", I wasn't referring to what SBRT or "Brightness" control "does", or the gamma settings, or for 480p/720p/1080i, "YGN" -- I was referring to the overall appearance of "brightness" of the entire picture(inclduing from SD/Composite S-video sources) ... as it soley relates to the "picture" slider control in user menu-- I.e. "contrast" related.

My picture looks beautiful with Picture slider at 22~25, brightness slider at 31, SBRT=28. There's nothing washed out about it. From what I can tell --- and it is hard for me to "read" those charts from the "Gamma" charts from AVIA/etc, it's at about gamma ~2.4 with all the Gamma settings in 2170P-4 = 0 -"(pro mode defaults). Which is where I have it set for all pic modes except "movie".

I was only trying to #1)find a way to be able to center the picture slider - NOT that it's important that I do so except just to Have it "centered" at a nice round 31. As I explained, #1) I Can't seem to do that with SPIC, or with SPIO, or by redefining the DRV values - as I'd need less than "0" for instance for the "BDRV" setting, and #2). I thought it would be interesting to Get some sort of an idea why some folks are needing to raise DRV values on their sets to center their picture slider, when I'd have to lower mine beyond what is actually possible, since I can only go down to BDRV=0, and that's not quite "low" enough for Picture=31, even with SPIC=0 for highest "contrast" sources.

ncasebee

01-31-06, 09:24 PM

Well, I decided to look up a service center for my area via Sony's website. I got one result and paid for a delivery of my TV to the SONY AUTHORIZED service center. The Sony KV36HS500 got over there, and there was a 75$ up front fee. I paid it.

The problems were convergence, focus, and geometry. The real problem was blurry/fuzzy at the left and right middle edges. Anyways, the technician there told me the TV was inside specs and that Sony does not build Televisions to be under this type of scrutiny. My Television was inside of specifications and I have to just live with it. I tried to explain that I did make headway with my television via the service menu, but he didn't want to adjust anything for me. He felt that adjustments on this CRT direct view, would never achieve anything constructive. I am receiving the exact same television that I sent over.

It's disheartening. Are the things he told me true? Do I just have to live with the geometry problems and fuzziness at the left and right? I am so very frustrated.

I thought I could just pay for a service, and not have to worry about juggling all these darn service menu settings. There were other settings that would cap out before achieving the best convergence. I hoped that they could fix this. What do I do now? I want my Xbox 360 to look very very good. THE darn text on Kameo is fuzzy on the right and left compared to the middle.

Nitewatchman

01-31-06, 11:25 PM

He felt that adjustments on this CRT direct view, would never achieve anything constructive.

All I can really say is that has not been my experience with Sony KD34XBR960.

I do think some of what he said is true -- but only to a certian degree ... For instance I think it is true "focus" on the screen edges is never going to be as good as in the center .... However, unless he checked it+tried to improve it, I don't see how he could say any "adjustments" wouldn't achieve improved results ....

Nitewatchman

01-31-06, 11:39 PM

I have a quick question for you guys.
Rob

I wish I had a good, quick answer for you :-)

I can quickly tell you that I had somewhat similar "thoughts" when I first turned on the sony set in comparision to other sets I have had - although certianly some of those have certianly had their problems as well, sometimes problems that couldn't be "fixed".

If anything, I've learned about this Sony set it is how *big* of an apparent difference small improvements can be - not allways visable "Right" when you change something, but after you have a chance to sit back and look at it/compare/etc. a bit. But, At the same time, I've also learned, given the complex nature of the SM and the MANY input+scan rate/pic mode specific settings/etc, how complex a task it can sometimes be to "track down" and properly implement and document(which is necessary) the "small improvments" you need to make. At times, quite a bit of time, effort and experimenting have been in order.

With my particular set, although I've made other "adjustments" as well -- probably the biggest improvements resulted from slightly modifying DRV~CUT values for "better" greyscale, using Dynamic convergence adjustments available in SM to solve a problem with convergence issues on far left and right side of screen(moreso left than right) visable on vertical lines in crosshatch pattern -- and -- saving the best for last -- utilizing Ken Tech's findings concerning "image processing" settings - and especially so where his latest findings are concerned.

ncasebee

02-01-06, 09:56 AM

I'm just frustrated about my expierence with this technician. A Sony Authorized Service Center, and they can't even tweak ONE setting for $75. Frustrating. There were also vertical convergence error, but he didn't want to use magnets because he says it's just not worth it. Huh. I wish they could have made my tv at least look somewhat better.

Anyway, it's a combination of focus and convergence error, so I might just try tweaking again. I just don't like doing it very much. The juggling act of settings makes me feel like I'm never getting anywhere.

bigkev4123

02-01-06, 02:36 PM

ADU

02-01-06, 03:14 PM

Update2 : Oh -- BTW -- even before your current editing, back at the time, I actually found ALL of the discussions/posts from yourself and other posters quite useful in one way or another ...That's surprising! ;)

If the gamma settings are at the Pro mode defaults, I think that should be the "darkest" setting.

ADU

02-01-06, 04:35 PM

I was wondering if any of you guys could help me out with something. i have a kd30xs955, and i was messing around with the picture settings, like mpin, pin, ucp, lcp,ppha, lang, vang, vbow, lbow, and for some reason the lines are a little off and I cant seem to get them right. Is there any way I could reset them to the original numbers, or if someone could give me their numbers and I could go from there. I wrote all the original numbers down but i must have threw them away cleaning. up. so any help would be greatly appreciated.kyle,

Per Ken's Post #871 (http://www.avsforum.com/avs-vb/showthread.php?p=7028963&&#post7028963) above, please DO NOT attempt to try to fix this by resetting the SM!! This will ERASE other CRITICAL factory adjustments.

I've got a couple other ideas, but I need to bounce them off a few people with a little more smarts first, hopefully tonight or tomorrow. So just hang loose for a little while, and let's see if we can come up with a better, safer and less destructive solution for situations like this.

Nitewatchman

02-01-06, 05:16 PM

That's surprising! ;)

Well, I think one of the best things about AVSforum is when knowledgable folks discuss their ideas concerning a particular issue, such as "gamma", as well as provide links to other sources/etc ...

Although I do feel it is important to post accurate infomation as much as is possible, and it is also good when that information is corrected by others -- every comment doesn't have to be 100% "right" in order for the reader to beneifit, as I think Some of this stuff is like following an "investigation" - which I think helps the reader "learn", improve their "thought processes" and make good decisions concerning what "to do" about issues they may be experiencing, especially when we don't have a nice big list of descriptions of service codes from Sony .... (The descriptions of listings from the XBR2 manual helps, but) ....

If the gamma settings are at the Pro mode defaults, I think that should be the "darkest" setting.

Yes, P2170P-4 GAMM, and the GAMS~GAMB "lookup" are all at "0" for a "0" GAMM setting, "brightness" slider+test patterns for setting black level work out perfectly with SBRT=28, with "brightness" slider at 31, and with UBOF's between 0~5 (0 only for my DVD player, 3~5 for everything else). But the picture slider needs to be 22~25(which works out just perfectly) or it hurts my eyes after a bit of viewing. (the "brightness" of the "contrast" is the best way I can think to describe it, there's probably a better way to say it). And, from what I can tell using "gamma charts" from AVIA/etc(another place where it helps If I take off my glasses and view it from a certian distance), Gamma appears to be right at about 2.4 with those settings(Including Pic slider at 22~25).

Again, all I'm talking about on this is that I can't find a way to "get" picture slider up to "31" midpoint value and achieve the same results as I'm getting with "Picture"=22~25 -- I don't want to spend much more time on this, as it really isn't necessary, but if anyone has any *suggestions or thoughts about how I could do this, or any thoughts concerning *why* this is occuring I would appreciate it.

* - Note that I tried redefining lower DRV settings(can't go low enough, as BDRV would have to be less than 0), I tried adjusting SPIO(which "sort of worked, but seemed to affect things "differently" than the picture slider, which Is something I didn't want), and as much as possible tried changing SPIC to lower values. Again, Not that it really matters, it doesn't , as "picture" slider at 22~25 works great.

--------------------------------------------------------------

Update to above : In short, looks like the weak link here(my eyes) were "playing tricks" on me ... I discovered this tonight as other family members are complaining about the "dim/dark" picture today .... So, I turn all the lights off and give it another look ... Surprise, surprise -- "Picture=28" appears to now be "the" setting to use ... Which, since previously I had the SPIC values at "4" for the highest contrast sources(which I now have at SPIC="0") makes sense, as I had looked at this closely enough on previous occasions when I'd found "Picture=24" to be "the setting", and that is where it was for a long time with no "complaints" from anyone ...

I expect what happened was something along the lines of : Along with going about improving "balancing contrast" settings among different inputs/etc, my eyes must have gotten "too used" to the dimmer picture with the lower Picture slider setting+in a complete dark room enviornment to the point I thought anything else was "too bright".

Although it might seem to be a simple thing, Doing this sort of thing "by eye"(even with the use of Test patterns for White levels) and getting "proper" contrast/white levels with "picture" or "contrast" controls has allways seemed to me to be one of the most difficult "eyeballing" adjustments. But, when it is set "just right", it shows ..

Now that I have Picture slider closer to 31, Perhaps I might revisit 2170P-4 "SPIO" again or lowering 2170P-1 DRV values a couple of notches to get Picture slider up to 31 ... Or, then again, maybe I won't ;)

KenTech

02-02-06, 01:18 AM

Again, all I'm talking about on this is that I can't find a way to "get" picture slider up to "31" midpoint value and achieve the same results as I'm getting with "Picture"=22~25 <snip> Update to above : In short, looks like the weak link here(my eyes) were "playing tricks" on meAh! Join the club! I have determined that, for myself, watching in a very dark room screws up everything, as though my eyes need a "context" in which to judge a "natural" TV contrast (overall perceived brightness). I set mine so Picture = 31 was at the dim end of my preference range, and I find myself using that, 36, and 41, depending on the material. That said, I find that my eyes compensate well for whatever choice I make -- as long as there is some light in the room. The "background light" test patch on one or both of DVE/AVIA test disks is useful; my illumination is similar, and it seems fine. My perception of color is properly dominated by the TV -- the argument for not having too-bright an ambient illumination.

Nitewatchman, have you tried to balance overall brightness with 2170P-4/SPIC? I found HDTV much too bright, ordinary SD broadcast about right, and my DVD player over HDMI to be impossibly dim. Accordingly, my SPIC settings for those three inputs are 1, 4, and 14, respectively, out of a range of 0-15. Original factory settings: all at 7.

KenTech

02-02-06, 01:31 AM

HELPFUL NEW CHARTS

Motivated by Nitewatchman's complex posts and to save myself from going nuts keeping track of all of the service-mode parameters that affect specific factors, such as black level, color saturation, etc., I have prepared some new charts to organize this stuff. Fill in your own values.

Columns are initially labeled according to Sony's service manual, but sometimes the meaning of the column labels changes a bit from place to place. Particularly ambiguous is the term "ATSC," as it certainly includes QAM cable-digital tuning in some places but not others. Digital SD broadcasts are included in the ATSC column in one chart but under "DSC" in another. For my set, the settings for 2170P-4/SPIC~SHUE that apply to broadcast SD and to my DVD player connected to V5 are inseparable, even though they appear in different columns. But I've shown the charts as they appear in the service manuals.

See attached PDF file of five pages. It prints in *wide* orientation. Pease call my attention to any errors and suggest improvements. Let me know if there are font problems on Windows PCs.

[Edit1: Made very small changes and corrections to charts. Decided to commit to Arial font. Edit2: Eliminated superfluous column for _DRV and _CUT tables; added color-matrix table (RYR~GYB) to section III.]

Nitewatchman

02-02-06, 12:52 PM

Nitewatchman, have you tried to balance overall brightness with 2170P-4/SPIC? I found HDTV much too bright, ordinary SD broadcast about right, and my DVD player over HDMI to be impossibly dim. Accordingly, my SPIC settings for those three inputs are 1, 4, and 14, respectively, out of a range of 0-15. Original factory settings: all at 7.

Yes, I believe I've mentioned doing that in recent posts, including revisiting that very procedure, currently. I'm am doing my best to continue to refine the "balance", which defintely seems more difficult than it sounds like it shoud be.

I experienced similar results with the factory settings - which were SPIC=7 for almost everything - except for "DRC"(SD 480i/etc), which was set at SPIC=4 from the factory. As you might imagine, this only helped make HD "burn out your eyes" "bright" compared to SD. The factory "SCON" settings in 2103-1/2103-1 seem to be right on the money, or close -- "4" for RF, "6" for YC/CV ... although, I can't quite decide if "5" or "6" for SCON is better for CY/YV at this point, or where exactly "SCON" for RF needs to be at -- 4~6 is in range, but even "8" isn't too high for some OTA NTSC stations. Right now, contrast for the 'brightest' OTA NTSC analog station I receive is matched with "SCON=4" to it's digital "HD" counterpart. So, Some analog OTA stations via the tuner are "dim" and some equally match their ATSC/"HD" 720p or 1080i(during upconverted local/syndicated programming or HD programming) counterparts. Oh well.

In my case, it is also my DVD player, (but only at 480p+via component) that's the "dimmest" source, but not quite that "different" from other sources ... I currently have SPIC for most sources at "0"~"1", with exception of 480p via component from DVD player at SPIC=7.

Probably the most difficult one to eyeball for "balancing" contrast among sources for me has been for 480i ATSC. It has allways been "bright" and "garish" and hard to deal with. I have Many stations that are multicasting and sending 480i "subchannels" -- several which are simulcasting their analog programming at 480i SD digital, only one I'm aware of (from talking to one of their engineers) that I know with a fair amount of certianity that the analog+SD digital should more or less "match", more or less. With it's SPIC at 0 also -- Last night, I was also able to make a improvment for 480i ATSC by lowering its UBOF setting by 1 value, and also help "balance out" things "brightness wise" among sources. I do recall when setting black levels among sources that I had a particuarly difficult time with 480i ATSC - The internal QM pattern with the relatively small white box "sort of" helped a bit, but not quite enough.

Also, I know there are other ways of doing it(such as lowering DRV values and raising other SPIC values), but It also turned out I found the slight modification of "CBGN~YGN" values I've been playing around with useful when it comes to the "too bright" HD I was dealing with. For instance ,YGN factory setting was "5", and since I'm currently using SPIC at "0" for say 480i ATSC or component -- With YGN setting at "5", SPIC=0 was still too bright for HD in comparision -- but, YGN at "4" is just perfect with "SPIC=0 or 1 (haven't decided 100% on that one yet)" for 720p/1080i ATSC.

Of course, YGN can't be just adjusted "independantly" on it's own -- If you have CBGN~YGN settings of 5-5-5, to lower YGN to "4", you'd have to lower CBGN~CRGN to 4 as well. Along with that "consideration", After further refining, also concerning "color balance" and striving to achieve the same results with the same RYR~GYB values for all sources - , right now I have CBGN~YGN at 5-5-4, changed from the factory default of 4-5-5. Haven't quite settled on the "exact" values for these yet, along with other adjustments for color(SCOL/SHUE/etc), trying to get the best match for using the same RYR~GYB values for all inputs/ sources+scan rates. I think I have pretty much ruled everything else out for CBGN~YGN other than 4-4-4, 5-4-4 or 5-5-4, so it will likely be one of those. If I needed it a notch "brighter", from what I can tell(I haven't looked at this extremely closely) changing CBGN~YGN to 6-6-5 would only change white balance as well as would be the case going from 4-4-4 to 5-5-5.

Anyhow, On the PICTURE=31 thing -- I did go back last night and look more closely at "SPIO". Lowering it to "7"(It may need to be "6" however) from the factory default of "10", and ended up with "Picture=31" while looking closely at programming material/etc did seem to work this time without causing any I'll effects. However, I also tried lowering SPIO farther and adjusting SPIC upwards as necessary, that may not have turned out so well -- I'm not talking about "big" differences here, but it still looks like to me that adjusting SPIO is for example, effecting gamma to a very slightly different degree than say, the Picture slider or "SPIC".

Update : Oh, I do have a ambient light set up (behind the set) ... but, I had decided not to use it when I was working on this a couple of days ago, in order to hopefully maximize the "differences" I'd be able to see while trying to further improve the balance of "contrast" among sources .... I think it did help in that regard, but, it didn't work so well for trying to "change" Picture slider up to mid-point values, which I was trying to do at the same time to "kill two birds with one stone" in a sense ... Which evidently didn't quite work .... I did turn on the light last night ...

Motivated by Nitewatchman's complex posts and to save myself from going nuts keeping track of all of the service-mode parameters that affect specific factors, such as black level, color saturation, etc., I have prepared some new charts to organize this stuff. Fill in your own values.

Thanks Ken -- that should be very benefical for folks. Along with use of printed out versions of the servicecode chart from the PDF, The method I've been using is to keep track of everything in a Word document, with "scribbled notes" used(and then added/edited into the word document)when actually making changes in SM. that word document includes not only my changed settings+the defaults/etc, but also notes/do it lists/thoughts/etc. It is a LONG document, and I don't know if anyone but me could deciper some of it. - It's so long in fact, when I update it, I usually only print out a seperate, "updated" version, unless a "lot of stuff" changes in the "big" file. At first, it was more of a hassle to deal with, but now, generally I'm not changing a lot of different settings/values, and so it has worked out well for me.

KenTech

02-02-06, 02:10 PM

All I can really say is that has not been my experience with Sony KD34XBR960.

I do think some of what he said is true -- but only to a certian degree ... For instance I think it is true "focus" on the screen edges is never going to be as good as in the center .... However, unless he checked it+tried to improve it, I don't see how he could say any "adjustments" wouldn't achieve improved results ....Nor mine with my 36XS955. I significantly improved overall focus by drilling an access hole to the physical focus control inside the set and redoing everything according to the service manual's procedure. This is what I wrote up in my previous article #13 starting here (http://www.avsforum.com/avs-vb/showthread.php?p=5774400&&#post5774400).

I called out a service tech because I had an overall vertical red fringe on horizontal lines, and I suspected that was a magnet tweak on the neck of the CRT. I was right: the tech fixed it in a minuite while I monitored the results for him.

But when I pointed out the noticable vertical misconvergence in the upper-right corner, a *slight* cyan patch left of center, and curvature of horizontal lines about 1/8 the way down from the top, the tech smiled and explained that those types of defects were theoretically correctable with magnets, but that you could lose your mind trying, as all of the changes interact! You twiddle a magnet in one place to cure the cyan patch, but now the convergence is out in the upper-left. You correct the horizontal curvature at the mid-upper-right, but now the upper-right corner is much too dark and reddish.

I understand, as I have gone nuts tweaking the convergenbce on an old RCA TV monitor with *only* magnets. So much of it interacts that making the correction is more art than procedure.

I have decided for now to take the tech's advice and leave well-enough alone. He wasn't avoiding work on my set; I believe he was being honest.

KenTech

02-02-06, 03:15 PM

I don't know, but actually "brightness" does actually work for me to best describe what we've been talking about when it comes to the "picture" slider or SPIC, rather than using brightness to refer to black level ... But, that's just me, and it would probably be confusing for folks, and it's no wonder some of this stuff is confusing.(I deleted the reply to your post, where I wring my hands about term ambiguity, as it no longer applies.)

Some very knowledgeabe folks who write good online technical articles on this stuff have opined that the three terms should be these:

Black Level = what you think it is. Replaces the idiotic "brightness" term forced on us by the manufacturers.

Brightness = Overall appearance of brightness in the picture as determined by the white-level setup. Corresponds to what you and I perceive but would horrible confuse folks who already have bought into the manufacturers' definition. I would reluctantly accept "contrast," although I don't like this because I am a digital-photo junkie, and this is NOT what contrast means there.

Contrast = mid-tone brightness, as established by the gamma setting. I really don't like this one -- just call it "gamma" ferchrissake! That would be precise.

Proposal: I would like a real "contrast" user control, meaning what it means in photography: a steepening of the response curve in the mid-tones, also known as an "s-curve." This is an era of digital processing of video, and it is a regular feature of pro-grade video cameras. Why not in the display? Some broadcast TV, no matter that gamma is maxed out on the TV, has really insipid mid-tone contrast, making eneryone look pasty. A "contrast" increase would fix this.

Switching to a lower gamma in the TV, an option we can set up for one of the four Picture modes, already compensates nicely for grimly dark TV. ("Lost" and several regular network dramas come to mind.) I did this for myself for "Movie" mode, and I now can't imagine doing without.

KenTech

02-02-06, 03:53 PM

Since the new charts make it easy for me to do this, for those who might be interested, here are the settings that define black level and contrast ("Picture") for my set. This results in normal user settings of Brightness slider = 31 and Picture = 31 to 41.

In 2170P-3: (6500-6700K white point, no pink/green tint, linear grayscale. References: 6500K-calibrated computer monitor, mid-day sunlighted white clouds.)

07-SBRT = 29

08-RDRV = 43
09-GDRV = 27
10-BDRV = 21

11-RCUT = 43
12-GCUT = 21
13-BCUT = 24

15-SBOF = 7 - 7
16-RDOF = 30 - 32 (for User-menu choices Cool - Warm, respectively)
17-GDOF = 31 - 31
18-BDOF = 32 - 30
19-RCOF = 30 - 32
20-GDOF = 31 - 31
21-BCOF = 32 - 30

In 2170P-3, my 13-UBOF settings vary from 3 to 5, balancing black level among the various inputs. E.g. SD broadcast = 4, HD broadcast = 5.

In 2103-1, 02-SCON is set to 6 and 8 for RF and CV/YC, respectively.

In 2170P-4, 01-SPIC is used to balance the contrast ("Picture") among the various inputs. My settings for DRC - V5,V6 - HDMI - MS/ATSC are 4 - 4 - 14 - 1, respectively.

The global adjustments, 04-SPIO ~ 06-SHUO have been left as set at factory, 7 for all.

Nitewatchman

02-02-06, 04:04 PM

(I deleted the reply to your post, where I wring my hands about term ambiguity, as it no longer applies.)

Note: I deleted my post also, to hopefully cut down on the "clutter" in my posts a bit, as the portion of it you quoted seemed more than suffiencent.

Some very knowledgeabe folks who write good online technical articles on this stuff have opined that the three terms should be these:

Agree 100% with your post, the terms they came up with(well except for their "contrast" term), and your take on them.

A "real" contrast control would be nice for that "pasty stuff" - which was causing me difficulty in some cases when trying to find the best values for SPIC, balancing overall brightness(as defined by the "experts") and user menu Picture slider/etc.

As I think I've mentioned, I also have movie mode set up with lower gamma -- Have GAMM=1 set up for GAMR~GAMB=3 for movie mode. Although It does come in handy at times, I must also admit that I rarely use it, and that I actually like the "dark" look of "Lost"(at least the HD version) ....

Nor mine with my 36XS955. I significantly improved overall focus by drilling an access hole to the physical focus control inside the set and redoing everything according to the service manual's procedure. This is what I wrote up in my previous article #13 starting here (http://www.avsforum.com/avs-vb/showthread.php?p=5774400&&#post5774400).

