THE SONY SERVICE CODES - Articles, Comments, Discoveries - Page 39

Below is a color bar pattern from my 36HS510 (4x3 set). I'm not sure where to start for fixing this. Is this focus, convergence? Please don't judge geometry or color accuracy by this image as it was shot from a tripod mounted camera and cropped for posting. It is only to show the colors bleeding into each other. I have placed arrows next to some of the most offending edges.

Only in recent weeks have I started SM tweaking, so I'm still a bit of a noob. As always, thanks for any helpful advice.

Jeff

EDIT- In the user menu, I subsequently tried adjusting picture, brightness, sharpness, hue, clear-edge to off with no improvement. HOWEVER when I reduced the color significantly, the edges cleaned up nice, but of course the color rendition was horrible.

I made overscan adjustment like you mentioned in the thread. My set was at 5% when I started and now it's down to 2%. One thing I notice was the adjustments seem to be global. I did my overscan adjustment on 1080i but that fixed overscan for all inputs and etc...

I maybe wrong here but you maybe able to fix that tilt by adjusting VANG and LANG on 2170D-2. give it a try.

Great thread. Five stars. There's more than enough info here to write a book on Sony calibration. Oh, and I also have a quick question.

QUESTION: Anyone here know of a service mode setting that will make closed captioning come on when the volume is muted? Even cheap TV's do this.

They seem to have every other possible setting tucked away in there, so why not this one?

Kentech,

Thanks. The focus patterns on DVE are razor sharp across the screen.

Thanks to you and this thread, my geometry is damn near peferect. Though I'm a wee bit concerned that I had to change VCEN to zero to acheive it (originally set at 8).

On convergence pattern (DVE, title 12, chap 20) I see a slight ingress of green and blue on the center portion of the uppermost line of the grid, but otherwise just clean sharp white throughout.

An interesting experiment for tonight will be to run color bars at 1080i on upconverting DVD player and see what happens. Otherwise I'll no longer be concerned about having sharp color bars.

I'll echo what others have already said THANKS KENTECH!

Deepfreeze- That image was just to show the edges bleeding, but thanks for tip.

I have an issue with the flicker problem on a KP53XBR300. I understand that you can do a work around by entering the service mode and changing the Video 5 lock to 1080i. However, I have been unable to locate the codes for entering the service menu to do this. Can anyone supply these and comment on making this change if you have experrience? thank you.

I've been working on evaluating in which cases I may like the new "tuned for" SYSM=3 settings(using MHLY=2/etc) "better" than our "tuned for" SYSM=2 settings. Since for the most part, as you say - the differences are very subtle(even or especially so with the use of test patterns such as AVIA TVL resolution patterns) it is making this task a bit difficult to accomplish and even more difficult to write about ....

At this point, the one thing I am fairly sure about is : I seem to prefer the new settings for SYSM=3 for 1080i from internal ATSC receiver and MS. Not only is there less ringing or overshoot, I'd say it looks like an improvement concerning fine detail as well. I also seem to be preferring the "EQ"(along the lines of your audio/graphic equalizer analogy) for 1080i concerning those "mid-treble" frequencies. That's true for HD and some high-quality images from MS scanned with 35mm film negative scanner, I'm not so sure yet about the varying quality of upconverted material from local stations.

I haven't given up on 720p+SYSM=3 yet, but at this point it also seems like I'm getting a clear "preference" for the old "tuned for" SYSM=2 settings with 720p. Unlike is the case with 1080i, It almost seems like MHLY=2 is causing a bit of a "blurring" effect with 720p that I can't seem to compensate for with MHYE/MHCE/etc, and I may also like the "EQ" a little better with SYSM=2. I've even tried it with a bit of Vertial EE with MVYL=3, and even some short experiments with 2170P-3 controls such as F1LV/LTLV ...