Using your article #13 (precision focusing - I had also checked dynamic focus adjustments in 2170D-4 QPAM~DQP and your article #7 on a previous occasion, but revisted them again when using article #13) made the procedure probably one of the "funnest" things I have done concerning calibration of this set. And although it just turned out to be the case with my set that it wasn't necessary, it's nice now that I don't have to wonder about whether or not it is as good as it can be. It's allways been a question I've had with various sets, and I end up checking the focus on just about all of them, and some of those sets DID need the adjustment.

It just happened to turn out in my case on my XBR that the adjustments pertaining to focus/dynamic focus apparently had already apparently been properly adjusted at the factory.

It is possible however a small improvement may have resulted via my adjustment of the focus pot -- I thought I may have noticed a slight improvement, anyway -- but it's hard to say. Because it was fine as it was, and if it the "knob" did end up being in a little different spot that it was after I adjusted it for best results with the "focus" test patterns/etc, it wasn't much ... say, something along the lines of probably 10:35 o'clock position for one of the "slits" on the knob intead of from the factory 10:45 or 10:30.

What *did* help on my set was adjustment of the D-CONV settings with a cross hatch pattern up, as I had horizontal misconvergence(and it was pretty bad, enough to cause "blurriness" in those areas - which is why It's good I did this before looking at the "focus" adjustments) visable in vertical lines in what would be in the "side bars"(outside of 4x3 "area") on this 16x9 tube - moreso on Left than right, and to a different degree in the corners. The D-CONV settings worked marvousely(spelling, I'm being lazy today sorry) to correct this.

I can detect a slight bit of vertical misconvergence(but only with a cross hatch pattern up) , at extreme top+bottom of screen -- more significant at right top corner perhaps -- but only present on the very "top and bottom" visable cross-hatch pattern lines. It is not effecting the next line "inwards" at all, so can't be effecting much more of the screen than say, 1/2" or less from top or bottom. It's not noticable at all during programming material, and I see no reason whatsoever to "mess with it" so to speak.

Geometry Isn't absolutely perfect, but for the most part I'm very happy with it, as it was from the factory. For the most part, I wasn't able to improve on geometry(straighting out "lines" that are for the most part already straight) with the relevant P2170D-1+D-2 settings.

I did reduce overscan a bit, and "sort of" centered the raster -- There is also a slight bit of a Horizontal Linearity issue - referencing screen center, occuring about in the left and right "middle" of each 1/2 left+right side of screen. It's not noticable -- excepting via "measurements" via comparison of the horizontal width of the "little squares" in a cross hatch pattern. I did notice(and I wasn't expecting this) that adjusting HCNT also improved this H-Linearity issue. HCNT is now at "40" it was "37" from the factory. Actually, I paid more attention to minimizing the H linearity issue than centering the raster when I noticed that HCNT effected H linearity (which is something I didn't quite expect). I didn't check the raster centering precisely by "opening the shutters(HBLK/etc)" yet, but have that on my list. V-linearity seemed to be right on from the factory, if it had been off, or I needed to adjust it because of other changes I make, V linearity of course is quite "adjustable" from within SM.

I haven't messed with any of the MID geometry/overscan related values yet, don't know if I ever will, except if I want to perhaps try to reduce overscan a little more(it's mostly at about 4% currently) -- Mainly, If I ever get around to it, I do want to reduce overscan more for "HD Zoom" mode specifically -- although I adjusted from the set defaults so proper aspect ratio would result for "zoom" or "HD zoom" via setting ASPT=52(the set default of 43 made for example - "squashed" circles), It looks like with "HD Zoom" It's cutting off "more" of the top and bottom(and the sides too for that matter) of the frame(say from a 4x3 upconvert from a station sending 1080i) than it should be, or more than what gets cut off with "zoom" mode from SD sources. This is especially noticable, and mostly an issue for me if letterboxed 16x9 programming is being sent within a 4x3 frame, within the 16x9 1080i or 720p ATSC format actually being sent by the station.

Anyway -- I did have the back off the set at one point just to "peek around", and noticed that it did get the chevron magnet treatment at the factory. There are 6 or 7 I could see placed in spots on the back of the tube.

I don't know how significant(or if they are really significant at all) deflection/geomtery issues concerning the alignment of the set would be if the set were instead, say being used in say, Japan, or if it had been made in asia/etc. In any event, in my case perhaps it is nice to know that my set is facing directly west, and that the PA plant where the set was made is only a bit more than 200 miles away from my location, at probably the same latitiude - such that the Earth's magnetic field probably has about the same effect at the set's manufacturing site as it does at my location.

I think I've covered just about everything somewhere or another that I've encountered or adjusted/etc on my set, I hope some of it is useful to someone in some way.

sanjoseskater

02-02-06, 04:10 PM

Any reason why I would not be able to download the files you have attached in this thread? I keep getting error messages that Internet Explorer cannot open because the site can not be opened or can't be found.
Thanks,
Derek

ptchristensen

02-02-06, 04:13 PM

Jediphish

02-02-06, 05:41 PM

I know this is probably impossible, but is there be a way through the service menu to re-lable the picture settings from "Vivid," "Standard," "Movie" and "Pro?" I'm not referring to the inputs.

Given all the tweaks to SNR, BRT, GAMM, etc., it would be nice to be able to rename the picture settings "HD," "SD," and "DVD," etc.

If not, I guess "Standard" can become the picture setting I use for SD, and "Pro" or "Vivid" could be for HD, while "Movie" could be for DVD.

Nitewatchman

02-02-06, 05:42 PM

Since the new charts make it easy for me to do this, for those who might be interested, here are the settings that define black level and contrast ("Picture") for my set. This results in normal user settings of Brightness slider = 31 and Picture = 31 to 41.

I know I've posted these before, in a more scattered fashion -- just to post them all in one place and perhaps for a bit of fun to "compare" with your settings -- here's those values for my set, for Brightness slider=31 and (hopefuly mostly) Picture slider = 31

2170P-1 - Hopefully ~6500K just via eyeballing and looking at clouds/etc. Again, sorry I can't offer anything more definitive than that. There is No pink/green tint or other color contamination(including checking all inputs with user menu "color" slider at 0, or grey scale patterns from DVE or AVIA/etc.), so it appears to be linear greyscale from what I can tell. Note that the factory "warm" offset seemed awfully close to 6500K, however it does have quite the greenish tint to it.
07-SBRT = 28

"neutral" setting for color temp :

08-RDRV = 32
09-GDRV = 16
10-BDRV = 7
11-RCUT = 32
12-GCUT = 17
13-BCUT = 9

for User-menu choices Cool - Warm, respectively :

Note that I've mostly used "cool" offset for experimental purposes - although it is currently set for a bit cooler color temp currently ..

Note that for "warm" offset, I have values set up so the RDRV~BCUT values are the same as what resulted with the factory defaults for "warm" offset(which were same as service code listing), and in brackets, I've added what those corresponding DRV/CUT values are.

15-SBOF = 7 - 7
16-RDOF = 31 - 31 [32](for User-menu choices Cool - Warm, respectively)
17-GDOF = 32 - 32 [17]
18-BDOF = 33 - 33 [9]
19-RCOF = 31 - 31 [32]
20-GDOF = 31 - 31 [17]
21-BCOF = 31 - 27 [5]
22-DCOL = 0 - 0 (note, set default was "1" for cool offset, "0" for Warm)

In 2170P-3, my 13-UBOF settings vary from 0 to 5, balancing black level among the various inputs. For my DVD player(480i or 480p component) UBOF=0, everything else ranges between UBOF=3~5. HD ATSC broadcast(or Memory stick) at UBOF=4, RF(for analog NTSC OTA broadcast mainly) is at UBOF=5, CY/YV is at UBOF=3 - for Dish network 311 receiver and JVC S-VHS Deck via S-Video connections.

In 2103-1, 02-SCON is set to 4 and 6 for RF and CV/YC(the factory defaults), respectively.

In 2170P-4, 01-SPIC is used to balance the contrast ("Picture") among the various inputs. note that On my set, I've noticed can't set a seperate value for 480i via V5/V6, it uses "DRC" value.". My settings for DRC - V5,V6 - HDMI - MS/ATSC are 0 - 6 - (HDMI not used) - 0 or 1(haven't decided for sure on that one yet), respectively.

As YGN this effects "white balance" for 480p/720p/1080i (via V5/V6 or internal ATSC tuner) as well Note that CBGN~YGN factory defaults for "component" were 4-5-5, my current adjustment is 5-5-4. I would need lower "SPIC" values for 1080i/720p(and they are at 0 currently), or lower drv settings/lower "SPIO" values with YGN at "5".

The global adjustments, 04-SPIO ~ 06-SHUO had factory defaults of 10-6-7, my new values(only changed SPIO) are 7-6-7.

--------------------------------------

It could never happen, but I was thinking how interesting it might be if we could do a "side by side" comparision of our sets Picture quality, using the same sources(of course, for example a cable STB's output might of course vary and require different settings from what I need for say, YV/VC for my Dish network receiver, or JVC S-VHS deck).

It would be like trying to guess how many sheep I can count jumping over the moon before I fall asleep tonight, and while I have no doubt there might be some differences -- I'd almost be willing to bet the results would be more similar than the differences among our sets for the above numbers might seem to suggest.

KenTech

02-02-06, 05:44 PM

Made small changes to the GS&ColorChart.pdf file, eliminating superfluous column in _DRV and _CUT tables and adding color-matrix table (RYR~GYB) to section III. See original post (http://www.avsforum.com/avs-vb/showthread.php?p=7046048&&#post7046048) to download current file version 04.

These files are openable by Adobe Reader, a free application from adobe.com. I have no idea why Explorer insists on trying to open it in Windows, as it clearly has a ".pdf" suffix. Another choice: Download it to a specific location as a ".pdf" file, then open that.

KenTech

02-02-06, 05:52 PM

Note that for "warm" offset, I have values set up so the RDRV~BCUT values are the same as what resulted with the factory defaults for "warm" offset(which were same as service code listing)I can't imagine where Sony got these values, as they violate common sense. If Warm using approximately these settings were tweaked to be correct, then why would you ever switch to Neutral, as it would be way too blue? Just keep in mind that these are *offsets,* not absolute _DRV and _CUT values. Setting Warm or Cool to 31-31-31 has *no* effect.

Like you, I "take over" Cool for experimentation so I can try out changes and live with them for a while before committing them to the Neutral _DRV and _CUT parameters. Not that there's any potential problem, as I log all of my previous settings. But one can easily switch between "old" and "new" settings that way.

Nitewatchman

02-02-06, 05:55 PM

If not, I guess "Standard" can become the picture setting I use for SD, and "Pro" or "Vivid" could be for HD, while "Movie" could be for DVD.

Perhaps I might be missing something, but, unless you want to use slightly different RYR~GYB settings for say HD vs SD for some reason -- for the most part I don't know why you'd really need different Pic modes for "SD or HD"/etc, as for example, The 2170P-3 settings allow you to set different image processing(and Color/Hue or black level offsets) which are BOTH scan rate and input specific. In fact, they are already set from the factory "differently" for different scan rates(HD or SD/etc), even via the same input. For instance, there are different SYSM settings available for 720p, 1080i and 480p from the internal ATSC or QAM tuner, and there are different settings available for SD from the internal NTSC/cable ready tuner. You don't have to "switch" between them, the change in "settings" happens "automatically" when with the internal tuner you say, switch from analog channel 7 to a digital station on virtual channel 7.1 sending HD. For more info,

See Ken Tech's article earlier in this thread on customizing picture modes, as well as his articles/posts on the image processing settings.

I don't know if you can "rename" those or not, I've never looked into that. But, that is one of the reasons, I have set up "standard" as my "main" viewing mode for everything. Actually, I rarely use anything else, except for my lower gamma settings for "movie" pic mode, or for expermimenting with various image processing settings with "vivid", or making comparisions to "pro" which I've left at it's defaults.

KenTech

02-02-06, 06:06 PM

It could never happen, but I was thinking how interesting it might be if we could do a "side by side" comparision of our sets Picture quality, using the same sources(of course, for example a cable STB's output might of course vary and require different settings from what I need for say, YV/VC for my Dish network receiver, or JVC S-VHS deck). What? You don't want to do a Giant Schlep between Oregon and Ohio? (I hail from Youngstown, BTW, thru high school, then on to Pittsburgh and CMU.)

What one *can* do for show-and-tell, however, is digitally photograph the screen -- or, better yet, parts of it so the picture files don't get huge. It won't prove anything regarding color, grayscale, etc., but for image quality, arguments about "edginess" and sharpening -- yes, it works. I keep thinking I will do this the next time I see someone complain about the quality of SD-analog broadcast on their set! (The DA-4-based sets are unparalleled at SD display, since they only minimally upsample 480i for 960i display. Garbage in, garbage out, however!) Trouble is, you have to use a freeze-frame, as most camera exposures are a bit long for anything moving.

An example for anyone willing to try: Display the attached freeze-frame from Monsters, Inc DVD (note the time in the filename), and take a picture. It's one of the best examples I know of minimally-enhanced video at the limits of DVD quality. View at 300-400% to see the gory details. (The picture is a pixel-accurate computer-grab from a DVD freeze-frame, *not* a photograph.)

Nitewatchman

02-02-06, 06:08 PM

Just keep in mind that they are *offsets,* not absolute _DRV and _CUT values. Setting Warm or Cool to 31-31-31 has *no* effect.

Absolutely. The warm offset values I provided, as well as the DRV_CUT values the "correspond to" should, at least in part demonstrate this.

Perhaps I didn't explain myself fully enough. To "get" the DRV~CUT values that "corresponded" to the factory "warm offset" values, I looked at what the offsets "were" in comparision to what the factory "neutral" values were. For instance, Factory BDRV value was "24". Factory value for BDOF for "warm Offset" was "16". So, looking at the number you need to subtract(15) from "31" to get "16" for the offset, means that the factory BDOF offset of 16 means the warm offset from the factory (BDRV = 24 - 15) was using "9" for it's "actual" "BDRV" value.

So, I figured that out(and double checked it numerous times) and wrote down the actual RDRV~BCUT VAlues that would produce the SAME results as the factory "warm" offset if I actually used them with RDRV~BCUT and "neutral". Then, given my new "neutral" values(actually what I'm using for RDRV~BCUT) I redefine new offsets for the "warm offsets (+ or - from 31 as necessary) so that the "warm offset" still gives me the exact same results as was the case with the factory "warm offset".

I know that seems like a lot of work, but wanted to be able to check the same color temp that resulted factory warm offset values for reference, and I just prefer to use "neutral" and the actual RDRV~BUT settings with my improved settings.

Nitewatchman

02-02-06, 06:16 PM

What one *can* do for show-and-tell, however, is digitally photograph the screen
.)

Yes, I do that often for various reasons in my local threads, and for my TV/DTV Dx'ing Hobby. You'll find some(but mostly from other sets) posted from me In Cincinnati or Dayton threads.

It won't prove anything regarding color, grayscale, etc., but for image quality, arguments about "edginess" and sharpening -- yes, it works.

And, since I'm using the same settings there(your lastest image processing "Findings"), including If I recall correctly from one of your posts even the same value for the user menu "sharpness settings" -- I don't think we'd see any difference in the image quality :)

The user's PC display might be something to think about though.

The DA-4-based sets are unparalleled at SD display, since they only minimally upsample 480i for 960i display.

I agree. Unlike many HD displays(not that some of the others are all that bad -- the farojuda chipset in my Toshiba 34HF84 does an awfully good job of it as well - just not quite as good as the XBR), SD (even from analog OTA sources) looks fantastic on this set.

KenTech

02-02-06, 06:18 PM

Absolutely. The warm offset values I provided, as well as the DRV_CUT values the "correspond to" should, at least in part demonstrate this.

Perhaps I didn't explain myself fully enough.No, I got it. My comments were not aimed specifically at you, but were to clarify for some readers who, not being engineers in their DNA, don't relate intuitively to the whole "offset" notion, especially when "zero effect" isn't zero but 31. For those who get it: I'm trying to walk that line between snowing some readers and talking down to those who already understand. Hope there's not too much repitition.

KenTech

02-02-06, 06:27 PM

The user's PC display might be something to think about though.For black level, contrast, and especially gamma, you're right. But it's fine for seeing the actual image on the DVD. Here (http://www.avsforum.com/avs-vb/showthread.php?p=6934240&&#post6934240) is a method that can be used on a Mac or PC with free software to get pixel-accurate 1:1 frame grabs for any DVD you can play on your computer. There are a few more frame grabs posted there and in the next few messages, too.

Nitewatchman

02-02-06, 06:57 PM

Hope there's not too much repitition.

Not at all. I also didn't mean to "moan and complain" that you were pointing something out I already knew -- but unfortunetly via this "medium" I think it's sometimes impossible to know "exactly" what is meant by any given comment or point(not that it is allways necessary to - as people come here I think mostly for the information that will help them), and it is also difficult (and really unnecessary) sometimes to "word" your posts so that the actual "intent" of the comment is known. If THAT makes any sense.

In any event -- I think add'l explanations/clarifcations/etc are useful in some cases - along the lines of what I was discussing with ADU concerning his earlier posts about "gamma" -- for instance Perhaps hopefully my additional clarification might also help others who are looking at utilizing their Cool/Warm offsets in similar ways. If I recall correctly, a post similar to mine helped me understand it back at the time, and saved me time trying to figure out what was going on.

---------------------------------------------------------------------------

digimiX

02-02-06, 07:02 PM

Nitewatchman

02-02-06, 07:16 PM

For black level, contrast, and especially gamma, you're right. But it's fine for seeing the actual image on the DVD.

I should also mention -- One issue I have run into when photographing a CRT screen, is "focusing the camera" issues and visable "moire pattern". That's as true for a "digital" camera as it is a film camera.(Yes, I'm also into photography+photo editing/etc - just not so much lately). In fact I have, and sometimes still even use a 35mm film(negative) scanner).

You don't see it with your eyes, but if you focus the camera for "sharpest" results, or use the "auto focus", I've found moire pattern will usually be present, and that has been the case for every CRT display I've ever photographed. Placing the camera a certian distance from screen can help, but how I usually handle this is slightly defocus before making the exposure - although, with my digital camera the little LCD display isn't the best for this. Generally, I use relatively low F-stop and shutter speeds of ~1/15~1/60 per second with settings on my digital camera which supposedly approximate ASA200. As for the shutter speed/etc, if I under or overexpose something/etc, I'll fix it by raising/lowering Gamma/etc. with photo editing software, since it isn't generally a priority to be exact, I just change it so it "sort of" looks good.

FWIW, attached is a screenshot I took of a portion of KD34XBR960 screen while in one of the "QM" "info" screens for ATSC information(PSIP info from the transport stream/etc), in order to demonstrate the station wasn't sending some proper info concerning some PSIP tables. The abbreviations "light up" for those tables if the receiver can properly recognize them/etc. In this case, it's a 1080i signal, and 1080i "full" mode being used.

I took this screenshot several months ago, when I was using your "old" image processing recommendations. I believe I was experimenting with greyscale also at the time, and don't recall what "other" sorts of settings I was using. This is very much a "close up" shot of screen, and is a part of a higher resolution image which I didn't resample/resize or use unsharpening or sharpening filters on -- which is something I often do for screenshots of CRT for other purposes(such as involving my TV Dx'ing hobby).

As you can see, there is a slight difference in focus on what is the far left hand edge of screen. I can't really see any of that to as much of a degree with my "eyes", even this close to screen, presently, so perhaps, unless it has to do with having the lens so close to screen/etc.(I just didn't take the time or effort to pay any attention to that) -- I might have also been experimenting with the dynamic focus adjustments a bit at the time. Also, note I wasn't really trying to get a "good shot" other than than to get a clear shot, just so the text was readable.

Update: oops! Just realized the attached file is not a very good example .. as Looking at that again, looks like I must've used a little too high JPG compression ratio when I took that image. Don't know when/if I'll ever get around to it, but I'll put it in my notes to try to get a better screenshot from something such as the captures from Monsters, INC Ken posted(can't do it from DVD, though), and I'll use the camera's "raw" format to create a uncompressed tif and be very careful about the compression ratio/etc.

ptchristensen

02-02-06, 08:15 PM

i recently bought a KD-34XBR960 and need to adjust the picture size/placement. all sources are cut off a bit...about 5% or so. isnt this overscan?? i cannot adjust in the normal tv settings. anyone know how to do this in the service menu?

If the picture looks great and its an even amount of overscan on all sides...leave it alone. You never know whats behind this overscan: bending lines in all directions, corner spots and other nasty stuff.

Nitewatchman

02-02-06, 10:57 PM

HELPFUL NEW CHARTS
I have prepared some new charts to organize this stuff. Fill in your own values.

Sorry to post twice on this but I just wanted to thank Ken again for putting the work into this and posting it, and for providing/adding the excellent descriptions. It should be very useful for folks, I've printed out the latest update and already found it useful especially as it is much less "confusing" to look at to see all the relevant parameters for "Black Level", "Contrast", "Color" All on one page in an organized way, along with the different columns for inputs/etc.

KenTech

02-03-06, 12:53 AM

I just did a little tinkering this evening with CXA2171/CBGN~YGN. I put up a completely balanced-looking DVD still frame from my V5-480p input. (I prefer 480i for DVD, but that doesn't go thru the CXA2171 chip.) I watched the picture while I tweaked the values of CBGN and CRGN to see what the "signature" of maladjustment was.

Other "misadjustments" are well knows: Wrong Hue setting = greenish or pinkish skin tones. Color amount wrong = garish or insipid colors across the spectrum. Red push = colors balanced, except for red, which stands out as exaggerated.

But why did Sony "unbalance" the factory settings for the two components Cb and Cr in the service-manual chart -- and apparently also for Nitewatchman's 34XBR960? Mine were set alike: CBGN - CRGN - YGN at 4 - 4 - 4 or 3 - 3 - 3, in different columns.

The effect of unbalancing the first two is interesting but in accord with what Cb and Cr signify: the degree of color mapped along either a blue-yellow axis or along a perpendicular red-green axis, independent of luminance. If you crank CBGN above its "partner" CRGN, there is considerable blue and yellow "push" added to the picture -- not one color or the other, but both. Same with CRGN, except the colors are cool-red and green that are exaggerated. Grays are not affected. If both are increased the same amount, all colors are increased together -- as though one has simply increased saturation.

Those two color-axis parameters are together paired with a Y parameter, which is the luminance of the substance of the picture: Increase it by cranking up YGN, and the picture gets brighter but less saturated; lower it and it gets dimmer but the colors get more intense. Raise and lower all three together, and you just increase and decrease the picture's contrast without affecting color.

I guess that if there is a characteristic in the color-processing chain that must be compensated, then Sony might go for the CBGN/CRGN unbalance. Or if one of the processing chips has a lower output for one of those color components. But my 36XS955 came with them at 4 - 4 - 4 out of the box. Go figure.

I note that the color-coding of the cells for these values in Sony's chart indicates that some are set "at F[actory] A[djustment]" and others "at C[ircuit] B[oard] A[djustment]." (My interpretation.) Perhaps if one of the circuit boards is replaced, service personnel have to attach an oscilloscope and use these settings to "trim" the color-component waveforms to the correct amplitudes.

If I see color aberrations that push blue *and* yellow or red *and* green, I'll now think about trying a correction in CXA2171 -- for 480p, HD video, or HDMI, that is. Anything that goes thru 480i DSC processing doesn't go anywhere near this chip.

Nitewatchman

02-03-06, 06:51 PM

first -- Another quick note on the new charts Ken posted yesterday -- they also helped "remind me" that although I had made sure I'd set up everthing properly to "customize" the picture modes for all the inputs/sources/etc, I was using at the time .. When I added "MS" recently, I had missed changing the value to "0" for the seperate column for "ABLT" for "MS" -- and finally realized this last night when I first glanced over the charts. Don't know why I didn't notice the slight change when switching between "Pro" and other pic modes with MS, but I thought there was something that was a little "off" ... now I know what it was ... Thanks again Ken ...

I just did a little tinkering this evening with CXA2171/CBGN~YGN.

Thanks for giving those a little more of a look. Your findings on this match my observations on this to a T. And, you have done a much better job explaining it than I have explained my observations and tinkering with this.

But why did Sony "unbalance" the factory settings for the two components Cb and Cr in the service-manual chart -- and apparently also for Nitewatchman's 34XBR960? Mine were set alike: CBGN - CRGN - YGN at 4 - 4 - 4 or 3 - 3 - 3, in different columns.

Unless someone tinkered with it(or the set was serviced) between the factory and the time I got the set(and for various reasons described below, that may be a possibility), the factory settings for CBGN~YGN on my set were 4-5-5.

I don't think I have seen much possible sign of any such tinkering or "add'l" servicing being a possibility elsewhere, although its hard to say. For instance, all 2170P4 SHUE/SCOL set defauts match the servicecode listing from the service manual, as do most other values you'd expect. I don't know, but, the "factory" SPIO setting of "10" along with most "SPIC=7" - (the big exception being DRC SPIC=4), might perhaps seem a bit unusual.

[update] Almost forgot : Other possible unusual "clues" that perhaps might suggest there was some "tinkering" or servicing going on with my set between the factory and the time I received the set : #

1). The "factory" default on this set for CBOF/CROF = 31/31 for 720p - although all other YOF/CBOF/CROF factory defaults match servicecode listing. So does perhaps(as I just happened to notice) the CBOP/CROP "factory" default of 9/36(36/37 in Servicecode listing).

2). Also, unusual perhaps especially as noone else has reported seeing these issues : a). the unusual, "sporadic" incidences I'm occasionally getting when only working in SM concerning values for specific columns of either 2170D-4 ASPW/ASPT or ZOOM occasionally "switching" on their own to values in the "next column"(and sticking) while I'm switching channels or screen modes at times when I'm working in SM(never outside SM). b). As well as a couple of odd things which have occured in SM concerning signals from internal NTSC or ATSC tuners. More detail avaialable concerning what I'm experiencing with these issues (a+b) as I reported in a more accurate and detailed fashion in Post #876 (http://www.avsforum.com/avs-vb/showthread.php?p=7034710&&#post7034710) [end update].

It was "supposed" to be a new set -- but, nevertheless I suppose it is possible it wasn't entirely "new". My KD34XBR960 is a warranty replacement for another set. On that other set, long story, but essentially a bad diode in an aux power supply led to a 3 month wait and a set replacement rather than a repair. I know, it sounds crazy, and there is more to it than that, but that is the only info on it I was able to verify as 100% accurate.

I do know we took delivery of the Sony set in July 05, date of manufacture on the sticker says Jan 05. Also, I did notice when at the store "picking out" which set to replace my RCA F38310 with(of course, I had already decided there was only one choice, given the choices I had), they had no fewer than 5 open box KD34XBR960's sitting on the shelf. I also noticed the owner's manual said I was supposed to get some sort of "welcome" screen when I turned the set on for the first time, but I didn't get that.

I guess that if there is a characteristic in the color-processing chain that must be compensated, then Sony might go for the CBGN/CRGN unbalance. Or if one of the processing chips has a lower output for one of those color components. But my 36XS955 came with them at 4 - 4 - 4 out of the box. Go figure.

I note that the color-coding of the cells for these values in Sony's chart indicates that some are set "at F[actory] A[djustment]" and others "at C[ircuit] B[oard] A[djustment]." (My interpretation.) Perhaps if one of the circuit boards is replaced, service personnel have to attach an oscilloscope and use these settings to "trim" the color-component waveforms to the correct amplitudes.

For whatever "reasons", on my set there seems to be slight mismatchs for "color" between #1). those scan rates effected by CBGN~YGN/CXA2171 chip(480p/720p/1080i) and 2.) everything else(DRC). As well as a very slight mismatch between a)480p and b)720p/1080i. And, for whatever reasons, My best guesses are either #1). is that whoever(factory or someone else? I don't know with 100% certianity see above) made any "CBA" or "FA" adustments managed to perhaps not find quite the "best" settings for their adjustments, or for CBGN~YGN(if they even adjusted those at all) to best compensate for these slight mismatches I'm getting or #2). This set is on and used by various family members on average about 12 hours a day. It just doesn't make much sense to turn it off much of the time, as someone will just come in 5 minutes later and turn it back on. Perhaps it's possible some "drifiting" has already occured, although I'm fairly sure this issue has actually been there all along.

Follows is just some of the evidence I have for this -- -- note that in one of my previous posts on this which might suggest somewhat different results -- at that time I was also "experimenting" with 2170P-4 SLO - In some cases changing it to "7", from the "factory" default of "6". This time, I left SCLO at the default "6" for everything.

#1). Using the Blue+white QM PATN 6, 46 and 66 for 1080i, 480p and 720p respectively, and turning on blue only gun via RGBS = 1 -- With the "factory" defaults for CBGN~YGN -- While I found I could adjust SCOL+SHUE for the "most even" blue(to where I can't even see a "boundry" between top+bottom 1/2 of pattern) --- For 720p AND 1080i ... I need different SCOL/SHUE(Or UCOF/UHOF) settings in order to get the same results from 480p ATSC(QM PATN=46). I think this may partly explain why finding out exactly how the "MTRX" control in SM worked and how to use it was important for me.

#2). Using the "factory" defaults for "CBGN~YGN" and adjusting SCOL/SHUE for best results with QM PATN= 6 (1080i), or PATN=66(720p) -- I do need to set SHUE=32(default 30), and SCOL=33(default 32) for 720p/1080i ATSC for best results. However, that change in SHUE from default doesn't really "work" well for Red gun only, or Green gun only tests/adjustments - I can get it to "work" to somewhat of a degree, but I have to use quite different, and unusual values for RYR~GYB than I need for everything else( in my case 480i DRC, basically) - which is 13-15-5-3. And again, the results for 480p(QM PATN #46) are somewhat different.

#3). I've played around with trying all the slightly different values for CBGN~CRGN I could think of to try, to see what works best to match 480p with 720p/1080i with the QM PATN x6 series of patterns, while adjusting SCOL/SHUE as necessary as well as trying to find the values which will allow me to use the same, or nearly the same values I need for RYR~GYB for 480i/DRC.

There is nothing I've found that "works" perfectly -- However -- I think the best I can find is CBGN~YGN at "4-4-4" -- It's still not 100% perfect solution, and CBGN~YGN "5-5-4" might be the better 'compromise" -- I'm not sure at this point. With CBGN~YGN at 4-4-4 - 480p and 720p/1080i with internal QM PATN #6 "series" match the closest, with "SCOL=34" for ATSC 480p/720p/1080i - best with 720p+1080i -- This also turns out to be the best "match"(more or less) I can find for 480p via component, but with SCOL=33. Note that SCOL=33 works best for 480p ATSC as well, but I can't set it differently - I can of course set SCOL=32 for 480p/720p/1080i ATSC, and then raising 2170P3 UCOF values for 720p and 1080i - before messing with anything like that though I want to find the permanant values for CGBN~YGN I'm going to use). No change in SHUE from "30" default is required for any scan rate.

Also note that concerning all the slightly different CBGN~YGN values I've tried, the factory defaults appear to be about the worst choice of what to use out of all of them concerning these slight "mismatches".

With CBGN~YGN at 4-4-4, The "best" settings for RYR~GYB for 720p/1080i ATSC are still not the same as the 13-15-5-3 I need for 480i/DRC -- Instead, after a little more "fine tuning" and checking last night -- the best settings for RYR~GYB for 720p/1080i using CBGN~YGN at 4-4-4 turn out to be "14-15-6-4" (and I was about to detirmine this time that is slightly noticably better than 14-14-6-4) -- Which Is perhaps interesting that Ken had posted he had came up with those same values for his set for 1080i with a recent recalibration using the QM patterns/etc. But, on my set for 480P component V5/V6, it turns out the best settings for RYR~GYB are the same as everything else -- "13-15-5-3". This provides for pretty much "perfect" color balance for all scan rates involved(only with 2 slightly different RYR~GYB settings - luckily which I can choose from user menu per input and pic mode), and I can't seem to find any "better" settings - not only concerning CBGN~YGN, but also the P21704 SHUE/SCOL offsets/etc.

#4). What I thought was also interesting was -- When I was experimenting a bit with setting "SCLO=7"(I know it says set it with a scope, but for one thing, I don't have one currently) several days ago instead of the "factory" default of "6" -- In which case, I found CGBN~YGN setting of 5-4-4 (servicecode listing default BTW) worked perfectly with 720p/1080i WITH RYR~GYB at 13-15-5-3. However, that wasn't the case for 480p -- including 480p via component. I also noticed raising SCLO to 7 also made the issue mentioned in #5 below involving S-Video input to become "worse". So, even though at first glance it seemed SCLO was a global setting that would have the same effect as turning up the "Color" slider, or adjusting 2170P4 SCOL, or 2170P-3 UCOL offsets -- that didn't seem to quite turn out to be the case.

#5). As you say, and I also have observed and reported-- 480i/DRC or NTSC isn't effected by these settings or the CXA2171 chip -- but -- neveertheless, I do wonder if perhaps the following issue might still in some way be "Related", but only in the sense that on my particular set I've also found a slight bit of a color balance "mismatch" between inputs in one other place, which doesn't involve CXA2171 CBGN~YGN :

Even though Color tests show "right on" results for 480i component via DVD(or 480i ATSC via internal test patterns), for All three guns used indivdually with :

#1). "SCOL=33" (factory default was SCOL=32 for everything, SHUE=30 for everything - again note that I have left "SCLO"=6, although I did experiement a bit with "SCLO=7") ---

#2). RYR~GYB 13-15-5-3 ...

Via CV/YC (S-video) (AVIA or DVE color tests played from DVD) -- I'm getting excellent results with these settings(and centered Hue/Color sliders) and 2103-1 SCOL=8, SHUE=10 -- with "blue gun", and "Red Gun" -- BUT, with "green" gun, for best results Hue slider needs to be set to "G1" - (or 2103-1 SHUE adjustment at SHUE=11/etc) Trying to change RYR~GYB instead doesn't help, GYR~GYB at 5-3 are still the best values.

Again, since I'm not getting this from 480i component or internal ATSC pattern at 480i, I don't know what is going on with this. Could be something with the DVD player via S-video, or perhaps the s-video input itself and any "CBA" adjustment/etc -- Although, I'm not getting this with this DVD player via s-video vs. component hooked up to a another set - so that would seem to take the DVE player out of the "mix".

If I see color aberrations that push blue *and* yellow or red *and* green, I'll now think about trying a correction in CXA2171 -- for 480p, HD video, or HDMI, that is. Anything that goes thru 480i DSC processing doesn't go anywhere near this chip.

In my case, I allways knew something looked a bit "wrong" about the color with 720p/1080i. I noticed it most being an issue with "green" - seemed to be too much of it, or "unnatural" looking Green. But, it wasn't really apparent enough to be able to "quantify it" as Red+Green push(or probably more like "not enough" Blue+Yellow), as opposed to perhaps issues with the broadcast material/etc. But, I'm quite sure now that's exactly what was happening.

After changing CBGN~YGN from 4-5-5 to 4-4-4 (or 5-5-5 - I haven't decided on that yet or for that matter SPIC=0 or 1 for 720p/1080i ATSC) I think I can be reasonably sure just a "tinge" of Red+green push is what I had been getting with the "factory" CBGN~YGN settings of 4-5-5, as, in addition to what I've seen with the color bars/etc, --- those "unnatural" greens(it also effected flesh tones slightly) are now gone.

If there is anything going on now with CBGN~YGN at "4-4-4", and "SCOL=34" for 720p/1080i ATSC -- it may be a slight "electric blue" "push" you'd mentioned you had an issue with ... . I'm not sure yet, and hopefully I was just seeing things --- There was a lot of "blue" in this week's "Smart Gardening" episode and it looked just wonderful -- but Last night the fellow's blue shirt on the PBS HD Mexico cuisine show was awfully "electric blue" looking ... Even though of course in my case my QM #6 pattern with blue gun is perfectly even across the screen with the current settings, and tests with Red or green only are showing excellent results as well. Hopefully, it was just his shirt and the lighting involved ....

If not, perhaps it may be turn out to be the case that "5-5-4" is my better "compromise" setting for CBGN~YGN - In which case It works best to lower "SCOL" for 720p/1080i to 32 (note that SCOL=32 also works best for 480p in this case) for best results, and use RYR~GYB=13-14-6-4 .. which in this case, also works best for 480p - unlike the differences in RYR~GYB needed between 480p and 720p/1080i with CBGN~YGN at 4-4-4.

I've tried using both "5-5-4" or "4-4-4" "options" and the differing required SCOL settings for each for CBGN~YGN now(but only for a few hours each), and at this point I'm not sure which offers the better "compromise". I think it's "4-4-4" but may require a little more (and longer) closer observation of actual programming. I know in this case I'm just changing the color saturation(gain) by lowering or raising both CBGN+CRGN with the sam values - but the question is I think - Is it a tad " more" or a tad "less" less saturation that is the best "compromise" here --- 2 "ticks" up for SCOL, or one "tick" up for one value for Cb/Cr.

So, In any case and in conclusion, I've been able to make improvement, but it seems there are still some slight color "mismatches" going on between inputs/scan rates that it seems I have no way of completely "fixing" with SM (or user menu for that matter) adjustments. It is slight enough that at least so far with current settings(with perhaps exception of the "electric blue thing at 720p/1080i), It's not noticable when viewing programming. I may be able to find a slightly better "compromise", and I'll continue to look, but it is difficult to do as It the mismatches are not very significant.

------------------------

2/4 Update : Last night, went through a couple of my different options again, and have finally decided CBGN~YGN = 4-4-4 turns out to be best for my set.

One perhaps interesting thing I noted was that in SM and using RGBS 2/4 and the internal QM PATN #4/6, #24/26, #44/46, #64/66 -- as I mentioned earlier RYR~GYB at 14-15-6-4 is best for 720p/1080i, but 13-15-5-3 (confirmed with AVIA/DVE via DVD) is best for 480p, just as it is for everything else (480i component, 480i ATSC/i.e. DRC). However -- I left the QM patterns up for 1080i and went outside of SM, and noticed that with the DVE green or Red color filters, RYR~GYB=13-15-5-3 appeared to be the better choice -- Now, I know there can be errors in these filters(and that seemed easy to spot for the DVE Blue filter), but what was odd was that using RYR-GYB at 13-15-5-3 outside of SM with the DVE Red+blue filters perfectly matched my results with RYR~GYB 14-15-6-4 + RGBS 2/4 inside of SM. Probably just a "coincidence", but it just seemed odd and thought I should mention it ...

KenTech

02-04-06, 04:44 PM

RYR~GYB at 14-15-6-4 is best for 720p/1080i, but 13-15-5-3 (confirmed with AVIA/DVE via DVD) is best for 480p, just as it is for everything else (480i component, 480i ATSC/i.e. DRC)My results are similar. East to determine whether it matters. Just set up one for Default and the other for Monitor color axes. Now you can observe a good subject pucture and switch between them.

I fond the difference is a tiny change in green intensity -- totally insignificant. SInce broadcast TV is poorly color controlled anyway, I'll happily "sacrifice" that for accurate DVDs. So 14-15-6-4 is my Default. I have almost no incentive to ever switch.

If skin tone had changed noticably, I would have an issue. But a tiny green intensity change? Hardly!

Ken

Nitewatchman

02-04-06, 05:28 PM

My results are similar.

That's good to hear. Not unexpected, but now I know besides the "factory" CBGN~YGN setting on my set, the very slight color mismatches I'm getting on my set are not unusual.

Just set up one for Default and the other for Monitor color axes. Now you can observe a good subject pucture and switch between them.

I fond the difference is a tiny change in green intensity totally insignificant.

Same here. I've also already checked broadcast sources(even HD) with both "default" and "monitor" choices I have set up, and I really, really have to look closely to see any differences, so either choice will suffice in any circumstance, really.

SInce broadcast TV is poorly color controlled anyway, I'll happily "sacrifice" that for accurate DVDs. So 14-15-6-4 is my Default. I have almost no incentive to ever switch.

If skin tone had changed noticably, I would have an issue. But a tiny green intensity change? Hardly!

Ken

Yes, what I did is set up RYR~GYB "13-15-5-3" for the "default" color axis choice avialable from user menu, and RYR~GYB "14-15-6-4" for "monitor". Only really because those are the very slightly "different" RYR~GYB settings that worked out for me, and because in the 34XBR960 for "default" color axis choice it's labled in the user menu as "emphasize red colors", and the "monitor" color axis choice is labeled as "de-emphacize" red colors - which just seems to make more sense in my particular circumstance.

On 34XBR960 with the "mode memory" function in user menu set to "on", I can use different "color axis"(either default or monitor) per pic mode, and per "input". So, since in my case, "13-15-5-3" is "most accurate" for everything (including DVD at 480i/DRC or 480p) except 720p/1080i HD -- I am using "default" color axis choice for all inputs except from internal tuner and MS, which I have set to "monitor", currently.

If there is anything that might cause me to change my setting for broadcast sources(from internal tuner), from previous experience with this sort of thing, it might be for the orange portions of Cincinnati "bengals" graphics or uniform as received from the "poorly controlled color" from some broadcast sources - mainly that of course involves SD or upconverted local/syndicated SD programming for various reasons, not HD from the network feeds.

I just find it extremely annoying when the "orange" portion of those graphics turns out being more of a weird "sort of" pinkish+reddish+Magnenta color ... And, I've found that easily occurs with a slight tad less "green intensity"(any "red push" doesn't help either in a slightly different sort of way) - even from a network feed during a game. And, As it so turns out, my "monitor" setting will probably end up being the best choice for this, anyway.

KenTech

02-05-06, 02:04 PM

On 34XBR960 with the "mode memory" function in user menu set to "on", I can use different "color axis"(either default or monitor) per pic mode, and per "input".I thought of this, too, but then one would be sacrificing one of the (only) four Picture Modes for a hairsplitting difference.

Consider how easy it is to "use up" the four modes for more important things. (Remember that Sony's mode-naming has no inherent significance. You just can't change it.) For example, my choices are:

Standard = My best overall settings at the current time. Used for all viewing unless something "special" is needed.

Movie = Exactly the same as Standard, except for gamma. For this mode, GAMM = 1 and the gamma amounts for that settings are set to 3 and GAMS to 0. Used to correct intolerable "gloom" in some programming.

Pro = More conservative settings than Standard, generally based on SYSM = 3, not 2, and tuned for smoothness. I try this when I think I am overdoing "enhancements" in Standard. 90% of the time, I go back to Standard.

Vivid = Ad-hoc experimentation. I make it identical to Standard, then change one or more settings to see if I have learned something new that's of value. I'm currently experimenting (again) with the VM-shaping parameters in 2170P-3. Another example: If I think I have spotted an issue with Standard, I "fix" it in Vivid, then live with it for a couple of days, switching back and forth to confirm results. If I have discovered a legitimate improvement, I transfer it to Standard, and then Vivid becomes available for another experiment. Vivid has become my "scratchpad" for settings.

Nowhere in this scheme is there room to assign an insignificant color-matrix shift to specific inputs. Mode Memory is OFF, since I might use any mode for any input. Obviously, I wish there were, oh, about a dozen or more possible preset "modes" some of which could be made input-specific.

Think of the problems folks have who simply want to balance black level among 5 discrepant inputs and have no inclination to do it in service mode! They're always playing with Brightness of leaving it set and complaining about the picture quality. (Or they don't see any difference. Ignorance is bliss!)

Nitewatchman

02-05-06, 02:59 PM

I thought of this, too, but then one would be sacrificing one of the (only) four Picture Modes for a hairsplitting difference.

Consider how easy it is to "use up" the four modes for more important things. (Remember that Sony's mode-naming has no inherent significance. You just can't change it.) For example, my choices are:

Yes -- I agree. I don't use different "Color axis" settings for different pic modes, and wouldn't want to for the reasons you mention. Which is part of the reason why I want "color" to match as much as possible from all inputs/sources/scan rates, was just posting that it is possible.

I actually brought it up to address the capability with "mode memory on" to use different RYR~GYB Color axis settings in an "input specific" manner, but whenever possible I usually like to be as complete and accurate as is possible so also mentioned it's possible to have different "color axis" settings for the pic modes as well. As I explained, for instance, I'm using "default" color axis choice for everything except sources from the internal tuner(NTSC/QAM or ATSC) and Memory stick for which I'm currenlty using "monitor" color axis choice. That's the only "difference" between Inputs I'm actually using "mode memory= on" for. [update : oops. I lied ;) I'm also having to use user menu Hue setting of "G1" or "G2" for my Dish network receiver via S-video, and Hue = "0" for a S-VHS deck via S-Video. [end update]

I use the pic modes pretty much exactly as you do, well not just pretty much --- I have them set up+use them exactly the same way as you are. I do keep trying to come up with some sort of other good use for "vivid" for when (if ever, LOL) I get done "experimenting".

I have been experimenting a bit with ABL settings, and if it were possible(which doesn't seem to be the case for YMLT), and If I ever get done "experiementing" with "vivid" it might be nice for instance to be able to set up say, "YMLT=2 or YMLT=1" with "vivid", as opposed to "YMLT=3" for the other pic modes.

Napoleon D

02-06-06, 11:39 AM

Ken - I had asked this question on another thread here (http://www.avsforum.com/avs-vb/showthread.php?p=7073004&highlight=sysm#post7073004) but this one might seem more appropriate.

For your Pro mode, where you have SYSM = 3, how are you working with that smoothness (which SYSM-3 creates) to bring out sharpness? SYSM-3 i think is a much more attractive picture that the 2 setting. Although SYSM-3 requires some kind of sharpneing, more definition to make the details more visible. You mentioned you used a MIDE 60 setting with the appropriate sub-category settings. I tried this, but it still did not sharpen up the picture enough. My calibrator has it all set this way for Pro (with a light MIDE setting). All edge enhancemnts (except for MIDE) are turned off. He also has sharpness slider in middle, which is convenient place to keep it. I was curious how you were sharpening the SYSM-3 image to the point where fine detail is easily visible, but not overshooting/ringing etc.