Everything else is "in between" for me, and it seems more difficult to decide on a clear preference ... For instance, At this point, much of the time I may be leaning towards SYSM=3 for 480p V5/V6, SD NTSC(RF or CV/YC), and for 480i ATSC for the reduced overshoot and ringing and more "3d-like" look. And, there are times I like the "EQ" better for those as well with the new, tuned for SYSM=3 settings. But, there are also times I like the "EQ" better for those with SYSM=2, perhaps especially for SD NTSC via RF input(OTA in my case) ... At times, it just seems a bit more "natural" looking with SYSM=2 even though the difference is very slight ...

Except perhaps for 1080i and SYSM=3, and perhaps 720p and SYSM=2 --- At this point, I'm thinking the question of which is better in any given case may end up boiling down to a matter of preference for me. If I can ever figure out what that preference is, I think it's going to take quite a bit of time spent "evaluating"/making comparisions between the two along with a little fine tuning of MHYE/etc in some cases to have any luck concerning making decisions on anything other than 1080i and 720p .....

Also at this point, except perhaps for 1080i, no matter what I decide it looks like I'm probably going to want to allways have a pic mode set up with the "Old" SYSM=2 settings for all inputs/scan rates ....

BTW --- given the descriptions in the XBR2 manual for MID5/MHLY, MHLC, MVLY, MVLC (LPF coefficient for "Y"(I assume luminance) or "C"(I assume chrominance), respectively) : I've also played around with MHLC/MVLY/MVLC a little as well with 720p HD to see if I could further "fine tune"/improve the "EQ" of those "mid-treble" frequencies a bit. While it looks like that "somewhat" may be the case in a very, very subtle way -- anything other than "0" for those(except MHLY=2) seems to be "too much" and involve "attenuation" of fine detail or add'l blurring effect, which we don't want ... I've only looked at it with 720p so far however ... I think It may also be somewhat interesting that for the "factory" defaults(from XBR960 servicecode listing, or from XBR2 Service Manual) for MID5 columns, although MHLY values of "1" or "3" are occasionally used -- they NEVER use MHLY=2, which seems to be turning out the be the value(other than 0) that's most useful -- for SYSM=3 at least ... go figure ....


So do I... What may surprise folks is just how similar the results are to our "tuned for SYSM=2" settings - What happens with SYSM=3 using MHLY=2 certianly surprised me!

Ken,

I'm probably just second guessing some of my original choices with this. As, after several evenings of viewing thought I had pretty much decided I preferred the "new" settings for SYSM=3 just about everywhere ... The main exception perhaps being 720p ...

From all I have read and learned about "tweaking" in these forums it was always apparent to me that the raster settings on these Sony models (2107D-1, 2107D-2) were global settings. However I just stumbled onto the fact that the 2107D-2 HSIZ and HPOZ can be set differently between video modes (i.e 480p, 720p, 1080i). I have the Sony 30HS420, and I was just wondering if anyway else was aware of this? As far as I can tell the VSIZ and VPOZ are global and can't be different between video modes, but I could be wrong about that. Any feedback would be greatly appreciated.

That might be the reason that there is no MID3 settings on the 420.

What exactly are MIDE settings adding to the image? For ONLY 480P material (what i mostly watch), I have SYSM=2, with VM turned off, sharpness leveled off, and most especially, MIDES set to zero. The last remaining sharpness setting in the SM i have is a SHAP-6 setting. Anything lower than that makes the image a touch too too soft. Still the image is plenty sharp.

I notice that adding MIDES to SYSM=2 make the image less film-like, as you are seeing more edging and layering to objects on screen. It can even create a very obvious 3-D effect when it is set a certain way. In this case, your eyes may have trouble following the material onscreen because they are constantly having to make a choice of what layer to focus on. From all i've observed, this is wrong for film. The only 3-D effect should be coming from the source material, not from some kind of layering the display adds to the image.

Maybe i'm just confused about the function of the MIDE column. It's funny, but on SYSM-2, when using a sharpness pattern, when you toggle between MIDE columns=0 and a minimum MIDE-column setting, the sharpness pattern respecitvely improves. Does this make sense? Although when i strip the image of MIDES, the image clears us much more. It looks far more film-like, and looks far more natural. Without the MIDES it looks like you are looking through a window to a clear and natural image, as opposed to having things artificially jump out at you. The former is what resembles film more.

From all that I have seen, MIDE only seems appropriate to SYSM-3. For me, it seems that setting SYSM-3 introduces far more variable into the equation, most especially VM's (which i personally detest). SYSM-2 seems to require a lot less running around and balancing things.

Someone like Ken would know more about this than i ever would, as his testing has shown. I only am giving you my own observations of what the images look like.

Does anyone want to comment on this?

That's what I'm starting to wonder too, I know that the HSIZ and HPOZ can have different values but the VPOZ and VSIZ seem to have to remain the same. As I look at my notes from the default values for the different video modes (480, 720, 1080) none of them are the same, so maybe there is a way to have different VSIZ and VPOZ in different video modes. I'll have to do some more testing to try and figure out, but even with the ability to change the HSIZ and HPOZ for different modes does not mean the MID2 settings still don't need to be tweaked (on my TV anyway, I have to use MID2 cause there are no MID3 settings, which you touched upon above). Right now all my notes on values are kinda cluttered on mulitple sheets of paper (which is not like me at all) so I'm gonna have to organize that and maybe help learn a little more.

In my service menu for the 34XBR960 I cannot have different HSIZ and HPOZ for different video modes and inputs, so I cannot really speculate. As I have already stated earlier most 2170D-1 and 2 codes are universal, MID2 is for 480i and MID3 are for the rest.

I would try to get an updated service manual for your TV. Somewhere in this thread there is an 800 number where you can order the service manuals.

As always - take baby steps an document them.

Sorry about the length of below, but hopefully some of it is useful in some way ...


I'm having a hard time understanding what you are trying to say here, which is why I'm not sure if you understand how P2170-3/ MIDE and MID5 table/Columns/controls "work" ....

If you set 2170P-3/MIDE to "0" for 480p V5/V6(component input), for instance, that is "assigning" MID5 Column "0" to be used with 480p scan rate from component input. If you haven't changed all the values to "0" for all the controls in MID5 Column "0", setting MIDE=0 certainly does not equate to NO MID5 processing, but setting MIDE=63 (a MID5 "all zeros" column -- assuming factory values) does, effectively "turn off" all MID5 processing ....

On XBR960 for instance, MID5 Column "0" is set up for the factory pro defaults as assigned for "RF" input(which is also set for "SYSM=1" from the factory, btw), With 2(low pass filters) of the 17 MID5 controls set to values other than "Zero" ---- MHLY is set to "1", and MHLC set to "3" Yuck, IMO ! -- But, using MHLY=2(even if all the other MID5 paramenters are set to "0") when using SYSM=3(also use P2170-3/SHOF=3 for "mid range" sharpness slider settings) is a whole "different" ball of wax, so to speak!

Given the "factory" settings on XBR960 -- MID5 Column #63 is one MID5 column that has nothing but "zeros" in it, so if you want to "turn off" processing from MID5, use a value of MIDE=63 for the scan rate/input/etc. you're using ..... In this particular case, MIDE=63 should be "sharper" than MIDE=0. ......

here's another hopefully useful(probably more understandable than the above) explanation from Ken Tech :

From Section #2 Of Ken Tech Article #10 - "Presets" and related Service Codes :


Here, in Post 88 Ken responded to question from Jsperm :



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


Rather than providing descriptions of the "useful"(IMO) MID5 controls here -- I think the answer to your question is mostly available in the descriptions of the Image processing controls(including MID5 controls) Ken provided in "IPChart06all.pdf" chart which is attached to Post #1039

Also be sure to check out the updated info on MID5/MHLY and SYSM=3 in Ken's Post 1138 .