Thanks Ken!

KenTech

02-06-06, 05:41 PM

Ken - For your Pro mode, where you have SYSM = 3, how are you working with that smoothness (which SYSM-3 creates) to bring out sharpness?First, I think the "smoothness" is an illusion. If you reduce Sharpness for the SYSM=2 setting, you get smoothness, too -- it's just a different look. And the reason is probably that SYSM=2 really does reach into the very high video frequencies for its sharpening, and too much looks "scratchy."

To address the several questions you ask . . . My working model (in my head) is that these controls and filters, taken together, act as sort of a "graphic equalizer" for the video spectrum, analogous to the same for the audio spectrum in a stereo system. One additional factor, hard to find an "equalizer" analogy for, is the matter of visual unsharp-masking, which, for a scanning TV, is a matter in the time domain, not the frequency domain. The effects are positioned on-screen relative to the details on which they are acting. Example: a sharp vertical edge from 20% to 80% gray (dark on the left) would have a slight dark line added just before the transition to 80% and a slight bright line added just after it. That's if you turn the effect up too high to prove the point; normally one would add just enough of this effect to cancel an inappropriate softening of that transition becuase of losses in the video chain somewhere. So it's a great conpensator. I think that is SYSM=2's greatest strength, and SYSM=3 can't compete.

Here's sort of how I see these effects:

SYSM=2 = Great unsharp masking, especially with PROV=3. Affects mid-detail and up, the effect increasing with finer detail, like a video frequency-response boost-ramp. Overdone, it looks scratchy. My Sharpness is generally set about 25-31 for this. No effect is NOT at Sharpness slider=0 but more like at 10-12. Below that, Sharpness has a softening effect!

SYSM=3 = Effect occurs almost entirely to the right of the detail affected, and it appears to me more like a simple upper-frequency boost-plateau or peak in the small-detail frequencies. Overdone, it produces a bright line following a dark-light transition -- overshoot, but not much multiple-cycle ringing. It does not reach into the highest video frequencies, and so the finest detail remains unaffected. That's likely why it looks so smooth. (But I don't want Bruce Willis' face in Fifth Element to look smooth!) I choose to build my enhancement base on SYSM=2, unless the video source is way ovesharpened to begin with or has lots of film grain. (But that's what the DNR in my DVD player is for, no?) No effect is 0 on the Sharpness slider, unlike SYSM=2.

Now for the filters. The coarsest for DRC-processed video (480i) is 2103-1/SHAP. Above 2 or 3, it gives that cheaply-sharpened "TV" look to good video. I hate it, and for now it is set to 0, with PREO set to 0 or 1 to avoid a slight ringing side-effect. Maybe SHAP=1 would be helpful, but I haven't tried that yet.

Slightly finer is the boost in 3D-COMB #18, YPFG, for SD from RF and CV sources. I don't want it either, and so it is set to 8, which Sony indicates is zero effect. (Below 8, the image is softened.)

Medium-fine detail is enhanced by 2170P-3/F1LV settings higher than zero. Good for SD sources.

Only the very finest detail is enhanced by setting MID5/MHYE from 1-7, providing MYHO=0 and MHYL=3. If MHYO is set to 1, the affected video frequencies are lowered, and you can again choose MHYE from 1-7 for increasing effect. in other words, MHYO=1 makes the enhancement coarser -- which is just what is needed for SDTV, but not HDTV or fine DVDs. I find it hard to see the difference between the MHYO=1 group of MHYE settings and the same thing applied by 2170P-3/F1LV. Maybe the affected video frequencies are very similar.

For fine sources from DVD, I do all fixed sharpening in MID5 with MHYL=3, MHYO=0, and MHYE at 3 or 4. For HD sources, roughly the same, with MHYE at 2 or 3.

**************
So -- what to add to SYSM=3 to try for increased detail? I would use a column in MID5, set MHYL=3 and MHYO=0, and crank up MHYE from 1-7 to see if you can accomplish what you seek. That's the only filter that affects the finest details and textures. All others affect coarser detail, contributing to that "scratchy" thing.

I am baffled by the purpose of 2170P-3 nos. 10-12, LTLV thru CTLV. Sony calls it "LTI" level. I know it's zero at __LV = 0, and LTMD positions the effect such that LTMD=1 fattens vertical black lines, and LTMD=0 fattens white lines. If you look at frequency-sweep and resolution patterns on AVIA, you can see that the "LTI" effect is mostly with large details in the picture -- "midrange" rather than high frequencies. I have never found it to complement any image, so I never use it. Wish I knew what its intended purpose was; it's not just a filter but appears to operate in the time domain.

An aside: Filters that operate in the time domain give themselves away by shifting the picture to the right -- a delay. Compare 2170P-3/SHF0=0 with 1 (off). Or SYSM=2 with SYSM=3. Not a bad thing, if something important is accomplished. But it's a giveaway that there is not just simple frequency filtering going on!

2170P-3/SHF0 is another puzzle. Sometimes I have convinced myself that setting to 0 enhances SD video, but then I notice that fine textures are slightly attenuated by it, even on SD material -- and so I want nothing ot do with it. I have used my Vivid mode to experiment to see if I can live with something like this. (See my recent post #920.) I always keep returning to SHF0=1 becuase SHF0=0 does something ugly to the picture. The whole "LTI" group is even worse, attenuating fine detail a bit. No thanks!

The bottom line is that it is a balance among these choices that give your display a certain "look." With most real-world video sources, especially 480i, a "flat" video frequency response is a pretty soft or dull look, and some enhancements nicely fool the eye at a decent viewing distance into thinking the picture is better than it really is. That's the "art" of this and the reason to experiment and share results.

BTW, VM plays its part! Used discreetly, it does nothing negative, helps to compensate for the finite resolution of a CRT, and suppresses some sharpening overshoot. I use it on everything, even HD. I have spent a week at a time living with certain VM settings in the 2170P-3 group. Current favorites are:

VMCR~VMDL = 0-3-1-4 with actual VM level (Medium) of 6 (confirmed with VMLV).

Sorry to have been so long-winded, but getting the image right is a pet endeavor of mine, being a Photoshop junkie and a photographer. I think this series of TVs is absolutely amazing in how well they're engineered to get it right! If they were only adjusted this way from the factory . . .

Nitewatchman

02-06-06, 06:54 PM

Napoleon D,

Not meaning to "butt in" here -- just in case it is of any assistance, per some of the comments in your post in other thread --- if you haven't seen it you might want also want to check out Ken Tech's more recent(from Jan 06) posts and findings concerning "image processing". Most significantly, post Post 707 (http://www.avsforum.com/avs-vb/showthread.php?p=6916789&&#post6916789) in this thread and the attached file "IpChart05tall.pdf".

First, I think the "smoothness" is an illusion. If you reduce Sharpness for the SYSM=2 setting, you get smoothness, too -- it's just a different look. And the reason is probably that SYSM=2 really does reach into the very high video frequencies for its sharpening, and too much looks "scratchy."

I think so too at this point. One recent experiment I tried involved using your findings from Post 707 (http://www.avsforum.com/avs-vb/showthread.php?p=6916789&&#post6916789) .

Utilizing those findings, I set up both "vivid" and "standard" pic modes so everything was the same EXCEPT I set SYSM=3 for "vivid" and SYSM=2 for "standard". I then observed the results with various programming(including HD/SD from ATSC OTA sources) via switching between those 2 pic modes, in some cases changing the sharpness slider a bit.

Much to my surprise, there was little, (nothing of consequence really) I liked more about the SYSM=3 setting when compared to SYSM=2 in this way, although that had never been the case for me when I'd tried similar experiments before while using (mostly) your earlier image processing findings.

What is also puzzling, I think, are not only the SYSM "pro" mode set defaults in some cases, but also the description from Sony in the XBR2 service manual for SYSM. It says :

SYSM Bandwidth [0: NTSC, 1: FF, 2: HD, 3: DTV]

Ok, so "bandwidth" makes sense ... However --- why #2 for "HD" and #3 for "DTV?" In other words, what is that supposed to mean ?

"DTV"(for digital terrestrial Television(ATSC) I assume - Includes various 480i, 480p, as well as 720p and 1080i formats, the HD formats being 720p and 1080i) ... Just a WAG -- Could SYSM=3 perhaps be a setting which is there as a means for reducing visable MPEG2 compression artifacts when too high compression ratio is utilized ? Which of course, is an issue that is going to affect "HD" or SD MPEG2(or probably other codecs as well) sources other than via "DTV" ...

If there is any one setting I'd like to hear about in detail from the engineers at Sony who worked on this, it would be SYSM. If we could get a Sony Engineer as a "Genie in a bottle" sort of thing, what would be your 3 questions :)

Current favorites are:

VMCR~VMDL = 0-3-1-4 with actual VM level (Medium) of 6 (confirmed with VMLV).

Allways something new to try ;) -- Just kidding! Looking forward to it, as I've noticed VMCR~VMDL 0-3-2-8 with VM level 6 (set up for "medium" VM choice for user menu - which is where I generally most like it) seems to perhaps be doing a little "too much", to the point of it perhaps involving a bit of a "noticable" VM look to it on say, small, bright white text.

KenTech

02-06-06, 07:40 PM

SYSM Bandwidth [0: NTSC, 1: FF, 2: HD, 3: DTV]

Ok, so "bandwidth" makes sense to a degree ... However --- why #2 for "HD" and #3 for "DTV?" In other words, what is that supposed to mean ?Another Sony description: 0 & 1 = narrow peak, 2 = broad peak, 3 = flat. I don't believe a word of this!I've noticed VMCR~VMDL 0-3-2-8 with VM level 6 (set up for "medium" VM choice for user menu - which is where I generally most like it) seems to perhaps be doing a little "too much", to the point of it perhaps involving a bit of a "noticable" VM look to it on say, small, bright white text.I was experimenting there with using VM-delay (VMDL) for the first time to keep the "center" of hot, bright objects from shifting to the left with the addition of VM. I now think VMDL=8 was a bit much, and 4-6 is better. Further, the higher you set VMF0, the more low video frequencies are excluded from VM. If all you wanted is more detailing, this might work, but VM is too crude for mere sharpening. The effect is far more subtle than that, and the ability to reduce overshoot increases with the lowering of VMF0. There's a big jump in VMF0 from 1 to 0; maybe the high-pass filter is removed entirely, who can tell. But VM effects get a little "fat" with VMF0 at 0. (Notwithstanding Sony's choice of VMF0=0 for HD!) I am trying this other compromise for a while: 0-3-1-4 instead of 0-3-2-8. Last night's viewing of DVD and ordinary SD cable broadcast was very fine. Is it me, or is it VM?

BTW, in the case of "tuning" VM, I believe the effects are subtle and not readily appreciated when quick-switching between otherwise-identical modes. I.e. if I set up both VM groups above for two modes, then switch back and forth during a live local newscast, I'm hard-put to see any difference. But when (how to say this?) my eyes have "locked in" on the picture, I eventually notice something very right or wrong about it -- which is the reason I have canned 2170P-3/SHF0=0: looks okay on a quick-switch, but after a while I notice that something isn't right. So I really believe that, on these subtle changes, one should live with an evening's programming to see if your recent changes "fit."

KenTech

02-06-06, 08:32 PM

SYSM Bandwidth [0: NTSC, 1: FF, 2: HD, 3: DTV]

Ok, so "bandwidth" makes sense ... However --- why #2 for "HD" and #3 for "DTV?" In other words, what is that supposed to mean ?

"DTV"(for digital terrestrial Television(ATSC) I assume - Includes various 480i, 480p, as well as 720p and 1080i formats, the HD formats being 720p and 1080i) ... Just a WAGPlease don't take these descriptions too literally! I have begun to believe they are (1) badly-translated from Japanese and (2) full of errors. In the last charts I published, I copied the column labels from the service charts, but in many cases they make little sense, confusing ATSC, "V5, V6," YUV, etc. Sometimes OTA signals are referred to as "RF," in other places "ATSC." QAM (digital over cable) is never mentioned; AFAIK, QAM is *not* a subset of ATSC.

Eventually I hope to make corrections in the charts, but not just yet. I'm a little weary of testing!

Nitewatchman

02-06-06, 10:11 PM

I eventually notice something very right or wrong about it -- which is the reason I have canned 2170P-3/SHF0=0: looks okay on a quick-switch, but after a while I notice that something isn't right. So I really believe that, on these subtle changes, one should live with an evening's programming to see if your recent changes "fit."

I had much the same experience with SHFO. In short, I had it set at "0" for quite some time until I decided I didn't like it.

BTW, in the case of "tuning" VM, I believe the effects are subtle and not readily appreciated when quick-switching between otherwise-identical modes.

Absolutely. I'll often go weeks before deciding on implementing a change, especially as I said earlier, I don't spend a lot of time "watching TV" on this set. Which is part of why I *very* much appreciate your work on the image processing settings, and, rather than spending a LOT of time on it experimenting(although it has been fun to check things here and there and do a bit of experimenting) I've been very happy using what you've found - especially your latest findings.

Anyhow, In addition to programming material(which we are most interested in of course), I also have a few things I look at closely when making or deciding on changes.

Such as : #1) various Local news graphics.

#2) A small ID bug one of my local PBS stations uses.-- It's a small circle, inside is white background with very small black text which Says "Think TV" (the "TV" is an even smaller "superscript"). If, for example I can't read that clearly from say 10 feet away, I know there's something wrong, either with my eyes+eyeglass perscription, or something has "changed" for the worse with the "image processing" settings.

#3). Another example(this one having to do solely with color) involving -- The EPG info from Dish network 311 receiver. For the channels I'm not subscribed to, they're "shown" in Red in the channel listing. If that red is a "cool" red, or undersaturated or oversaturated I immediately notice it.

#4). Also, concerning the 311's EPG info -- If I press the "info" button once on the remote, the programming description/guide/etc. pops up with small, bright, white "hot" lettering on a "clear" background". For instance, Some even Subtle changes in VM settings can especially be visable here - but we aren't talking big differences, and it might take a while before I notice a subtle change has changed things for the worse, or better for that matter. Press the info button a 2nd time, and the text turns black with a fairly bright "eggshell white" background(it's a rather large "box" which fills about 2/3 or more of a 4x3 frame). Any "Ghosting" issues can be especially apparent around that black text.

Please don't take these descriptions too literally! I have begun to believe they are (1) badly-translated from Japanese and (2) full of errors.
."

That was one of the exact points I was trying to make, when I said "what is that supposed to mean?" -- Perhaps another clarification that might help --" WAG = Wild A** Guess. "

It's sort of like those "super green motherboard" manuals I used to seem to come across a lot in the Mid-90's that were written in Taiwan/china and translated to English. Stuff like "Jumper the M54 setting PIO mode 4 "/etc ...

Just especially gets very confusing at times in this case when your going through and looking at the charts at all the "columns ... "

AFAIK, QAM is *not* a subset of ATSC.

It's not, QAM is the signal modulation that is most used by cableco's for "digital" cable. 8VSB(Trellis coded 8-level vestigal sideband) is signal modulation used for OTA "digital"(ATSC) signals in U.S. There is also a little known signal modulation in the ATSC specs for use by digital cable that is "16VSB", and requires less FEC(forward error correction) than the OTA version. I don't think anyone is using it in U.S. as I think Cableco's(and decoders/"digital cable ready sets") have pretty much "adopted" QAM as the standard they want to use, but not sure.

16VSB has about the same data payload in a 6MHZ RF channel as QAM256 -- about 38Mb/s . Because of the extra error correction needed for "robust" OTA transmission/reception, with 8VSB, the data payload rate is 19.38Mb/s for 6MHZ wide RF channel. Thus, cableco's can "fit" 2 entire 19.38mb/s datastreams from 2 seperate digital ATSC stations into a single 6MHZ channel. Thus, there are often "two" slots per QAM channel. I've even read that it might be possible it can be 3 stations and 3 slots in one 6MHZ channel with the use of "rate shaping"(which doesn't require an add'l decoding/reencoding process).

With these sets which are "digital cable ready", the receiver "front end" has ATSC+QAM "tuners". In most cases, (in all cases I know of in my area), when a cableco distributes a station's digital "signal", the Transport stream(MPEG2 Video/AC-3 audio/etc) remains the "same" as what OTA viewers get(even the PSIP info), the only difference is the cableco doesn't Use the ATSC 8VSB "RF"(signal modulation) portion of the signal, and instead distributes the datastreams via QAM 256 on their system.

I suppose MPEG2 could in a way be thought of as usedas an important "subset" of ATSC - --- and you'll find a LOT of "MPEG2" info in ATSC white papers, but of course MPEG2 is not a "ATSC only" thing(DVD's use MPEG2 for example), and other codecs besides MPEG2 CAN be used with ATSC OTA as well -- although no current receivers I know of will work with anything else other than MPEG2.

Sometimes OTA signals are referred to as "RF," in other places "ATSC."
."

Without going into great detail or accuracy here -- in the charts, I think it's working out that "RF" is refering to the NTSC (analog) "tuner" -- including for either OTA or analog cable(which uses the same "tuner", and channel numbers/frequencies on VHF as OTA, but different channel #s along with in part different frequencies on UHF cable bands). I think ATSC on the charts is referring to both ATSC(OTA) and QAM -- as, even though the tuners and RF signal de-mods that need to occur are "different" for 8VSB+QAM -- after that, For instance, I think whether you're getting a datastream originating from a station's MPEG2 encoeder via OTA via the OTA 8VSB RF signal or via cable via the QAM RF signal the cableco is sending via the "wire", the MPEG2/AC-3 audio/PSIP/etc. streams involved are going to be the same, and are also going to use the same MPEG2 decoder in the set+ probably the same signal paths(at least mostly)/etc.

Update :

I can actually hook an antenna up to the "cable" RF input on KD34XBR960, and it will work just fine with 8VSB ATSC signals on VHF(I have a local station transmitting on VHF 10) via the "cable" input - even though it isn't QAM - shows up as C9.1 with the PSIP virtual remapping. It Will NOT work for UHF ATSC stations.

end update

I'm a little weary of testing!

Me too! I do know I had the chance to watch quite a bit of TV on the XBR960 yesterday, Including the ABC HD superbowl coverage and everything looks so good at this point -- I don't feel all that motivated to do too much more of anything .. Of course, I said that about 6 months ago as well ...

loadams

02-07-06, 10:53 AM

Ken, don't mean to ask a silly ? here, but............

I've got the older, bigger brother 40" gorilla. Now with 9 of 10 dvd rentals being in 16x9, why would I want to continue using 480p (component) when the set upconverts those dvds to 960i anyway? Pros.....cons ?

And that being said..... I have balanced my color decoder (13,14,6,3) with 480i and tweaked 480p with CXA2151, however, there is NO WAY to have a happy medium with 480p and 1080i on my set !!! Which is giving me a good reason to dump 480p altogether for the sake of decent color decoding in 1080i.

I guess I'm asking............ if 960i doesn't touch CXA2151, dump 480p ? Your opinion ?

SurfingMatt27

02-07-06, 11:09 AM

justsc

02-07-06, 12:00 PM

You want a sharp picture, turn all the MIDE5 settings to 0, your PQ will be ultra sharp with no enhancements neccessary.Just look at text and tell me how perfect it is.
Matt,

I'm with you on almost all of the settings you have suggested - and have a very nice picture - thanks! ;) I realize now that "adding" to the original picture (signal) can actually take away from the overall fidelity.

But, what if there's elements in the signal that I'm not seeing - elements that could make it even sharper? This is the direction my thinking is going. Like when raising the black level on a overly dark "dusk" scene and seeing how much detail I was missing (just an example).

I'm just thinking there may well be more "sharpness" to be enjoyed by "bringing it out" rather than "adding it in."

So, I'm focusing on finding ways to "expose" levels of fidelity that are already there but I haven't seen - yet.

This is why I find the latest "back and forth" between KenTech, Nitewatchman and Napoleon D so interesting. This is especially important since my 34HS420 doesn't have the SFP tube.

tlniec

02-07-06, 12:07 PM

Great thread, lots of information here! I have a question regarding a standard-def Sony CRT (KV-32FS100). I've tweaked the user menu settings and gotten some good results. However, I think I'm ready to take the next step and delve into the service menu. I have logged all the existing service menu settings (but have not yet changed any). Do you guys have any advice for a Sony CRT that only sees 480i (maybe 480p for DVD) input from DVD over component, VCR over composite, and Digital Cable over composite? I suspect I might be able to get by tweaking a relatively small subset of the parameters KenTech and Nitewatchman have posted, but I want to make sure. Thanks! :)

SurfingMatt27

02-07-06, 07:40 PM

Matt,

I'm with you on almost all of the settings you have suggested - and have a very nice picture - thanks! ;) I realize now that "adding" to the original picture (signal) can actually take away from the overall fidelity.

But, what if there's elements in the signal that I'm not seeing - elements that could make it even sharper? This is the direction my thinking is going. Like when raising the black level on a overly dark "dusk" scene and seeing how much detail I was missing (just an example).

I'm just thinking there may well be more "sharpness" to be enjoyed by "bringing it out" rather than "adding it in."

So, I'm focusing on finding ways to "expose" levels of fidelity that are already there but I haven't seen - yet.

This is why I find the latest "back and forth" between KenTech, Nitewatchman and Napoleon D so interesting. This is especially important since my 34HS420 doesn't have the SFP tube.

Well Steve,..You can't really add detail that is'nt there,all you will be doing is introducing stuff into the picture that does'nt belong resulting in noise, or artifacts.My advice is to keep the signal clean and you will thank yourself later on trust me ;)

I was skeptical as you a year back and found out that it's not really neccessary at all, since i noticed all your doing is adding false information that is'nt there.

I just keep it simple and feed the signal to my tv without any processing, easy breezey :cool:

As a matter of fact i'm thinking of recalibrating my tv, mainly only touchup.Brightness and color mainly is my only issues, since i think i may have a low setting right now and it could be higher,And color although good, i think i can tweak it a bit.It may not be D6500k but it will look good none the less by simple refinements.

Beback in a few.

ADU

02-07-06, 08:52 PM

Apologies for jumping in the middle of a discussion, but just thought I'd throw in my 2c re some of the 2170P-3 edge controls, since I've been experimenting with these the last few days. This is still sort of a work in progress, so editing / retractions / updates (without notice) are likely, if not certain.

Some of the controls seem to operate with finer granularity depending on which SYSM setting is used, so I've broken them down into two categories based on that.

All of these controls seem to have their strongest effect in a horizontal direction and little or no effect in a vertical direction, so their behavior is usually easiest to see on vertical edges. And when I refer to "vertical edges" or "horizontal detail" below, I'm really talking about the same thing.

All of these observations were made using patterns/video content at 480p (input at 33.75khz) via DVI on a 34XBR800.