Also, If you haven't already done so, I'd recommended digging up the service code chart PDF for your set and downloading it/printing it out so you can more clearly see how "scan rate or input specific" SM settings work. And, in the case of P2170-3/MIDE and MID5 columns, to perhaps get a clearer idea of how those settings "relate" to one another ... I don't recall which set you are using, but the servicecode chart for XS955+XBR960 is attached as/labeled as XS955:XBR960 Service Data.pdf, here at Post 715

I would also recommend thinking about "customizing" one or more of your pic modes other than "pro", so otherwise your "customized pic mode" is the same as "pro" -- This makes it possible to easily make "side by side" comparisions of different image processing settings set up in different pic mode. See KenTech's Article #5 - Customizing Picture Modes for info on how to do this.

For even more background+info concerning image processing settings, including info on MID5 table controls and what they do --- You might also want to read through some of Ken Tech's previous articles on "Image processing". Such as :

#1). Ken's Article #12 - Optimizing Image Decoding, Shaping, and Enhancement". Note that the "meat" of the article is contained in the PDF file attached to that post.

#2) post #707 (tuned for SYSM=2 values). Note : well, looks like Post 707 is actually post 706, now, as someone must have deleted an earlier post ... LOL ....

Just keep in mind, some earlier "discoveries" (such as Ken Tech's article #12) have been improved upon in later articles, so some of the settings or info described in earlier articles are not allways necessarily the "currently" relevant info on those thesttings, although I think there is still much useful info to be found in those earlier articles/posts for those who haven't been "following along" and trying these settings out ....

Utilizing info in links I provided above, I think you may be able to gain a better understanding of the image processing which is occuring, the controls that can be used to improve and "fine tune" image processing and EE, and, perhaps a better understanding concerning how to go about "tweaking" for improvments to get what you want - Including the MID5 controls, some of which I think can be quite useful. I believe there are also a few other posts earlier in this thread which may be benefical to you in this regard as well, but I don't have time currently to dig up their URL's ....


I disagree, and I just don't see how that can be the case. Various settings available in P2170P-3, MID5, and 2103-1(the latter only in the case of SD that passes throgh 2103cxa chip) -- are all *relevant* and approriate for image processing no matter which value of SYSM you are using .... And, how those settings(besides MID5 table settings) are set in your case for any given input/scan rate are *relevant*, important factors here as well.

In the interest of not making this post overly "long" -- I won't address 2103-1 settings, or the 2170P-3 VM settings here : Otherwise --- Specifically for 2170P-3, in additon to SYSM this would involve - SHOF, SHFO, PROV, F1LV, LTLV, LTMD, CTLV. VM settings can pretty much be thought of "seperately", and for now, let's just assume you have VM "off". Mostly Per Ken Tech's recommendations, along with what I "like" from these settings with my own evaluations/testing, I have these set (for all inputs/scan rates, and for my "customized" Vivid/Standard/Movie Pic modes - I've left "pro" at it's defaults for now) at :

SHOF = "0" for SYSM=2, "3" for SYSM=3
SHFO = 1
PROV = 3
F1LV = 0
LTLV = 0
LTMD = 1
CTLV = 0

If you've got something different from the above -- especially perhaps LTLV or LTMD -- then, certianly we are going to see something quite "different", even if we are both using a MID5 Column with all its controls set to "0".

That doesn't necessarily mean, for instance you'd want to "change" the value of P2170-3/SHF0 because of a different value for SYSM is being used, only that the setting used for SHF0 is a relevant and "approrpriate" setting that is involved, no matter what value is used for SYSM .... For SYSM=3, You *will* however want to have 2170P-3/SHOF to "3" for midrange "sharpness slider" setting, and for SYSM=2, SHOF=0 for mid range sharpness slider setting ....