NOT (OR DOESN'T APPEAR TO BE) SYSM-SENSITIVE

SHF0 seems to change the "frequency" of sharpening in the image, creating a subtle "loosening" or "tightening" effect on horizontal details. 0=loose. 1=tight. It's easiest to see the effect with Sharpness raised on a pattern like this (http://www.avsforum.com/avs-vb/attachment.php?attachmentid=50979) with SYSM=1. Though you can also sense the tightening/loosening effect in video content at nominal Sharpness as well. I don't think this control "scales" with SYSM, but it does effect the character of edge-distortions (ie ringing/overshoot) with each SYSM setting.

PROV is also easiest to initially see with Sharpness elevated. And seems to shift the "emphasis" to either the left or right of the edge in small increments. Like SHF0, the effect of this control is pretty subtle at nominal Sharpness settings, but it can be perceived as sort of pulling your eye towards either the left or right side of the screen. So it may have some value in achieving better "equilibrium" in the picture.

PROV may need to be tweaked differently depending on how SYSM and other controls are configured. Although it's easier to see the behavior of this control at higher Sharpness settings, I seem to get more reliable results adjusting it at nominal Sharpness, based on the way it seems to pull my eye toward either the left or right of the screen.

LTLV functions differently depending on how LTMD is configured.

If LTMD=0, then LTLV seems to smooth out details that are similar in brightness/color, while sharpening those that are aren't. So you may be able to think of it as sort of a combination low-con noise-reducer/hi-con edge-enhancer. It particularly seems to reduce "noise" (and maybe also some clarity) in darker areas of the image, ie shadow detail.

If LTMD=1, then LTLV seems to have a more generalized sharpening effect, basically reducing softness in the picture across the board, including shadow detail. This could be useful for enhancing shadow detail on 480p sources with SYSM=1. Whether it would be beneficial with the other SYSM settings, I don't really know yet. Preliminarily it looks like it might perhaps be a little heavy-handed with SYSM=2.

To see any effect from the two LTMD settings, LTLV has to be higher than 0 (ie between 1 and 3). LTLV=3 has the strongest effect. And the behavior of these controls is probably seen most easily at nominal Sharpness probably with SYSM=1. You can probably get a better feel for the difference between the two LTMD settings if you set LTLV to 3, and then flip back and forth between LTMD=0 and LTMD=1 while watching something with a pretty good level of detail like AOTC.

SYSM-SENSITIVE

F1LV seems to better define vertical edges, by boosting brightness on the lighter side(s) of the edge. Its effect can be seen fairly easily on patterns of fine dark detail (http://www.avsforum.com/avs-vb/attachment.php?attachmentid=50979) or text against a lighter background, with Sharpness at nominal settings. This control is sensitive to how SYSM is set (ie it seems to operate at higher frequency/smaller granularity with SYSM=2/3), and it's easiest to see it's effect with SYSM=1.

Although this does increase "noise" in the picture (especially at SYSM=1), it may have some "defining" value in terms enhancing shadow details, especially at the higher granulity with SYSM=2/3.

CTLV seems to control the saturation of color on fine horizontal detail where two different colors meet. You can see the effect fairly well on the attached CTLV patterns with SYSM=1 at nominal Sharpness. With SYSM=2/3, CTLV seems to operate at a higher frequency/smaller granularity (watch the thinnest lines in the CTLV1.JPG pattern with SYSM=2/3).

KenTech

02-07-06, 10:13 PM

CTLV seems to control the saturation of color on fine horizontal detail where two different colors meet.I very much appreciate your additional insights on these "enhancement" parameters. Some of them are very hard to describe, and it's good to hear another view.

I'll bet that these parameters similarly named but starting with "L" or "C" are the same filters acting on Luminance and Color information, respectively. Whatever Sony's mysterious "LTI" is, there seems to be an equivalent "CTI," controlled by LTLV and CTLV. When I experimented with them, I didn't like what they did to light/dark boundaries, and I did notice a slight softening of fine textures. But I never thought to check for shadow vs highlight effects, as you described. I will now have a look, too -- although I am skeptical that I will ever find a use for them.

In the MID5 parameters, there are groups of filters that enhance color sharpness (MHCR, MHCE, MHCO) just as the ones for luminance do (MHYR, MHYE, MHYO) -- except here the luminance is referenced by a "Y." There is a downside to sharpening color, as color boundaries, especially between certain color pairs, are already screwed up by the color encoding during compression or by the NTSC broadcast standard. I once noticed that enhancing sharpness for color alone was not increasing the attractiveness of the picture but drawing attention to these artifacts. So I have chosen to leave color alone for now, as evidenced by all of the zeroes in the MID5 color parameters that I have recommended. The eye is known to be remarkably uncritical of color resolution, so maybe it's a lost cause.

Live and learn, though. Maybe there is a role for these color-signal sharpening features. I just haven't found it yet.

BTW, where did you get those lovely color patterns? Are there more?

KenTech

02-07-06, 10:24 PM

I've got the older, bigger brother 40" gorilla. Now with 9 of 10 dvd rentals being in 16x9, why would I want to continue using 480p (component) when the set upconverts those dvds to 960i anyway? Pros.....cons ?What I can't predict for you is how your set handles these different video standards visually. On my 36XS955, for example, the set is more competent dealing with a 480i input in producing a marvelous picture, whether interlaced 960 or the CineMotion equivalent. If I feed 480p via component inputs, there is a limit on the detail somewhere -- the picture has lost some important texture information, and it can't be recovered. So I never use it. (HDMI is better, but that's another discussion.) Admittedly I can't tell if this is a shortcoming in my Panasonic S97 player's Faroudja/Genesis chip or a characteristic of the TV. I just know there's nothing superior to the component inputs for 480i, and so I would sacrifice 480p in a heartbeat!

So, that said, I would dump 480p, assuming you're happy with how it handles the 480i. Then you can twiddle your 2151 chip's settings to make 1080i similar in color rendition.

KenTech

02-07-06, 10:35 PM

You want a sharp picture, turn all the MIDE5 settings to 0, your PQ will be ultra sharp with no enhancements neccessary.Just look at text and tell me how perfect it is.This certainly should be, IMO, one of the columns available for your MID5 table. On mine, it's #63. But to say all-zeroes is all you need in all cases is ignoring what MID5 can do for you in many cases. All of the settings from #5/MHYR thru #18/MVCE are sharpness *enhancers* that have no visible downside. That is, they don't exaggerate some detail while squashing finer detail. The use of MHYR=3, MHYO=0, and then choosing a MHYE from 1-7 really restores fine detail and texture nearly lost in the video-processing chain. Let your eyes be your guide! Changing MYHO to 1 coarsens the effect, which is useful for SD video. Do what looks right to your eyes. There are no "best" settings!

KenTech

02-07-06, 10:52 PM

'm just thinking there may well be more "sharpness" to be enjoyed by "bringing it out" rather than "adding it in."

So, I'm focusing on finding ways to "expose" levels of fidelity that are already there but I haven't seen - yet.Well, this is the philosophy behind most of the "enhancements" discussed here. What Matt said, that you can't recover what isn't there, is true, indeed, and oversharpening is ugly! But sometimes detail and texture are really there and attenuated, like rolled-off highs in a sound system. A little boost restores the fidelity -- as long as you can tolerate a bit more noise, because you'll be boosting that, too.

The second reason to add sharpness to a soft-ish picture is to fool the eye when the picture is viewed at an appropriate distance -- the basic technique with broadcast SD, such as analog cable. Get too close, and you can see what crap the picture is; but get back a ways, and the picture integrates in your eye/mind and doesn't look so bad. If you add a little sharpening to the mix, that illusion can be improved, even if there is some outlining when viewed close. (Solution: Don't view close!)

As is obvious from reading Matt, viewing preferences vary a lot from person to person, and there is no "right" choice for everyone on what makes a great picture. Further, many of these judgements can't be made in a snap -- one has to live with a setup for a while to see if it's good.

ADU

02-07-06, 11:23 PM

I very much appreciate your additional insights on these "enhancement" parameters. Some of them are very hard to describe, and it's good to hear another view.Agreed. See some of my already add'l edits above re the "tightening" effect of SHF0, and the "pulling" effect of PROV. :)There is a downside to sharpening color, as color boundaries, especially between certain color pairs, are already screwed up by the color encoding during compression or by the NTSC broadcast standard. I once noticed that enhancing sharpness for color alone was not increasing the attractiveness of the picture but drawing attention to these artifacts. So I have chosen to leave color alone for now, as evidenced by all of the zeroes in the MID5 color parameters that I have recommended. The eye is known to be remarkably uncritical of color resolution, so maybe it's a lost cause.

Live and learn, though. Maybe there is a role for these color-signal sharpening features. I just haven't found it yet.Raising CTLV seems to reduce the color saturation at boundaries, softening, rather than sharpening color detail. So it might perhaps be useful for reducing certain color artifacts, or taking some of the bite off high-contrast color edges in some instances.BTW, where did you get those lovely color patterns?Made 'em myself in Photoshop!

SurfingMatt27

02-08-06, 01:15 AM

Well, this is the philosophy behind most of the "enhancements" discussed here. What Matt said, that you can't recover what isn't there, is true, indeed, and oversharpening is ugly! But sometimes detail and texture are really there and attenuated, like rolled-off highs in a sound system. A little boost restores the fidelity -- as long as you can tolerate a bit more noise, because you'll be boosting that, too.

The second reason to add sharpness to a soft-ish picture is to fool the eye when the picture is viewed at an appropriate distance -- the basic technique with broadcast SD, such as analog cable. Get too close, and you can see what crap the picture is; but get back a ways, and the picture integrates in your eye/mind and doesn't look so bad. If you add a little sharpening to the mix, that illusion can be improved, even if there is some outlining when viewed close. (Solution: Don't view close!)

As is obvious from reading Matt, viewing preferences vary a lot from person to person, and there is no "right" choice for everyone on what makes a great picture. Further, many of these judgements can't be made in a snap -- one has to live with a setup for a while to see if it's good.

Exactly Ken :)

I feel your pain..it seems me and you are in the same boat always trying settings and seeing if you can live with them and test it with real material then constantly im changing parameters left and right.

Well after all this tinkering i realize you are never going to get a perfect picture but at least it's watchable,Even ISF is'nt guaranteed perfect given the limits of CRT technology.

I'm in the market for a replacement anyways, not that i don't like my sony it's just that me and my brothers are looking for a bigger display for gaming and movies and the closest we have found to match crt in PQ is DLP, so that's what i'm going to be looking into, especially the new 1080p OLED DLP's coming soon or the HP DLP tv.

peace,

Matt~

Q of BanditZ

02-08-06, 09:41 AM

Exactly Ken :)

I feel your pain..it seems me and you are in the same boat always trying settings and seeing if you can live with them and test it with real material then constantly im changing parameters left and right.

Well after all this tinkering i realize you are never going to get a perfect picture but at least it's watchable,Even ISF is'nt guaranteed perfect given the limits of CRT technology.

I'm in the market for a replacement anyways, not that i don't like my sony it's just that me and my brothers are looking for a bigger display for gaming and movies and the closest we have found to match crt in PQ is DLP, so that's what i'm going to be looking into, especially the new 1080p OLED DLP's coming soon or the HP DLP tv.

peace,

Matt~

You got a spare kidney you wanna part ways with? ;)

justsc

02-08-06, 10:48 AM

...I'm in the market for a replacement anyways...

...the closest we have found to match crt in PQ is DLP, so that's what i'm going to be looking into, especially the new 1080p OLED DLP's coming soon or the HP DLP tv.

peace,

Matt~
Matt!

A 1080p OLED? Are you sure that's what you really mean? The largest yet shown was about 20" and they couldn't even estimate a price because there's still many bugs to work out. OLED and DLP are not technologies used together, it's either - or.

We are researching FOLED display technology for the Heads Down Displays in our simulation cockpits and the prices are astounding. For a 13" display you're looking at about $40K, and that's not 1080p.

We're probably about 2-3 years out from SED at a decent price point, and about 5-10 years out for OLED at any kind of size and decent price point.

I recommend you look at HP's 1080p DLP - it's very, very nice.

Cheers! ;)

SurfingMatt27

02-08-06, 10:49 AM

LOL yeah, my left nut:D

But seriously though, i'm not going to make thepurchase to about late march oe april so it gives me plenty of time to save up.The sony's been great these past 2 years, i just wanted something bigger.I was planning on buying a bigger tv in the first place but did'nt have enough cash at the time when i purchased my sony.

Any thoughts on specific DLP models i may be interested in please feel free to PM me.Or better yet i'll just make a topic thread in the RP display thread.

That way we can get back on topic of Ken's thread.

Matt~

SurfingMatt27

02-08-06, 11:00 AM

Matt!

A 1080p OLED? Are you sure that's what you really mean? The largest yet shown was about 20" and they couldn't even estimate a price because there's still many bugs to work out. OLED and DLP are not technologies used together, it's either - or.

We are researching FOLED display technology for the Heads Down Displays in our simulation cockpits and the prices are astounding. For a 13" display you're looking at about $40K, and that's not 1080p.

We're probably about 2-3 years out from SED at a decent price point, and about 5-10 years out for OLED at any kind of size and decent price point.

I recommend you look at HP's 1080p DLP - it's very, very nice.

Cheers! ;)

Thanks for the advice and yeash your probably right, maybe i meant to say LED DLP are'nt these supposed to be coming out soon later in april?Oh and thanks for the suggestion of the HP i've been hearing a lot of good things about it and will look into that as well as any other advice would be greatly appreciated.

Well i'll be making another thread now in the RP display forum, feel free to help me out anytime since i'm new to DLP and would like some good suggestions.

I'll be selling my sony too by the way for cheap, anyone interested please PM me, since i really have no need for it anymore once i get my new replacement soon.My father wants me to get rid of it any ways since theres really no room to accomodate it now because of the bulk and weight and we have enough tv's as it is in our house about 8 by the way :p ....So i can't really keep it.

Peace,

Matt~

justsc

02-08-06, 11:39 AM

...Oh and thanks for the suggestion of the HP i've been hearing a lot of good things about it and will look into that as well as any other advice would be greatly appreciated.

...i'm new to DLP and would like some good suggestions...
Matt,

The newest 1080p sets have alot to offer. I recently had the pleasure of calibrating a Sharp Aquos 45" 1080p LCD set - what a fantastic picture! In my search for that set I looked very closely at Sony's 50" 1080p SXRD set - awesome as well.

HP is getting excellent reviews of it's two new 1080p DLP sets, and they stand apart as some of the only ones that can actually accept as well as display 1080p.

Here's a piece from CNET:

The MD6580n also offers extremely impressive all-around performance, and although it didn't score quite as well during testing as Sony's SXRD-based KDS-R60XBR1, it came close enough to make it our favorite DLP-based HDTV yet.

You can also find good reviews here:

http://www.ultimateavmag.com/rearprojectiontvs/106hp/

http://www.hdblog.net/2006/01/23/review-hp-md5880n-ultimate-av/

Best of luck in your search! ;)

SurfingMatt27

02-08-06, 12:18 PM

Thanks steve :)

Anybody you know looking for a sony 34" for cheap?If anybody's intertested in my sony 34hs420 just e-mail me or PM me.

By the way is'nt there a forum section for selling stuff?

Matt~

Nitewatchman

02-08-06, 02:06 PM

It might be interesting to hear from someone who has worked with the relevant 2170P-3 and MID5 settings on both a HS420 and one of the SFP sets.

Perhaps we are mostly dealing with issues relating to what any given person "likes", and as Ken said -- there is no "right" choice of settings for everyone ... Perhaps such issues as differences in output from DVD players also are involved.

Nevertheless, I wonder if it is possible we might also be getting something a little different on this depending upon if we are using SFP tube or no, or maybe there are some other differences between the different models which may be involved. Also -- apologies if this has been covered -- when, for instance we are talking about, say 480p or 480i from DVD via V5/V6 component inputs and SYSM=3 and MID5 column's at all Zero's -- are we also using the same, other relevant 2170P-3 settings? I think ADU's excellent info on these demonstrates how slight differences in how they are set up can be important as well.

-----------------------------

Anyway - FWIW - With ADU's excellent info on the relevant P2170-3 settings printed out and brought along -- Played around with these in a "seat of the pants" manner a little last night With KD34XBR960 and 480p DVD(Zenith DVD2381) via component input. Difficult to say how much of this, if any is "specific" to my DVD player.

I was working with these in my "experimental" pic mode(just happens to be "vivid"), to easily make comparisons to the settings I normally use, which are set up for "Standard" pic mode. Along with test patterns from AVIA, "The Phantom Menace" happened to be the first DVD I grabbed - and although not the best(It seems to have quite a bit of EE "built in" in spots for instance), various scenes from it, as well as its "THX Optimizer" along with Text+graphics in my DVD Players "setup" menu is what I ended up looking at.

Started with a MID5 "all zeros" column and SYSM=1 to more easily see what was going on. Then tried SYSM=2 or 3, making slight adjustments as necessary. I also played with changing MID5 columns set up with different settings along with making small changes to the "vivid" mode settings and trying different MID5 settings/etc.

Again, Just doing "seat of pants" stuff here - FWIW (not taking the time as is required to investigate in detail, or spending a lot of time "living with" small changes) - and just given a couple of hours playing with it :

#1) I thought it was interesting what one can do w/o use of MID5 column when one knows a little more of what is going on with these 2170P-3 settings.

#2). I do think I was probably able to come across the best results I've managed to date for 480p DVD with SYSM=3, utilizing the following applicable 2170P-3 settings - In brackets are my current settings I use with "standard" and "movie" pic modes (Note I have customized my Pic modes per KenTech's "customizing Picture modes" article earlier in thread - update : also note I included the VM "
shaping" settings, as well as I was using "VM" "medium" in user menu, with P2170-3 #19 VMM = 6 (set default was 8) :

#0 SYSM - 3 [2]
#2 VMCR - 0 [0]
#3 VMLM - 3 [3]
#4 VMFO - 2 [2]
#5 VMDL - 8 [8]
#6 SHOF - 0 [0]
#7 SHFO - 1 [1]
#8 PROV - 3 [3] (Note : I first ended up with PROV=2 for SYSM=1)
#9 F1LV - 1 [0]
#10 LTLV - 2 [0]
#11 LTMD - 1 [1]
#12 CTLV - 1 [0]
#16 MIDE - 60 [60, 48 or 49 - haven't decided yet]

And following MID5 column (Settings not noted are all set at "zero")

#1 POP - 60
#6 MHYL - 3
#7 MHYE - 3
#10 MHCL - 3
#11 MHCE - 3

For reference, here's My MID5 Column #48 (again - Settings not noted are all set at "zero")

#0 POP - 48
#6 MHYL - 3
#7 MHYE - 3
#10 MHCL - 3
#11 MHCE - 3
#14 MVYL - 3
#15 MVYE - 1

And MID Column # 49 (Settings not noted are all set at "zero"):

#0 POP - 49
#6 MHYL - 3
#7 MHYE - 3
#10 MHCL - 3
#11 MHCE - 3
#14 MVYL - 3

With previous experiments (which didn't work out very well) trying to use SYSM=3 for 480p DVD --- which mostly involved emphasis on "tweaking" MID5 column settings with(except for SYSM=3) the same 2170P-3 settings as shown in brackets above -- I was never able to bring out the "details" without adding too much EE or "ringing".

Last night, with SYSM=3, I found If I increased MHYE slightly to 4 or 5 for the experimental "vivid" mode settings noted above, I was just starting to approach some slight, noticable "ringing" occuring as evidenced in text in the DVD player's "setup" menu, and MHYE=2 may have been best.

I think it was (with LTMD=1) LTLV at a higher value which "helped" the most with this, concerning an overall increase in "sharpness" using SYSM=3. However, although I can see where it would be beneficial in some cases, one issue I had with raising LTLV is that I didn't really *need* to increase "shadow detail".

#3). Using SYSM=2 instead of 3, I noticed I was also able to get somewhat "close" to what I'm getting from my current "standard mode" settings for 480p DVD. Except with the use of "all zero's" in the MID5 Column. "Ringing" didn't seem to be an issue(which is especially noticable when it occurs in text for my DVD player's "setup" menu), nor was visable EE being an issue. -- but, I couldn't get "close enough" when it comes to "bringing out" the details.

#4). Using SYSM=2 and bringing the MID5 settings back into the equation -- I still couldn't seem to find anything better than my current "standard" mode settings, although again, this is "seat of the pants" stuff -- I tried to "back off" on MHYE/MHCE/etc, and also tried to "back off" several of the 2170P-3 settings while adjusting MID5 settings as necessary to try to find a better "compromise". I couldn't find one, even if the only MID5 settings that weren't "0" were MHYL=3, MHYE=1 -- For example, that resulted in "ringing" in the text on the DVD player setup menu as compared to my "standard" mode settings. I did however end up spending quite a bit of time looking closely at a CTLV setting of "1", instead of "0", and, if anything, it looks like it might be possible a different CTLV setting may be something I might want to look at more.

#5). While I didn't spend much time with test patterns for anything of importance(and can only use MS with the patterns ADU posted, as I don't have a DVD recorder), perhaps of some interest - some of the most apparent "changes" I noticed using various test patterns involved the "sharpness" bars in THX optimizer when changing values for "SHFO, F1LV or LTMD/LTLV", and the boundries/edges between color bars (such as from AVIA) and adjusting "CTLV". The effects of increasing LTLV with LTMD=1 were also noticable on THX optimizer "black level"("brightness") adjustment screen.

Update: Oh - for all above experiments, I left user menu Sharpness slider at 31 - 30~31 is also what I have it for normal viewing with "standard or movie" pic modes.

loadams

02-08-06, 02:37 PM

Thanks Ken for your reply, will probably go in that direction.

And yes, after watching 480p for a number of years, watching 480i w/ tweaks is very pleasing !!! Great stuff as always guys.

ADU

02-08-06, 04:07 PM

A couple more thoughts on the 2170P-3 edge controls...

SYSM
So far SYSM=3 seems to be giving a bit of a biased look to edges on my TV that's a bit uncomfortable to my eyes. It's possible that tinkering with other controls may be able to make this a bit more manageable, but so far I haven't spent much time on that, and have been focusing more attention on SYSM=1 and SYSM=2. Edge distortion is lowest with SYSM=1 at nominal Sharpness on my TV. SYSM=2 seems a little "noisier" looking, but I can see alot of potential in the higher granularity and generally better definition it seems to confer especially to lighter details and shadow information in the picture versus SYSM=1.

Per something Napoleon D said in another thread, rather than simply randomly flipping back and forth between the different SYSM settings while working with the other controls, I think there's probably something to be said for focusing attention on one SYSM setting at time (once you've got a pretty good grasp of the effect each 2170P-3 edge control has). And trying to break down the best configuration of edge controls to go with that particular SYSM setting. Then move on and repeat the process with the next SYSM setting, working in sort of hierachical fashion to narrow down the best configurations or range of potential adjustments for each (with a given input/signal).