Now, that being said, I suppose a case could certianly be made for using an "all zeros" MID5 Column, no matter what value for SYSM is used .... that may be more a matter of preference. However :

#1). I'm likely going to want to use MID5/MHLY=2, and 2170P-3 SHOF=3 if I'm using SYSM=3, and MID5/MHLY=0, 2170P-3/SHOF=0 if I'm using SYSM=2. Apparently(from what we've discovered so far) Those are the only *siginficant* changes and balancing that need to occur for SYSM=2 vs. SYSM=3. Additional, but VERY subtle "fine tuning" of MID5 columns for SYSM=3 instead of using SYSM=2 are useful, but it's a VERY subtle thing --- for SYSM=3 MHYE/MHCE +1 or +2 from the SYSM=2 settings, and in some cases(For SD, Not DVD) with SYSM=3 use of slight vertical EE via MVYL=3 seems to work well.

Try this : Using SYSM=3, SHOF=3, and sharpness slider at "31" -- With 480p from DVD, Check it out starting with a MID5 column with all zeros. Such as factory defaults for column #63 -- use 2170P-3/MIDE=#63 which assigns MID5 column #63 to 480p. Except, change MID5 Column #63's MID5/MHLY to "2". Then, compare that with MID5/MHLY=0, by changing MID5 Column #63's MHLY back to "0" --- which makes it "all zeros" again, and turns MID5 processing off. Switch between MID5/MHLY=0 vs MID5/MHLY=2 -- put up AVIA Sharpness pattern and especially look at what happens with the horizontal sweep portion of the pattern when you change MHLY between 0 and 2 ... Now, compare SYSM=3/SHOF=3/MHLY=2 to SYSM=2, using SHOF=0, MHLY=0 and sharpness slider=31 .... Do the same thing, but with *real* program material ..... Any differences you might see should be quite subtle ... Also, don't forget that how you have other settings in P2170-3 are also "relevant" to some extent here - such as PROV, F1LV, LTMD, SHFO/etc ...

# 2) I personally prefer to "fine tune" the MID5 columns a bit for "fine detail" in a subtle way, and that is as true with SYSM=2 as it is SYSM=3. What I mean by "fine tuning" involves different MID5 columns set up and "assigned" via 2170P-3 MIDE for different scan rates/inputs --- and, as is explained in KenTech's articles -- specifically "fine tuning" with the following MID5 controls(everything else in the MIDE "assigned" MID5 columns are set to "0", except for MHLY=2 if I'm using SYSM=3) : MHYL/MHYE, MHCL/MHCE. For SD(but not DVD) in some cases something a little less subtle using MHYO/MHCO, and also in some cases(480p DVD and in some cases also with SYSM=3 for SD) a little vertical EE via MVYL=3.


For the most part I disagree. You can't just look at either SYSM=2, or SYSM=3 "on it's own" when looking at the overall effect on the picture, IMO. You have to take all the other settings into account which effect, or can "control" various factors involving image processing/EE or "sharpness" as well ....... Now, from what I've seen I do think it may be a bit more difficult to "fine tune" MID5 controls such as MHYE/MHCE for SYSM=3 for best results with any given scan rate/input than is the case with SYSM=2. Otherwise, I think the "running around" and balancing things is just as important for SYSM=2 as it is for SYSM=3 in my experience ....

As for VM, I've found subtly "configuring" the VM parameters per Ken Tech's recommendations to be quite benefical. You might be surprised at just how "subtle", and "minimal" VM IS with P2170-3 VM settings of VML=2(for "clearedge VM=LOW" in User menu), and VMCR~VMDL = 0-3-0-6, or say, 0-3-0-8 actually are compared to VM=OFF .... It's a far cry from the "factory" setup, and, I'd certianly want VM "off" if the factory VM settings were my only other choice ...


Keep in mind, 2103-1/SHAP does not effect 480p, 720p or 1080i -- those signal paths don't use the 2103cxa chip. It does effect SD that uses the DRC processing -There are seperate columns for SHAP, allowing for different settings for RF(NTSC SD) CV/YC(NTSC SD), V5/V6(480i, such as from DVD), HDMI(480i), and ATSC(480i digital from internal ATSC/QAM tuner).