However, I would not make the mistake of assuming that just because a particular 2170P-3 parameter doesn't "help" with one SYSM setting, that that control should necessarily be thrown out of the mix for the other SYSM configurations. Because as mentioned above some controls seem to work with finer granularity and more subtle effect with SYSM=2/3 than they do with SYSM=1. I'm thinking particularly of F1LV here, which seems to have a much finer and potentially more beneficial "defining" effect with SYSM=2/3 than with SYSM=1.

SHF0
In addition to a loosening (SHF0=0) and tightening (SHF0=1) effect, I think I'd also describe SHF0 as having sort of a deepening (SHF0=0) versus shallowing (SHF0=1) effect on the picture. The higher "frequency" of the detail with SHF0=1 seems to "flatten" the picture a bit. Although one generally thinks of depth as being a good thing, I wouldn't necessarily describe the shallowing effect of SHF0=1 as "bad" on my TV. Although it seems to add a bit more noise/grain/texture, whatever you want to call it, the shallow picture may perhaps be a little easier on my eyes in some ways. The greater depth, clarity, looseness of SHF0=0 has some positives as well though. So far, it's been kind of a tough call. Another way I'd describe the difference is SHF0=0 seems more "TV-like", and SHF0=1 seems more "monitor-like", if that means anything.

Since there only two choices here, one might want to try to work out separate configurations for each, and then decide which is preferred.

PROV
No noteworthy additions on this yet.

F1LV
Although this generates fairly coarse looking "noise" with SYSM=1, as mentioned above it seems to have some greater potential with SYSM=2/3, and may have some value in elevating shadow information so it's a bit less strenuous on the eyes.

LTMD=0/LTLV=1-3
Still tinkering with this. Haven't found much use for it yet, but might have some potential in small amounts with certain configurations. And might be useful in cleaning up especially noisy signals. The effect LTMD=0 has at high LTLV settings reminds me alot of the "Vivid" mode on Panasonic plasmas, the way it smooths details similar in color/intensity, and over-defines high-contrast edges.

LTMD=1/LTLV=1-3
Although this seemed to have potential for better-defining shadow detail with SYSM=1, my enthusiasm for it has diminished a bit after watching it for awhile. The effect of this control is pretty heavy-handed (especially with the greater "granularity" that's inherent with SYSM=2/3), and after playing a bit more with some of the other controls it seems as if there may be other ways of working out shadow issues with possibly a bit more finesse. If more "definition" or "edge" is all you seek though, look no further than LTMD=1.

CTLV
0=highest color saturation at edges. 3=the least, and generally softest-looking color detail. Seems fairly straight forward.

ADU

02-08-06, 04:31 PM

#1) I thought it was interesting what one can do w/o use of MID5 column when one knows a little more of what is going on with these 2170P-3 settings.Good point Jeff. Since the signal I use bypasses the digital MID controls, the 2170P-3 analog controls are basically the only ones I have for tweaking edge detail on my TV (aside from Clear Edge VM, which I've mostly ignored so far, and basic Sharpness.)Nevertheless, I wonder if it is possible we might also be getting something a little different on this depending upon if we are using SFP tube or no, or maybe there are some other differences between the different models which may be involved.Quite possibly, if not probably. Using the MID controls & VM could certainly put a different spin on what settings work best as well.

Even if the "best settings" for these edge controls follows no particular pattern from one TV/user to another (or even one input/signal to another), I think you may be able to get a better feeling for the behavior of these controls and possibly where they might benefit you by playing around with them.I think it was (with LTMD=1) LTLV at a higher value which "helped" the most with this, concerning an overall increase in "sharpness" using SYSM=3. However, although I can see where it would be beneficial in some cases, one issue I had with raising LTLV is that I didn't really *need* to increase "shadow detail".Another good point. See some of my related and somewhat similar impressions above. for all above experiments, I left user menu Sharpness slider at 31 - 30~31 is also what I have it for normal viewing with "standard or movie" pic modes.This is similar to what I refer to as "nominal" sharpness in my earlier post. Nominal sharpness to me is the setting which produces the least edge distortion, but without introducing unwanted "softening" effects either. On my TV it's around (but a few ticks below) the middle of the slider (and I've reconfigured 2170P-4/USHP to make that setting my Pro mode default.) Lately I've been using this pattern (http://www.avsforum.com/avs-vb/attachment.php?attachmentid=50979) to set that adjustment, looking, as I say, for minimum sharpening or softening distortions on the vertical lines.

Since there are other Sharpness offsets in the SM (such as 2170P-3/SHOF), the adjustment of this setting could vary a bit for different inputs/signals.

Tks for the kudos btw. Just trying to expand a bit more on Ken's already detailed info on these controls earlier in the thread.

Nitewatchman

02-08-06, 05:47 PM

If I feed 480p via component inputs, there is a limit on the detail somewhere -- the picture has lost some important texture information, and it can't be recovered.

Very interesting. Using component input -- While I do find I need to do "more" with the image processing settings for 480p DVD that is the case anywhere else -- I haven't really noticed a limit on detail. Maybe I should try to look into this in more detail<pardon the pun<g>.

Problem is, I've never been happy with the 480i from this particular player(Zenith DVD2381), including when using it with other sets, but I've allways been very happy with 480p from it.

Oh BTW Ken -- I'd responded to your PM reply and also had sent you another PM yesterday/etc with a few updates on various matters. But, the way the email notifications seem to "work"(or not work), sometimes(or maybe I managed not to click on the read recepit "OK" button for your message), I'm not sure whether you've seen those yet or not, if not hopefully they are in your PM inbox.

Nitewatchman

02-08-06, 05:52 PM

So so far, it's been kin dof tough call. Another way I'd describe the difference is SHF0=0 seems more "TV-like", and SHF0=1 seems more "monitor-like", if that means anything.

What excellent descriptions! I have never been able to put into words my thoughts about SHF0, now I have some words I can use. There was a period over about 3 months that I changed nothing else except SHF0 for various inputs. For those 3 months, I was using SHFO=0, but finally decided I liked that "flat", "tightened" response and "monitor-like" aspect of SHF0=1 better, and I've been using it ever since.

Per something Napoleon D said in another thread, rather than simply randomly flipping back and forth between the different SYSM settings while working with the other controls, I think there's probably something to be said for focusing attention on one SYSM setting at time (once you got pretty good grasp of the effect each 2170P-3 edge control has). And trying to break down the best configuration of edge controls to go with that particular SYSM setting. Then move on and repeat the process with the next SYSM setting, working in sort of hierachical fashion to narrow down the best configurations or range of potential adjustments for each (with a given input/signal).

However, I would not make the mistake of assuming that just because one particular 2170P-3 parameter does"help" with one SYSM setting, that that control should necessarily be thrown out of the mix for the other SYSM configurations. Because as mentioned above some controls seem to work with finer granularity and more subtle effect with SYSM=2/3 than they do with SYSM=1. I'm thinking particularly of F1LV here, which seems to have a much finer and potential more beneficial "defining" effect with SYSM=2/3 than with SYSM=1.

I agree. I think it's probably best to spend quite a LOT of time "equalizing" other controls with the Use of a single SYSM setting and finding what is best. In my case, concerning 480p DVD specifically - over the past ~7 months, that's involved long periods using either SYSM=2 or SYSM-3 for 480p with various other settings, and generally using Ken Tech's finding as a "base" to work from and do "fine tuning" from there.

Last night -- Given my quick "seat of pants" experiments I explained in last post on this -- I should note some of my "experiments" and what I found were not only based on that more "seat of the pants" experimenting session, but also from previous occasions when I'd also played around with those settings more extensively with "SYSM=2" or "SYSM=3", but over much longer periods of time. The info ADU has posted did however, greatly help me not only "put 2+2 toghether" concerning what those controls were actually doing so to speak, but also helped me to find better results with SYSM=3 and 480p DVD than I had been able to come across in previous experiments.

Although the "best settings" for these may follow not particular pattern of consistency from one TV/user to another (or even one input/signal to another), I think you may be able to get a better feeling for the behavior of these controls and possibly where they might benefit you by playing around with them.

Absolutely, very well put. In my case, it has turned out to be the case that Ken Tech's findings(especially the most recent) have worked out very, very well for me, and other than fine tuning the MID5 Columns, I haven't really been able to "add" any signifcant improvments to using the settings he has come up with - and, it hasn't really been necessary, either.

The only real exception to that was for 480p DVD "back" when I was using SYSM=3 for everything. Prior to the info in his Post #707, I also wished I'd found the time to experiment more with SYSM=2 for other inputs besides DVD, as well as a couple of other things in 3dcomb and 2101-3 pertaining to "SD" via DRC from S-video or internal "RF" tuner. I also "used" to occasionally use a pic mode set up to use a MID5 Column with all "zero's" for all inputs(given use of SYSM=3, and 2170P-3 settings very similar to those in brackets in my earlier post).

Besides, it's "fun" to play around with, when I have the time. Perhaps Especially when you already have a good set of "settings" set up in one Pic mode that are working well for you and can just play around with trying to make improvments in another, "experimental" pic mode you have set up. Ok, sure, I'm a little crazy to enjoy such leisure activities ...

(and I've reconfigured 2170P-4/USHP to make that setting my Pro mode default.)

Just a FYI -- Even though it's shown in the XBR960 servicecode listings, there is no 2170P-4 UPIC~UTMP, on my KD34XBR960. I have 2170P-4/#6 SHUO, then next is 2170P-4/#7 RYR. "GAMM is control #11", for instance, and so on.

UPIC~UTMP are However present in either the "QM" or "QT"(I'll have to look to be sure which one) "undocumented" sections of SM -- I do recall looking at UPIC at one point to see if I could see what it does, and it seems to behave quite "differently" than is the case with your XBR800. If I recall correctly, it has seperate values per pic mode, and instead of actually "doing anything", it merely contains info on the Set defaults for the "picture" slider for the various picture modes -- On the XBR960, there are "little dots" on the sliders which correspond to where the slider was for the set defaults for the diffierent pic modes, and perhaps that may be what UPIC~UTMP are used for on my set -- to place those little "dots".

justsc

02-08-06, 06:31 PM

KenTech

02-09-06, 01:18 PM

FWIW - With ADU's excellent info on the relevant P2170-3 settings printed out and brought along -- Played around with these in a "seat of the pants" manner a little last night With KD34XBR960 and 480p DVD(Zenith DVD2381) via component input. Difficult to say how much of this, if any is "specific" to my DVD player.A couple more thoughts on the 2170P-3 edge controls...
These entire posts were interesting to me because they describe some of what I have observed in different words. I think we're verging on discussing "flavor," here! In particular . . .
So so far, it's been kin dof tough call. Another way I'd describe the difference is SHF0=0 seems more "TV-like", and SHF0=1 seems more "monitor-like", if that means anything.Yep, this works for me, too. I am trying to avoid the "TV" look and going for something more film-like, more photographic. One of the signatures of the TV look, I think, is the smoothing of fine texture along with oversharpening of medium-fine detail. I generally screen these parameters for their effect on fine texture, and if it is attenuated for some settings, then those settings are verboten on my set. SHF0 is one of those, and so I have abandoned it, leaving it at 1.#14 MVYL - 3
#15 MVYE - 1The trouble with vertical sharpening is that its scope of action (distance) is always a minimum of one scan line, and on 480i/p material, that's not too fine! My DVD player has a little built-in, so I don't dare add more. Further, if one is trying to rescue fairly soft video, a bit of vertical sharpening "balances out" whatever horizintal sharpening you may be adding. Many good video cameras and both HD and SD broadcasts show an irreducable vertical sharpening, too. But I generally don't like its effect on good source video. What you've listed is a very small amount, however, and, believe it or not, there is still a just-detectable amount still in effect with MVYE=0 if MVYL=3. MVYL has to be at 0 to shut it off completely.

KenTech

02-09-06, 01:52 PM

So far SYSM=3 seems to be giving a bit of a biased look to edges on my TV that's a bit uncomfortable to my eyes.I have come to think of it in audio tone-control terms: SYSM=3 raises high frequencies, but as a *plateau,* much like the standard trebel control on a stereo. When cranked up, there is a response-curve ramp-up at some frequency, and then it levels off for higher frequencies. Thus, very fine detail is not boosted proportionately to fine detail but equal to it. SYSM=2, however, seems to me more like a tilting ramp upwards on the high-frequency response curve. When cranked above "neutral", medium detail is enhanced a bit, fine detail a lot more, and the finest textures most of all. So it has a very different "look" from SYSM=3. At a given SYSM=3 Sharpness boost setting, one could add those finest textures, I guess, by having a corresponding MID5 setting for MHYE. Trouble is, MHYE is *fixed* and SYSM is variable. So it would appear balanced only at one Sharpness setting -- not a very versatile use of Sharpness.

SYSM=0/1 is scary to me! Its scope is quite large (distance of action), and it creates weird double-edge ghosts when cranked down. I have a feeling it was intended for heroic image-rescue, but I can't match it up with anything I use as a source.F1LV: Although this generates fairly coarse looking "noise" with SYSM=1, as mentioned above it seems to have some greater potential with SYSM=2/3, and may have some value in terms of "elevating" shadow information so it's a bit less strenuous on the eyes.'Scuse a little nitpicking, here: none of these really "generate" noise; I think you meant "accentuate" for what noise that's already present. I'm one of those lucky enough to have impeccable analog-cable service, and so I see very little noise as I try to optimize these SD broadcasts. My greatest source of noise is from grainy DVDs, and they abound! E.g. The Matrix trilogy is *very* sharp, but the film grain is clearly visible on smooth textures. In-player 3DNR is very useful, here.CTLV: 0=highest color saturation at edges. 3=the least, and generally softest-looking color detail. Seems fairly straight forward.Very clear description. Motivates me to experiment, too. Gotta keep an open mind!

KenTech

02-09-06, 02:02 PM

Very interesting. Using component input -- While I do find I need to do "more" with the image processing settings for 480p DVD that is the case anywhere else -- I haven't really noticed a limit on detail. Maybe I should try to look into this in more detail<pardon the pun<g>.

Problem is, I've never been happy with the 480i from this particular player(Zenith DVD2381), including when using it with other sets, but I've allways been very happy with 480p from it.What is particularly frustrating is that, absent a few $$$ pieces of test equipment, none of us can really tell whether a particular artifact or characteristis is caused by the source (say, DVD) or the TV. My Panasonic S97 is well-known to have stunning component output, and I have read reviews of other players who performs far less well on component than HDMI. Component performance seems to vary a lot!

Another example: I don't like the *very* soft look to my player's upscaling to 1080i over HDMI -- on my TV, anyway. I have a friend who just bought the Sony D50XBR1 SXRD 1080i RP unit. He bought the same player I have. His component desplay is good, but is forced to upcscale to 1080p. *His* 1080i over HDMI is great from this player! Uh . . . sort of implicates the 1080i/HDMI capabilities of our beloved TVs, eh? Go figure! 1080i digital input on a (purported) 1080i-native TV. Wonder where the fault lies in the processing chain.

KenTech

02-09-06, 02:08 PM

Lately I've been using this pattern (http://www.avsforum.com/avs-vb/attachment.php?attachmentid=50979) to set that adjustment, looking, as I say, for minimum sharpening or softening distortions on the vertical lines.It's good to double-check with edges that are not part of 1-2-pixel lines, too, since a 1-pixel line can cause an abnormal ringing that almost never appears in real program material. The "Sharpness" pattern on the AVIA disk is particularly revealing, with its crossed 1, 2, 3, and 4-pixel lines, plus white and black blocks that are devoid of antialiasing. (The circle and text are antialiased.)

Of further help on the same disk are the various sub-menus. On the big colorful main menu screen, almost everything *is* antialiased (softened) slightly, but the sub-menus are hard-edged graphics. Getting those images free of artifacts -- or even the main menu -- is quite an acid test!

KenTech

02-09-06, 02:28 PM

This confuses me a bit. And maybe I've been misusing the terminology. I have been equating edge distortion with noisiness or harshness. I found on my set that SYSM=3 was the smoothest, least harsh setting. I find now that SYSM=2 or 1 seems too sharp or too harsh.We're all reduced to metaphor and simile. Next thing you know, we'll all be referring to "airy" detail and "muddy" lower-upper-mid-bass. Er, I mean -mid-detail. (Sorry. I have worked in the high-end audio industry. Hope nobody holds that against me!)

I make all of my comments about edges from close observations, ignoring the picture -- and generally using known test patterns, the AVIA disk being especially valuable. Beyond that, there are certain frames on Monsters, Inc and other good DVDs that are revealing of texture and diagonal/curving fine lines (Sully's hair!). I see ringing and overshoot mostly from this close-up vantage point.

But when I sit down to watch a known-good DVD at a reasonable distance, paying attention to the overall fidelity of the image, I see characteristics that are best described as "smooth" or "harsh" or "grainy" or, well, *ugly* or *beautiful.* The eye/brain combination is integrating everything and "locking" onto the image. The 11pm newscasts here all all different from each other, but one is *very* fine. I can look at the faces of these (perfectly-lighted) newsreaders, and they might jump out as nearly 3D and authentic, or have lots of texture and detail but are *wrong* somehow. That's how I've come to reject certain of the filters: They do something that theoretically *should* help (according to test-pattern observations) but something's not right when viewing real video.

So, for me, ringing, edge-distortion, overshoot -- those are all thing I can see and identify only up close. What effect they have on real video I describe as harsh, unauthentic, finely-detailed, balanced, "TV" look, etc. Occasionally something that looks like (e.g.) overshoot on a test pattern really helps real-world viewing. I think that's the "compensation" factor coming into play. Describing it here is kinda tough.

SurfingMatt27

02-09-06, 04:51 PM

ADU

02-09-06, 04:55 PM

Just a FYI -- Even though it's shown in the XBR960 servicecode listings, there is no 2170P-4 UPIC~UTMP, on my KD34XBR960. I have 2170P-4/#6 SHUO, then next is 2170P-4/#7 RYR. "GAMM is control #11", for instance, and so on.Interesting.UPIC~UTMP are However present in either the "QM" or "QT"(I'll have to look to be sure which one) "undocumented" sections of SM -- I do recall looking at UPIC at one point to see if I could see what it does, and it seems to behave quite "differently" than is the case with your XBR800. If I recall correctly, it has seperate values per pic mode, and instead of actually "doing anything", it merely contains info on the Set defaults for the "picture" slider for the various picture modes -- On the XBR960, there are "little dots" on the sliders which correspond to where the slider was for the set defaults for the diffierent pic modes, and perhaps that may be what UPIC~UTMP are used for on my set -- to place those little "dots".Actually that sounds pretty much the same as on the XBR800. On the XBR800, 2170P-4/UPIC through UTMP configure the default positions of the sliders and temp settings for each picture mode in the User Menu. For slider controls, the numbers correspond to ticks, with 0 representing the first tick, and 31 representing 32 ticks. If, for example, you wanted the Brightness slider to default to say 27 ticks instead of the usual 32 in Pro mode, you could change that by making UBRT=26. Same for Sharpness, Contrast, etc.

So why use this approach to reconfigure Sharpness? Well, because the 2170P-3/SHOF sharpness offset is already at it's lowest setting for the DVI signal/input I'm using on my XBR800, and Sharpness is still not "optimal" w/o further reducing the slider in the User menu.

Perhaps I should have mentioned these adjustments earlier when you were having some of your Brightness dilemmas. But I figured you were probably already aware of this, and wanting to find a way to get Brightness in the SM to work with the existing slider configs in the User Menu.

justsc

02-09-06, 05:55 PM

I've been tinkering with those MIDE-5 tables last year,and aftyer al the tweaking i come to the conclusion that simply turining them all offto 0 is the best solution to a better picture quality.

No matter how many you have on or little as possible, you r still adding stuff to the picture that just is'nt there.You can't expect a stock honda to beat a porsche ;) lol.

Maybe it's me, but i find the less enhancements the better the PQ looks and more cleaner iand sharper it looks.With edge enhancements on it agve me a sharp picture but it was unnaturaly sharp kind of like a cookie cutter sharpness appearance.

I would rather have my tv do no processing and just feed the signal to the screen,any EE will be from the source material not my tv set.

The PQ looks razor sharp without them EE enhancemnts,so they really are'nt neccessary it just makes the PQ unnatural.With EE off the PQ is naturally sharp they way film should be.

just my 2 cents.
Matt,

I really do understand where you're coming from.

I'm just wondering - don't you think there's a chance that there is some detail that may be there, but not yet visible? Kinda like when it's a dark scene and by turning up brightness you find that there is detail you didn't even know was there? So maybe you bump up your brightness setting just a little so you can see what the director intended for you to see.

Now that's not adding anything to the signal, it's just "bringing out" or "exposing" the detail.

My interest is in exposing as much of the "intended" detail as I can. And I see there may be more than one way of doing that. And it's important to understand that our tv sets, as they've been manufactured and set-up at the factory, may not come to us in such a condition that we can view all this intended wonderful detail - that's why we calibrate right?

The first task is to explore the settings to discover those that "expose" those details without adding things in.

The second comes from having a full understanding of what the tv's s/w and h/w is doing to the signals. We might come to learn that there's circuits or code that actually "dial back" or attenuate details within the signal so that something must be done to "dial" those details back in. But maybe there's no suitable adjustment that does this, and the signal has to be massaged so that, in the viewers perception, from the appropriate viewing distance, this once attenuated "detail" is now exposed. And it could very well be that the only way to see this detail is by "adding in" something, possibly a little edge enhancement, so we can "perceive" what would have been there.

Either approach is simply to bring out what we were supposed to see. This is why I don't exclude the possibility of some edge enhancement.

Nitewatchman

02-09-06, 06:32 PM

Perhaps I should have mentioned these adjustments earlier when you were having some of your Brightness dilemmas. But I figured you were probably already aware of this, and wanting to find a way to get Brightness in the SM to work with the existing slider configs in the User Menu.

1st off, please stay with me here until the next section and attached screenshot ....

No, there was no need for you to mention those, but thanks. I'm afraid you are still misunderstanding the issue I had with getting the "picture" slider centered, instead of using it at "24" which worked just perfectly, all I was trying to do was to get the picture slider to read "31"(32 ticks) and achieve the same results as was the case with "24"(25 ticks). I didn't have any problem "issues" whatsoever with brightness, or "contrast" or balancing those(or the "sharpness slider either for that matter) among inputs/etc, just wanted to center the Picture slider at "31".

I perhaps used the wrong terminology --- In those posts I was talking about "brightness" as it relates to "white balance" and "contrast", NOT brightness as it refers to the "brightness" slider, or "black levels", the latter I took care of long ago with 2170P-3 UBOF and 2170P-1 SBRT.

It's all water under the bridge, now though, as I've already explained I was able to figure it out using a lower value for 2170P4 SPIO (and, long ago I had already used 2170P-4/SPIC to balance contrast among inputs, although I touched those up a bit anyway when I was working on this). Sorry, I probably did not do a good job explaining that, but I can't think of a better way to say it than I already have(along the lines of what Ken said, sometimes it is awfully difficult for me to talk about these issues and know that my comments are "understandable" by others), and I'm afraid going into that again and repeating it again won't be useful.

Interesting.Actually that sounds pretty much the same as on the XBR800. On the XBR800, 2170P-4/UPIC through UTMP configure the default positions of the sliders and temp settings for each picture mode in the User Menu. For slider controls, the numbers correspond to ticks, with 0 representing the first tick, and 31 representing 32 ticks. If, for example, you wanted the Brightness slider to default to say 27 ticks instead of the usual 32 in Pro mode, you could change that by making UBRT=26. Same for Sharpness, Contrast, etc.

So why use this approach to reconfigure Sharpness? Well, because the 2170P-3/SHOF sharpness offset is already at it's lowest setting for the DVI signal/input I'm using on my XBR800, and Sharpness is still not "optimal" w/o further reducing the slider in the User menu.
.

Interesting. I don't think it's quite working the same, at least with "mode memory" turned on on 34XBR960 in as "useful" a manner.

Fom what I recall from checking this, On KD34XBR960, QM UPIC~UTMP do nothing to actually "configure" the sliders(update: except to allow you to hit "reset" button to return to a certian set of default values as Ken mentions below/End update) , other than if you use the little "dots" shown in the slider menu as a reference to return the values for the sliders back to factory defaults for any given pic mode, or if you change the values for UPIC~UTMP, your "new" values would correspond to new posistions for these "little dots", which otherwise mark the factory defaults for the sliders for the different pic modes.

They only SHOW you [update again : or let you use the reset button to return to a certian set of "defaults" - either the factory, or values you've set up ; end update], in the "video" user menu what those factory defaults were, with little "dots". They don't effect what happens to the picture, you do that by changing the slider yourself, which you can do on KD34XBR960 in a Input specific manner(in other words you can use different slider values for different inputs(or pic modes for that matter with "mode memory/On) by turning "mode memory" on in user menu, or you can use the same slider "settings" for all inputs by turning "mode memory" to : off.

:update :

Again, as noted in my reply to Ken farther below - sorry ADU for my "misunderstanding" of what you were talking about on this -- Entirely my fault, but, rather than editing them except for these updates as noted, I think I'll go ahead and leave my comments as they were in this post, if nothing other than a demonstration of how easy it can be to "misread", or "misinterpet" things - again, in this case, entirely MY fault.

Because of the availability "mode memory: On" function and different slider settings(or other input/scan rate/etc. offset controls available in SM - such as "UBOF") for all the different sources I use, I've never really thought of a reason to make use of the reset button, or to define a certian set of "defaults" with QM UPIC~UTMP settings for the User mode "sliders" ... Although, thinking about it more, it could perhaps be useful for me in order to more quickly "switch" to a "default" set of slider settings when "checking" things, or, perhaps to set the sliders quickly back to my "normal" mid range settings after adjusting to preference to compensate for "differences in programming material."

:end update

Perhaps a picture is worth 1,000 words. So, attached is screenshot "sliderdots.jpg"(pardon the quality of the shot and the reflection from a lamp in the room) -- I've adjusted the sliders(this is NOT what I use them at, LOL) so you can see all the "little white dots" which UPIC~UTMP correspond to for "factory Mode" for each slider. For instance -- UBRT for this Pic mode = 32, and that's right where the little white dot is for the "brightness" slider that shows the "brightness" slider factory default for this pic mode. The "brightness" slider would JUST "cover up" the little white dot if I had the slider set at "31" as I normally do. If I change QM/UBRT to "45" for this pic mode, the little white dot will move up to correspond with the "44" value for the slider itself, but, it doesn't effect the actual value of the slider itself in any way shape or form -- nothing else will change except the posistion of that little white dot. Perhaps it may behave differently however if I wasn't using "mode memory" set to "on" - I didn't try that.

KenTech

02-09-06, 08:38 PM

No matter how many you have on or little as possible, you r still adding stuff to the picture that just is'nt there.Actually you are increrasing the level of stuff that *is* there, but if you don't like the look, then by all means leave them all at zero. You have to please only yourself in this matter. But I wouldn't advise that this is somehow a "correct" setting.Maybe it's me, but i find the less enhancements the better the PQ looks and more cleaner iand sharper it looks.With edge enhancements on it agve me a sharp picture but it was unnaturaly sharp kind of like a cookie cutter sharpness appearance.Then you really had too much of something added in, and maybe it came from *before* the MID processing. If the source is 480i, it could be the 2103 chip, which adds a horrible "TV-like" edginess if left untamed.

KenTech

02-09-06, 08:45 PM

They only SHOW you, in the "video" user menu what those factory defaults were, with little "dots". They don't effect what happens to the picture, you do that by changing the slider yourselfDoesn't clicking the Reset button on the remote (or is it in the user menus?) move the sliders to the "default" positions all at once? Then these presets establish what those "factory" settings will be. On my 36XS955, it's moot, as those "dots" are established by read-only settings in one of the "blue" SM groups.

Nitewatchman

02-09-06, 09:11 PM

Doesn't clicking the Reset button on the remote (or is it in the user menus?) move the sliders to the "default" positions all at once?

Yes, and although I had a hard time understanding ( my fault Not his) I realize now that's the function ADU is likely referring to.

I suppose that would be useful if either/or : #1). you couldn't set up different slider settings for different inputs if Mode memory = on isn't available on your set model, or 2). if you have chosen to use "mode memory" off, or 3). also to return to "preferred" values in any case when experimenting or making certian adjustments in SM relating to the slider in question. OR Perhaps most importantly, 4). To quickly return to your "preferred" settings after making adjustments to the sliders to compensate for differences in program material.

Then these presets establish what those "factory" settings will be. On my 36XS955, it's moot, as those "dots" are established by read-only settings in one of the "blue" SM groups.

On my KD34XBR960 It's in the "blue" QM section - For instance, UBRT is "QM" Control #15. It's not read only, though, I just confirmed this by Changing "UBRT" for "standard" pic mode from "32" to "46" and writing the change ... The white dot position changed for the brightness slider, and it stayed that way(at 46) after I powered the set off, and powered it back up outside of SM.

Note that the white dot didn't change when I was "viewing" the user "video" settings while making the change, but after I wrote the change, exited the User menu, then reentered the user menu, it had changed to reflect my new value.

SurfingMatt27

02-09-06, 09:42 PM

Matt,

I really do understand where you're coming from.

I'm just wondering - don't you think there's a chance that there is some detail that may be there, but not yet visible? Kinda like when it's a dark scene and by turning up brightness you find that there is detail you didn't even know was there? So maybe you bump up your brightness setting just a little so you can see what the director intended for you to see.

Now that's not adding anything to the signal, it's just "bringing out" or "exposing" the detail.

My interest is in exposing as much of the "intended" detail as I can. And I see there may be more than one way of doing that. And it's important to understand that our tv sets, as they've been manufactured and set-up at the factory, may not come to us in such a condition that we can view all this intended wonderful detail - that's why we calibrate right?

The first task is to explore the settings to discover those that "expose" those details without adding things in.

The second comes from having a full understanding of what the tv's s/w and h/w is doing to the signals. We might come to learn that there's circuits or code that actually "dial back" or attenuate details within the signal so that something must be done to "dial" those details back in. But maybe there's no suitable adjustment that does this, and the signal has to be massaged so that, in the viewers perception, from the appropriate viewing distance, this once attenuated "detail" is now exposed. And it could very well be that the only way to see this detail is by "adding in" something, possibly a little edge enhancement, so we can "perceive" what would have been there.

Either approach is simply to bring out what we were supposed to see. This is why I don't exclude the possibility of some edge enhancement.

Steve i understand what your getting at but you have to understand about the source material as well.

DVD for example, it's not going to look any sharper than what i see,it's as good as it gets, if your expecting ultra sharp pictures that detail just is'nt there..By that thought your thinking about HDTV and you are wanting HD resolution from a DVD that's only limited to 480i, it's just not possible to add stuff that's not there.

I've fiddled with those settings a while back on the MIDE5 table when i first got my sony in 2004, and the only sharpness enhancemnt i have seen from using that MIDE5 table is just giving the Picture a cookie cutter appearance.It tricks the eyes into being sharper by making the outlines thicker,but in reality it's just making the picture unnaturely sharp and grainy.

I'm satisfied with no enhancemnts being done to the picture, and DVD on my set is being produced as good as it can possibly be,it's not going to get any sharper,by then your just thinking HD when it's only a DVD resolution

I guess to get to the point what i'm saying is that your expecting too much from DVD.DVD is not HD, people are expecting HD resolutions from DVD when the detail just is'nt there in the first place.

KenTech

02-09-06, 10:01 PM

YOn my KD34XBR960 It's in the "blue" QM section - For instance, UBRT is "QM" Control #15. It's not read only, though, I just confirmed this by Changing "UBRT" for "standard" pic mode from "32" to "46" and writing the change ... The white dot position changed for the brightness slider, and it stayed that way(at 46) after I powered the set off, and powered it back up outside of SM.Well, well, if this is true for my set too, I will be a very happy camper, as I am always returning my menu sliders to "center" when tuning in new program material. I thought I tried this last spring. I will try again and report back!

ADU

02-10-06, 12:09 AM

Hmm. Not familiar with the mode memory feature, so perhaps that's something that was added after the XBR800s. And perhaps that's why the UPIC - UTMP controls are in different places in the SM. The User Menus aren't very smart on the XBR800. From what I gather from your remarks above, it sounds like my TV operates the same as yours with mode memory OFF.Doesn't clicking the Reset button on the remote (or is it in the user menus?) move the sliders to the "default" positions all at once?I realize now that's the function ADU is likely referring to. Precisely. So far, Sharpness (USHP) is the only User control I've reconfigured like this, because as mentioned before, it simply wasn't possible to drop the value low enough in the SM to fit my prefered level to the default Pro/USHP slider setting of 32 ticks (or dots as the case may be). if this is true for my set too, I will be a very happy camper, as I am always returning my menu sliders to "center" when tuning in new program material.I'd give it a try Ken. As Jeff mentions, you may not see any change to the picture while adjusting UPIC - UTMP in the SM. But when you press RESET on the remote, the User Menu sliders should snap to the new positions. Sorry for doing such a lousy job of explaining this earlier. :)

Nitewatchman

02-10-06, 12:57 AM

Hmm. Not familiar with the mode memory feature, so perhaps that's something that was added after the XBR800s.

I think I'd read somewhere(I think it was in a review of the XBR960 in one of the A/V mags) it is a new feature they added to at least some of 2004 "line" of sets -- not sure which all models are included. But, maybe the XBR910's have it as well, don't know.

. Sorry for doing such a lousy job of explaining this earlier. :)

Again, it was MY misunderstanding - My fault, and again my apologies -- Thought you did a fine job after Ken mentioned the reset button and I realized what you were talking about .... If you would have mentioned the "reset" button, what you were talking about would have finally made it "into" my thick skull ... <g>

Would be interested if Ken or anyone else finds a setting on this to be able to control the user menu "ClearEdge VM" choice when you hit the reset button. If there is one in SM somewhere, I couldn't seem to find it, so it reverts to factory defaults for the "clear edge VM" settings for different pic modes when you press reset button : defaults to "high" Clearedge VM setting for vivid, and standard(I think, might be remembering that one wrong), and "low" for Movie Pic mode, and "off" for "pro", which makes it less useful for me given my customized pic modes, and that I have "standard" for normal viewing.

It also didn't work for "mode memory=on" to turn "mode memory=off" first and set these settings just one time for all inputs, and have the my settings for the "little dots"(centered) actually apply to the sliders for all inputs given use of reset button/etc after I turned "mode memory=On" back on. Except for "pro" pic mode of course, which already had centered values with the factory defaults for all inputs.

So, with Mode memory set to "on", I had to go through every input and change QM/UPIC~UTMP for all inputs+pic modes to my preferred values.

I did learn something useful out of this that I'd never noticed before, as I'm not using HDMI currently, and normally have the HDMI input ("video 7") set up as "skipped" in User menu : I found I can "look at"(or modify) the HDMI specific SM columns when I'm switched to "video 7"(the HDMI input on this set). Where there are scan rate specific settings, looks like It's only 480i I can "look" at. Even though there's not a real "signal" there, the "HDMI" info screen in "HDMI Info" section of SM shows "480i 4x3". Checked a few values for HDMI specific columns such as "CBGN~YGN" settings for HDMI just to make sure. So, I can use that to some degree to set up HDMI specific settings for customizing pic modes, as well as (but just for 480i apparently) the 480i HDMI specific P2170-3 image processing settings - This is helpful as that way I won't have as much "work" to do when I add a HDMI device in the future.

Funny thing -- -- I recall trying it once back when I first got the set just to see if the button worked, but this time I had to go into the owner's manual to help me "figure out" how the reset button worked (LOL) -- It says you have to be in the "video" menu to use the reset button to reset the picture settings to "factory defaults" ... Puzzling, since I tried selecting the "video" menu Icon from the menu (in which case the video menu appears), but nothing happened when I hit "reset" ... So, reading that DID however(oddly enough) make me realize I had to actually "move over" into the "video menu settings" themselves in order for it to work ... A wise "fumble fingers" precaution I suppose, nevertheless, I'd bet they get a few calls about that<g> ...

ADU

02-10-06, 01:20 AM

Would be interested if Ken or anyone else finds a setting on this to be able to control the user menu "ClearEdge VM" choice when you hit the reset button. If there is one in SM somewhere, I couldn't seem to find it, so it reverts to factory defaults for the "clear edge VM" settings for different pic modes when you press reset button : defaults to "high" Clearedge VM setting for vivid, and standard(I think, might be remembering that one wrong), and "low" for Movie Pic mode, and "off" for "pro", which makes it less useful for me given my customized pic modes, and that I have "standard" for normal viewing.FWIW, on the XBR800, it's 2170P-3/VM. 3=High, 2=Med, 1=Low, 0=Off. This and 2170P-4/UPIC - UTMP are strictly picture mode sensitive on the XBR800. The intensities of the Clear Edge VM settings (High, Medium, Low) can be individually customized for each picture mode as well with 2170P-3/VMH, VMM, and VML.

ADU

02-10-06, 01:47 AM

KenTech

02-10-06, 02:40 AM

It's not read only, though, I just confirmed this by Changing "UBRT" for "standard" pic mode from "32" to "46" and writing the change ... The white dot position changed for the brightness slider, and it stayed that way(at 46) after I powered the set off, and powered it back up outside of SM.This evening I duplicated this process exactly on my 36XS955, and no-go. It won't remember the new setting. It does a WRITE, but nothing is really written, apparently. Rats!

ADU

02-10-06, 03:05 AM

Hmm, the new value isn't retained by the SM, and pressing RESET on the remote with the User Menu/picture mode highlighted does nothing?

Napoleon D

02-10-06, 11:20 AM

I have since switched on back to SYSM=2. My ISF calibrator had set it to SYSM=3. As much as he knows, this i ultimiately think is the incorrect setting. I've switched back and forth on this subject, and it could be a matter of taste. But SYSM is probably the most major setting in the service menu, at least it has the most dramatic effect on how your image will look clarity/sharpness-wise.

I speak only from my experience with 480P dvd/component content, but i'm sure this applies elsewhere just the same.

SYSM=3 produces a VERY smooth image. It is an image so smooth, that raising sharpness really has no meaningful or useful effect on the picture. The image would look terrific if you were watching it in a store, but to actually sit down and watch a movie, or anything for a period longer the 27 seconds... not going to work very well. The picture, prior to adding edge-enhancment, is so smooth that it is VERY VERY difficult for the eyes to focus on. I don't know terminology too well, but i would assume undershooting is the correct word for what happens with the image. While the picture looks free of noise, and almost glossy, there is a severe lack of fine detail to focus on. I'm not saying the picture is blurry, it is just so soft that your eyes cannot latch on to any part of the image to focus on. The only way to fix this is to add some sort of artificial edge enhancement in the service menu - using mide, or perhaps a small degree of VM. While you will come out with a nice looking image from this adjustment (and one you can actually focus on), you will now notice there are bits of noise and artififacts that collect around certain images, almost like snow collecting on the sides of a tree. This is very much a tradeoff, you will get a smooth image, but an image with notceable edge-enhancement and ultimately unnatural.

SYSM=2.... This image is not QUITE as smooth or as computer screen-like as SYSM 3. But, fine detail is brought more cleanly and naturally to the surface. When you see a closeup of a person, you don't have to strain to see details, it comes straight out. The advantage with SYSM=2? You can remove every single one of those god-awful (in my opinion) edge enhancements in the service menu. Once you do all this, the only thing you need worry about is having the sharpness in the user menu at the right setting. By this point, you've created an image so natural and clear (and one that is still VERY easy to focus on) that the only noise you will get is from the sharpness control, or from any original artifacting from the actual source. This creates much more of a realistic image, as even film has a small degree of noise to it itself. IMO a movie image is not supposed to look overly smooth and processed, but ever-so slightly rough in certain places. This is what SYSM-2 creates. Your eyes will also note the positive difference that you are seeing an image far more natural and clear, without that extra edge "crap" blocking it. Again, all in my opinion.

(SYSM=1 is basically a lesser version of 2 (from what i've seen) i have no idea what it's designed for other than antenna tv. )

Basically i would rather strip as much as possible off an image to make it cleaner (SYSM-2) rather than add fake layers to the image to make it stand out more (SYSM-3). The good focus and convergence should be doing the core work to make a good image.

Does this all make sense?

By the way, in my short experience, i've noticed that all those MIDE, VM, LTLV etc. settings are no longer needed when SYSM is set on 2. Nor do i think they should be on the screen to begin with in any situation. I have every MIDE, VM, LV's, RSVP's etc set to zero (exception of PROV, LTMD, and SHFO) and have the best picture because of it.

Actually, Chad Bilheimer did my ISF. I'm sure you all have heard about him by now, as he is probably the best ISF calibrator out there, based on his testimonials, and based on how my set is looking. The reason he was the best, is because he cared so much about other people enjoying displays as much as he would, the level of perfection he went for was unreal. At any rate, he confirmed shutting all these above settings off. The only disagreement i had was that he set SYSM on 3 (for compnt dvd), with a MIDE setting to suppliment it. I ultimately just changed SYSM to 2, turned off the MIDE and couldn't be happier with his work, and how the display looks.

KenTech

02-10-06, 11:54 AM

Hmm, the new value isn't retained by the SM, and pressing RESET on the remote with the User Menu/picture mode highlighted does nothing?I deduced that nothing was retained because the positions of the "white dots" hadn't changed upon re-entry into the user menus. Pressing Reset means I have a bunch of work to do restoring my defaults to all four Picture modes. AFAIK, those dots are the positions specified by the user-default parameters, and they didn't move.

KenTech

02-10-06, 12:31 PM

I've switched back and forth on this subject, and it could be a matter of taste. But SYSM is probably the most major setting in the service menu, at least it has the most dramatic effect on how your image will look clarity/sharpness-wise.The experience you detailed in this post exactly parallels mine; we just use different words to express it -- that is, if one of us catches his own typos! I completely obscured the main point of my post #955 by leaving out the numeral "3." Here is a corrected restatement of the point:

"I have come to think of it in audio tone-control terms: SYSM=3 raises high frequencies, but as a *plateau,* much like the standard trebel control on a stereo. When cranked up, there is a response-curve ramp-up at some frequency, and then it levels off for higher frequencies. Thus, very fine detail is not boosted proportionately to fine detail but equal to it. SYSM=2, however, seems to me more like a tilting ramp upwards on the high-frequency response curve. When cranked above "neutral", medium detail is enhanced a bit, fine detail a lot more, and the finest textures most of all. So it has a very different "look" from SYSM=3. At a given SYSM=3 Sharpness boost setting, one could add those finest textures, I guess, by having a corresponding MID5 setting for MHYE. Trouble is, MHYE is *fixed* and SYSM is variable. So it would appear balanced only at one Sharpness setting -- not a very versatile use of Sharpness."

***********
I'm sure this is not technically accurate in implying how these effects are implemented in the hardware. After all, SYSM=2 has a time-domain shift so edge enhancement can be added *before* an edge. That's why the whole image shifts to the right when SYSM is changed to 2 from 3. (And that's a clue for the evaluation of other parameters as well. 2170P-3/SHF0, for example.)

If I were to infer from the evidence a major factor that distinguishes SYSM=2 from 3, it is that 3 seems to be merely a frequency boost in the high end of the video spectrum (and a plateau at that), and so it cannot add an effect *before* an edge occurs. SYSM=2 has no such limitation: The effect is derived from the signal in real-time, the unaltered signal is delayed slightly, and then the "effect" is added on top of the *delayed* original signal. Much more sophisticated, and it mimics true unsharp-masking in a digital photograph, where each pixel's brightness is derived, not only from its own original value, but from the value of pixels to the left, up, down, and right of itself. This is easy for a static image. Our TV are scanned, however, and "to the left" means "before" in time. Proposed bottom line: SYSM=2 mimics more accurately an actual *optical* correction and entrains the eye for focus much more successfully.

(Your recurring point about "capturing" the eye for focus without strain is extremely well-taken! This is likely the case for oversharpening really bad video and then viewing it a suitable distance so the eye/brain can work with the video on *its* biological terms.)

Nitewatchman

02-10-06, 12:52 PM

FWIW, on the XBR800, it's 2170P-3/VM. 3=high, 2=med, 1=low, 0=off. This and 2170P-4/UPIC - UTMP are strictly picture mode sensitive on the XBR800. The intensities of the Clear Edge VM settings (High, Medium, Low) can be individually customized for each picture mode as well with 2170P-3/VMH, VMM, and VML.

On my XBR960, 2170P-3/VM is set at "0" for all pic modes, I use customized(lower than factory defaults) settings for VMH/VMM/VML per Ken tech's findings. None of These have any effect on the behavior that occurs with the "reset" button. The sliders move to my values I have set up for UPIC~UTMP when I press the "reset button", but the "Clear Edge VM" setting in the same "video settings" user menu still defaults to it's "factory" setting(either High, Medium, Low or off) for each pic mode when I press the button. Which especially doesn't work for me for my "customized" Vivid, Standard, and "movie" pic modes - In most cases, I use the "medium" ClearEdge VM setting for all of those. So, I'm looking for a setting which will allow me to control which Clear Edge VM setting each pic mode "defaults" to when I press the reset button when I'm in the "video" menu .. Just like setting UTMP=1 causes the "color temp" setting to use the "neutral" setting on my set. (0=warm, and 2=Cool).

[edited]On my set(34XBR960), looking at it more, I'm not sure what P2170P-3/VM does. I had thought it set the "intensity" of SVM when you choose the "cleardedge = Off" choice in user menu, if so, I'd want that one at "0" so it is actually "off" when I choose Off. However, It only allows "values" of 0-3. I'd thought, although it would make more sense for "off" just to mean "off" and not need any sort of "value" involved or SM control for it --- I'd thought if it were for an "off" setting, it might just allow 0-3 to only allow for the most "subtle" of SVM intesities if it were to correspond to the "off" menu choice. However, the "0-15" settings that are available for all all the others(VMH/VMM/VML), as well as the servicecode listing defaults(which match for pic mode the "high/low/off" defaults I'm getting per pic mode per the labels in user menu "clearedge VM choices) would lead one to assume it should "work" as it does for ADU's XBR800, although that is not what I'm getting according to the "low/Medium/High" "labels" in the user menu it's defaulting to regardless of how I have the 2170P-3/VM value set. [end edit]
[end edit]

If I weren't utilizing VM and VMH/VMM/VML(for high, medium, low settings) I suppose I could set those all at "zero" and then VM would be "off" no matter what the factory "reset" value for the "clearedge VM" user menu choice" for any pic mode actually is. But, I am using them, the VMM and VML settings specifically, I don't use VMH(High), except when testing+experimenting with the 2170P-3 VM "shaping" paramenters(VMCR~VMDL)

BTW, in case anyone is interested, I didn't throughly check this, but I don't think the "advanced" video menu settings (where you choose, or customize different DRC "pallete's" or chose a different "color axis" (for different settings for RYR~GYB/etc), are effected by the "reset" button.

Nitewatchman

02-10-06, 01:07 PM

I deduced that nothing was retained because the positions of the "white dots" hadn't changed upon re-entry into the user menus. Pressing Reset means I have a bunch of work to do restoring my defaults to all four Picture modes. AFAIK, those dots are the positions specified by the user-default parameters, and they didn't move.

Interesting, I would have thought it would probably work on the XS955, as it is working on XBR960. Allways good to find out about the little differences we experience with different set models.

Note that it didn't change the slider settings for different pic modes, just the one I was using at the time I pressed "reset". I didn't check to see if with "mode memory=off" if it was the case "input specific" values were still being stored(as is the case with Mode memory=on) - one wouldn't think so, but I guess you never know..

As I mentioned to ADU, however, I can't find a setting to "control" what happens with the "clearedgeVM" user menu setting when you press "reset" button. It's still defaulting to "high" Clearedge VM user menu choice for example for "vivid" if I press "reset" when in Vivid pic mode+video settings menu. Which makes this feature much less useful for me, as I'd have to change "clearedge" VM to my preferred value for any given pic mode everytime I hit "reset" button.

Nitewatchman

02-10-06, 02:08 PM

The experience you detailed in this post exactly parallels mine;

Same here.

You can remove every single one of those god-awful (in my opinion) edge enhancements in the service menu.

You can, but where I differ a bit with you on this is - at least at this point, that at least in my case I find a bit of "fine tuning" and mininal enhancements in MID5 useful(and fine tuning those for different sources), per Ken Tech's findings on this per post #707, as well as "subtely(spelling+if that's a word)" set up VM settings useful, even with SYSM=2.

Perhaps however differences in our set models(such as SFP or no) might account for why some of us may be be experiencing something a little different in this regard.

I do agree with what you've said about the 2170P-3 edge enhancement settings - at least from what I've seen so far. Although, If I wasn't using MID, and I was using SYSM=3, I'd have a use for them, as I'd have to use them to some degree.

(SYSM=1 is basically a lesser version of 2 (from what i've seen) i have no idea what it's designed for other than antenna tv. )

OTA (antenna) looks awfully good to me with SYSM=2 or 3 (given you've "equalized" and fine tuned other relevant settings accordingly) vs. SYSM=1 .... That's including NTSC (analog) OTA, and the relevant 2103-1/3d-comb settings/etc ....

I do use an outdoor, high gain directional antenna setup, which is in a sense "optimized" for excellent analog NTSC reception(which is an important factor when it comes to NTSC analog PQ) - which luckily also turns out to also work great for OTA ATSC(digital TV/HD/etc) reception.

I've personally never found a use for SYSM=0 or 1, except when experimenting/testing/etc. Perhaps however things would be different If I was using an earlier model set, or if I were(don't know at this point if you can even set that up on the newer sets, or what the "effects" would be) using the "PT" for 1080i as ADU is.

While the picture looks free of noise, and almost glossy, there is a severe lack of fine detail to focus on.

I also think this is an especially excellent description of what I've also experienced with SYSM=3 -- "Almost Glossy" is exactly what I've thought of it as well. In my case, SYSM=3 especially does not work well for 480p DVD.

As (hopefully) should be apparent from my post concerning experiments with the 2170P-3 edge controls per ADU's excellent info --- What I thought did seem odd was that in order to get SYSM=3 to "somewhat" work for 480p DVD, I had to add all sorts of enhancement from either MID5, or 2170P-3 - something which really isn't the case for SYSM=3 for other sources/scan rates/etc = and, what does work best for SYSM=3 and 480p DVD is a "balance" from both 2170-P3 and MID5 EE controls ... However, with SYSM-2, I don't really have to do anything "different" for 480p DVD than I am for 720p HD to achieve excellent resluts -- And all that involves really, is minimal enhancement from MID5.

SYSM=2.... This image is not QUITE as smooth or as computer screen-like as SYSM 3. But, fine detail is brought more cleanly and naturally to the surface.

I've noticed the same thing here.

Actually, Chad Bilheimer did my ISF. I'm sure you all have heard about him by now

From all accounts I've heard, Chad is very dedicated and seems to do a wonderful job. I also recall enjoying, and reading many of his posts on another forum several years ago.

I've seen him post on AVS a few times, perhaps If he ever gets the oppurtunity to do so, I'd certianly very much enjoy hearing his thoughts here concerning the signal processing of these Sony DV CRT's we're discussing, as well as the "image processing"/ "EE" settings/etc. available in SM which we are working with.

ADU

02-10-06, 03:15 PM

I deduced that nothing was retained because the positions of the "white dots" hadn't changed upon re-entry into the user menus. Pressing Reset means I have a bunch of work to do restoring my defaults to all four Picture modes. AFAIK, those dots are the positions specified by the user-default parameters, and they didn't move.Check the SM to see if the new values you entered for UPIC - UTMP are still intact. If they are, then the new slider config may be "in effect". You may not see it take effect though until pressing the RESET button on the remote (while highlighting that Picture mode in the User Menu). That's the way it works on the XBR800 anyway.

All the RESET button does on the XBR800 is snap the sliders/parameters in the User Menu to the positions defined in the SM by 2170P-4/UPIC - UTMP (and 2170P-3/VM for Clear Edge VM) for the currently selected Picture mode. That seems to be the sole purpose for these features on the XBR800.

If you have the MODE MEMORY feature on your TV and have spent alot of time tweaking the User Menu/Picture modes differently for each separate input, then I can understand why you might not want to use the RESET feature. But I use this all the time on the XBR800 to reset the sliders in Pro mode back to their default positions (as I've configured them to work in the SM.)that is, if one us catches his own typos! I completely obscured the main point of my post #955 by leaving out...Been there, and done that too. :)

ADU

02-10-06, 03:28 PM

Nitewatchman,

You might wanna check what 2170P-3/VM does on your TV again. FWIW, on the XBR800, 2170P-3/VM seems to behave exactly the way you're wantin it to (per Post #971 (http://www.avsforum.com/avs-vb/showthread.php?p=7102355&&#post7102355) above). Ie, it controls which setting Clear Edge VM defaults to (either High [3], Medium [2], Low [1] or Off [0]) for each Picture mode when you press the RESET button on the remote.

Nitewatchman

02-10-06, 04:48 PM

Might wanna check that again. FWIW, on the XBR800, 2170P-3/VM seems to behave exactly the way you're wanting it to (and the way described in Post #971 (http://www.avsforum.com/avs-vb/showthread.php?p=7102355&&#post7102355) above). Ie, it controls which setting Clear Edge VM defaults to for each Picture mode when you press the RESET button on the remote (either High [3], Medium [2], Low [1] or Off [0]).

I double checked it before posting my last post on this , and just checked it a third time.

P2170-3/VM stays at "0" If I press "reset" button on the remote while in "vivid" mode+in user menu "video settings" section, but the user "video" menu "clearedge" choice for "vivid" reverts to factory default choice of "high" Clearedge VM setting(thus it uses the P2170-3 VMH setting in this case). Note I pressed the reset button "outside" of SM, and entered SM afterwards to check the P2170-3/VM value(as I didn't "look up" what might happen hitting "reset" button while in SM, so didn't want to try that) - 2170P-3/VM still at 0, Clearedge VM user menu still at "high" for "vivid" pic mode.

If it was using the 2170P-3/VM setting and is working as it does on your set, then it should be defaulting to "clearedge:"off" choice in user menu when i hit "reset" for "vivid" pic mode given the P2170-3/VM=0 value for "vivid" pic mode on my set, correct? It is not doing that. It is "defaulting" to "High" user menu clearedge choice.

It's no big deal for me really. But, wish it would work, as Ken noted it would be especially useful to quickly+easily return to my normal "preferred" values for the sliders after making any slight adjustments to sliders to account for different program material/etc ... While it works for the sliders and UTMP "color temp" setting, It's just not as "convienient" or very useful if I'd have to keep changing back the Clear edge VM setting "manually" every time I'd use the reset button.

After "UTMP" in the "QM" section, the next control was "CVSB", but I don't think it has anything to do with the Clearedge setting, as it's value was "0" in all cases for all pic modes, whearas, with the reset button it's defaulting to Clearedge : "high"(Vivid or standard), "low"(Movie) or "off"(pro) depending upon Pic mode.

Thanks for your input+thoughts on this, though.

Update : just to check it a 4th time, In a little different way, I changed 2170P-3/VM to "1" for "vivid" mode. As I expected, Clearedge VM user menu setting still defaulted to "HIGH" when I went out of SM, and into user menu "video" settings and used the reset button(I'd set it "manually" to "medium" first, just in case, and so I could "see" it change).

Went back into SM, and P2170-3/VM was still set at "1" just as I had changed it(I then changed it back to "0" of course). Note that just to keep it a completely "controlled" experiment(although I don't think P2170-3/VM stores "input specific" values, only "pic mode" specific values), I "stayed" with the same input/source throughout the experiment, without switching to other inputs/sources/etc at any time.

Nitewatchman

02-10-06, 05:13 PM

If you have the MODE MEMORY feature on your TV and have spent alot of time tweaking the User Menu/Picture modes differently for each separate input, then I can understand why you might not want to use the RESET feature.

If Mode memory is on on my XBR960, using the reset button only effects the Pic mode and Only the input you are using at the time, not other inputs or pic modes(including pic modes for that single "input" your using at the time). With Mode memory On, you do have to configure your "UPIC~UTMP" values for EACH input, and EACH picture mode. Otherwise, the inputs you "don't" configure, will still have the factory defaults for the "little dots".

You won't see it take effect though until pressing the RESET button on the remote (while highlighting that particular Picture mode in the User Menu). That's the way it works on the XBR800 anyway.

Interesting. The "little dots" did move to reflect my new UPIC~USHP values on my XBR960 without me needing to hit the "reset" button first(of course, that' no actual "effect" to the picture). As I noted in earlier post, the dots didn't change In User Menu WHILE I made the change, but, after exiting User menu, and bringing it back up (the "video settings" portion of the menu), my new "dot" posistions were indicated instead of the previous factory settings.

Nitewatchman

02-10-06, 06:44 PM

Might wanna check this again. FWIW, on the XBR800, 2170P-3/VM seems to behave exactly the way you're wantin it to

I do see what you are getting at here, and what you are saying certianly seems to make sense, as my updates/info as mentioned below should also hopefully also point out ....

However, the "behavior" I'm actually getting with this is exactly as I've described, and I would agree THAT doesn't "make sense" but, nevertheless that is what is happening ...

I realize I had left an important word out when I said :

On my set(34XBR960) The P2170P-3/VM setting sets the "intensity" of SVM when you choose the "cleardedge = Off" choice in user menu, I of course want that one at "0" so it is actually "off" when I choose Off.

I thought I had written(important missing item in Bold) :

On my set(34XBR960) I think The P2170P-3/VM setting sets the "intensity" of SVM when you choose the "cleardedge = Off" choice in user menu, I of course want that one at "0" so it is actually "off" when I choose Off.

But, now I'm not so sure about that either, and have edited this section of my above corresponding post to instead read :

[edited]On my set(34XBR960), looking at it more, I'm not sure what P2170P-3/VM does. I had thought it set the "intensity" of SVM when you choose the "cleardedge = Off" choice in user menu, if so, I'd want that one at "0" so it is actually "off" when I choose Off. However, It only allows "values" of 0-3. I'd thought, although it would make more sense for "off" just to mean "off" and not need any sort of "value" involved or SM control for it --- I'd thought if it were for an "off" setting, it might just allow 0-3 to only allow for the most "subtle" of SVM intesities if it were to correspond to the "off" menu choice. However, the "0-15" settings that are available for all all the others(VMH/VMM/VML), as well as the servicecode listing defaults(which match for pic mode the "high/low/off" defaults I'm getting per pic mode per the labels in user menu "clearedge VM choices) would lead one to assume it should "work" as it does for ADU's XBR800, although that is not what I'm getting according to the "low/Medium/High" "labels" in the user menu it's defaulting to regardless of how I have the 2170P-3/VM value set. [end edit]

I guess what I'll have to do when I get some time is temporarily change the VML/VMM/VMH settings to higher values(and maybe the 2170P3 VM Shaping values as well), so I can easily, and quickly "spot" the VM or differing intesities(my current settings are really to subtle to easily spot this, except (to a very SMALL extent) for the very slight differences seen when "moving" through the different Clearedge VM User menu choices(low/medium/high).

As, maybe it is possible 2170P3/VM could be "working" on my set as it is on yours,EXCEPT that the labels for the "defaults" when using "reset' for "high/Medium/Low" are being shown even if a actual "different' value is being used .... I don't know though -- I think I can tell the difference between my VMH (high) setting from the "off" setting, and it certianly doesn't seem to be "off" when it defaults to saying "high" for "vivid" mode even though "vivid" mode 2170P-3/VM value is "0".

Or, (and this is what I suspect) maybe 2170P3/VM is really doing nothing at all on my set, and is a "leftover" in the SM from earlier models that is not being implemented properly, and instead, the High/medium/low VM defaults for the different pic modes for the "reset" button are "hard coded" somewhere, or available somewhere else in SM(although I can't find it). Perhaps, Ken is also getting the same thing on his set, but instead with the "sliders" and little dots and "reset" button and QM UPIC~UTMP.

ADU

02-11-06, 03:30 AM

I see what yer sayin now. Sounds like the MODE MEMORY feature is making things a bit more complicated on the XBR960. If I were in your shoes, I wouldn't bother trying to determine if the TV's using a different level of Clear Edge VM than the one shown in the User menu. That wouldn't make much sense. More than likely, it's gonna use whatever's listed. And if the parameter isn't changing to the new VM setting when you press the RESET button, then it sounds as if there may indeed be another switch or parameter involved somewhere else.

Perhaps the 2170P-3/VM only works when MODE MEMORY is OFF, and there's another parameter involved when MODE MEMORY is ON? Any idea how many different options the CVSB parameter contains (e.g. 0-3, or 0-4)?

Nitewatchman

02-11-06, 11:19 AM

If I were in your shoes, I wouldn't bother trying to determine if the TV's using a different level of Clear Edge VM than the one shown in the User menu. That wouldn't make much sense. More than likely, it's gonna use whatever's listed.

Yeah, I agree, I'm not going to mess with that. Was just trying to think of all the possibilities that might explain what's going on, no matter how unlikely.

Perhaps the 2170P-3/VM only works when MODE MEMORY is OFF, and there's another parameter involved when MODE MEMORY is ON?

2170P-3/VM doesn't work for this on my set with "mode memory" off, either. I tried it earlier, and just doublechecked it.

Another useful thing I did find out when doing this however is that when I turn "mode memory" off, and then back on, I don't lose any "mode memory On" input specific(or pic mode specific) user menu settings.

Any idea how many different options the CVSB parameter contains (e.g. 0-3, or 0-4)?

I don't know, I didn't check it as I don't know what it does. CVSB is showing a "0" value for all pic modes+inputs that I checked, so wouldn't think that would be reflective of the different user menu clearedge "defaults" for the different pic modes when the reset button is used(High for Vivid, Low for Movie/etc).

The Section of the SM CVSB(and UPIC~UTMP for that matter) is in is the "QM" section, which is an undocumented section at least it's not shown in the servicecode listing from Service manual. There is also a CVSB "entry" in the undocumenteed "QT" section of the service menu, it's also at "0" everywhere I checked.

----------

Which reminds me --- On another note -- an update on the "HDPT" PM I sent you and Ken the other day. Among other things, if you missed it, In the PM, it explains how I followed the info on this for older sets on other threads, and read all the approrpiate info in the "vertical bar" thread/etc.

Anyway -- I played around with HDPT=0 a little more in one pic mode last night. HDPT has Pic mode specific setting on my set - for instance, you can have it at "0" for vivid and "1" for Standard. With 1080i sources(which is all HDPT effects I believe), The screen going "scrambled" on my set(like you've lost V Sync or H Sync, and somewhat similar, in a sense to what you may get in some occasions if the 2171CXA "FIXS" paramenter isn't correctly set)occurs ONLY if you bring up the user "display" function, and only does that when your in SM - although, even outside SM, the user menus, channel banners/etc, are still "garbled" and not correctly displayed. Otherwise, it seemed to "work" to "by-pass MID".

I didn't find any "good" use for HDPT=0 it however. Except involving "differences" involved with values for the "PT"(passthrough) columns for some controls which differ -- (such as YOSW value, the 1080i PT specific CROP/CBOP(interestingly enough, CROP/CBOP AND YOF~CROF all are offsets that effect the "1080i PT" signal, as well as PT column CBGN~YGN/etc) from the "component" or 'ATSC" specific columns, I couldn't find any differences using HDPT=0 concerning the picture quality of 1080i sources as compared to using HDPT=1 and using all Zeroes in the MID5 Column.

Or, If there were any differences in "image quality"/etc, I couldn't "see" them. I also noticed, with "HDPT" at "0" there was often a "quick moving", rather wide Vertical bar moving across the screen at times(only on 1080i sources of course). Also note that MID1/"DPSW" "blank screen" on my KD34XBR960 if set to "1", and on this set "0" is the "normal" setting for it. If I recall my reading about older sets correctly, it's the other way around.

**disclaimer** At this point, I do not endorse, or recommend that anyone, at least with one of these newer Sony sets "check out" MID-by pass and the "HDPT" setting! As, (just based on my short experiments), #1). I don't think we know "enough" about what it does on these newer sets, it might not be 100% "safe", #2) I don't see that its "MID by pass" does anything "useful", and you can use all Zeros in a MID5 column+accomplish the same thing, at least that seems to be the case given my short experience with it, and from what I can tell. I could be wrong, of course. #3). Also, when you're in SM at least, you might find the screen becoming "all garbled" if you set it to "0" making the Service Menu info "unreadable", therefore you'll be in a real fix if you don't know a way "out" of it. #4). At least on my set, when using HDPT=0, the User menus and channel displays/banners/etc. do not display properly.

ADU

02-11-06, 02:35 PM

I got your PM on the above. And apologize for not getting back to you on that yet. Just a little backlogged at the moment.

http://www.avsforum.com/avs-vb/images/icons/icon4.gif A WARNING: The HDPT parameter that you refer to above has the potential to damage Sony TVs if it's not properly implemented by an authorized technician, especially on models which lack certain hardware. http://www.avsforum.com/avs-vb/images/icons/icon4.gif

No foolin or kiddin around. Tampering with this setting has the potential to damage your TV. That's why I have been reluctant to get much into it outside of certain threads which take the time to go into the subject in greater depth, such as the 1080i scrolling bar thread (http://www.avsforum.com/avs-vb/showthread.php?s=&threadid=190083) and the Sony signal processing thread (http://archive.avsforum.com/avs-vb/showthread.php?threadid=293741). If you want to know more about it, I would strongly advise reading both these threads in their entirety to get a better idea of the risks/requirements/headaches involved with this parameter. Please also see my cautionary remarks to Ken about it in Post #29 (http://www.avsforum.com/avs-vb/showthread.php?p=6944207&&#post6944207) and Post #31 (http://www.avsforum.com/avs-vb/showthread.php?p=6955693&&#post6955693) of another thread.

For the vast majority of users (especially owners of newer models with better processing), I don't think messing with it would be worth the associated risks and headaches that it may involve.

Nitewatchman

02-11-06, 05:09 PM

No foolin or kiddin around. Tampering with this setting has the potential to damage your TV.
.

Absolutely, and I can't stress enough that you HAVE to know what you are doing before you MESS WITH ANY SM control. See my Disclaimer in last post as well.

I read the info in the threads you mention in depth, as well as closely studied the "signal flow" diagrams. Nevertheless, with the lack of info available on it for the newer sony sets, without anyone trying it on these newer sets, or posting about it+without a "sony" engineer "in the know" I can call, I certianly STILL took a bit of a "risk" with it --- a risk I was very aware of the possible consequeses involved , although I took every step possible to ensure it was a "carefully" mitagated risk, as I knew I had a quick "way out" ...

Believe it or not, I am quite knowledgable about Electronics hardware, and actually DO know a little about what I'm doing when it comes to these issues and inner workings of TV's and service menus and such. I've been doing this sort of thing for 30 years now, including working on the INSIDE of sets .....

Update:

Here's another "cautionary" tale involving a certian model of set+SM I recall from the late early mid 90's. This particluar set's Service Menu happend to have "VFREQ"(sets vertical scanning frequency) as it's FIRST SM item. It was also the case that, the "code" one needed to enter on the remote(or front panel set controls) was such that I believe it was very possible for a user to "accidently" enter SM without knowing it, and, perhaps thinking they were turning the vol down or switching channel(don't remember which), they could EASILY lower the VFREQ setting(which "stuck" immediately, you didn't have to do anything to "write" the new value) without knowing it - which would not only make it impossible to "see anything" on the screen, it would also cause the set to turn off after a few seconds. And the set would NOT work again, unless the EEPROM was reprogrammed or the EEPROM chip with the firmware on it was replaced, or perhaps a VARIAC was used to step down the A/C current to where you could enter the SM without the set turning off in a few seconds, and be able to "see" the value on the screen, and change it back to correct value.

: end update

For the vast majority of users (especially owners of newer models with better processing), I don't think messing with it would be worth the associated risks and headaches that it involves.

From what I saw from it, I could be wrong but I don't think anyone will really gain anything from it on these newer sets.

Which Is the important reason WHY I thought I'd go ahead a post a little about it on this thread, as I thought I'd pass along that I can see no benefit whatsoever from using it with newer sets. THEREFORE, I don't think there IS any reason for anyone else with the newer sets to Mess with it. .

Maybe that wasn't a good enough reason, however. If so, and if necessary, and if asked I will delete my comments on this matter.

At first, thought I'd just send you+Ken a PM about it given your recent conversation on that other thread, but thought the info(that info being : From what I can tell there's no REASON whatsoever to try it with these new sets) might be useful for folks. On my set anyway, again, the short time I did check it -- it didn't accomplish anything "beneficial" for the "picture" over just setting all "zeros" for MID5 Column -- and I did look fairly "closely" at that.

In other words, my advice would be, don't mess with it.