In my case, 2103-1/SHAP=6 certianly seems WAY High for my set, I wouldn't use anything higher than SHAP="1" --- Ok, I've thought of wanting to use "2" occasionally - but rarely. Anything above "1" is just too much and is too "scratchy" looking, IMO. The exception being for my Right Twin-view window via 2103-2/SHAP, which I have set to "4"(RF) or "5"(CV/YC) as for some reason, the right twinview window is "blurrier" than the left, which is controlled(where applicable for SD) by the 2103-1 settings. My best guess would be, perhaps it involves something different going on with the 2nd tuner's alignment at the factory, or different signal paths used for the 2nd NTSC tuner/rerouting CV/YC through it/etc ...

The appropriate MID5 column settings can be tuned for "better", and much more "subtle" and pleasing results than using "SHAP=6" for any given "SD" input/scan rate/etc, IMO. Also, I'd recommend checking out Post 1137 for excellent info on "tuning" the approrpriate 2103-1 controls for excellent results with SD/480i from NTSC, 480i DVD or 480i ATSC ... Just keep in mind, You'll also still need "appropriate" values set up for various 2170P-3 controls/MID5 parameters/columns ...

I think I was mistaken earlier, I believe that only the HPOZ can be different. I really do need to get a newer version of the manual, I think that I'm trying to change the wrong things for the wrong video modes (I do document everything though). I need to figure out which settings for the 30HS420 control the 480, 720, and 1080 pictures modes cause I think I'm not doing something right.

Just curious Ken, what is your SBRT setting set at?

Mines at 29 very close to sonys intended range of 31. Ive tried 31 but it's a tad too much on the smokey side and 29 seems to be the best compromise for shadow detail and black level. if not perfect!

Just curious that's all since it seems we have found similar discoveries with our settings,maybe we think alike no

Is there anyone here that knows how the video settings to fix overscan on the newer 30HS420s work? I think mine was manufactured in November of 05. What I mean by video settings to fix overscan is what to use to set the screen up for 480p, 720p, and 1080i. I know 2170D-1 and 2170D-2 are for the raster image, but what exact settings are for the rest the setup on the newer models is rather different than the older ones. I was using the MID2 settings for 720p and 1080i, but I'm starting to think that's not the right approach (like I said the settings are rather different than the older models). Any insight whatsoever?

Here is the problem I have come across with my Sony 30HS420. I own a Xbox 360 which I spend as much free time playing as possible, for the longest time I was using the MID2 settings DHHP, DHHS, DHVP, DHVS to size my 720p and 1080i input modes. I thought all was fine until the new Ghost Recon came out, when I first turned it on the HUD was awfully overscaned to my amazement. I tinkered a little and got it to where it fit on the screen, and that was not an easy task one bit. After this I figured I had now properly configured my TV for my 360, until I put in another game and saw that the sides of the TV had black bars (in other words there was a space between the image on the screen and the edge of the screen that was not supposed to be there.). I figure that this can't be the setting for changing the image on top of the Raster, cause if it were it should have not created any black space correct? It should have filled out the rest of the screen with the image the game was presenting, am I correct in this thinking?

I have a problem. I have the xbr910 and I have adjusted sbrt to 29 and ubof to 2 in the Service menu. My user menu setting for brightness is set at the midpoint. But the picture still seems a little to dark. I raised the ubof to 3 but it seems like the blacks get washed out.

Interesting, that's exactly how mine was shipped from the factory as well.Only difference with my tv is i set UBOF to 0 for all the other inputs any higher and it's smokey.

Ken, thanks for the good information, and the specific settings for MIDE appropriate to what i have. I was missing the point of MIDE, which is what i thought.

The VM's i refer to are in that one column, i forget the number, but they include "VM" values. I have set those to all zero from day one. I also kept "clear-edge" on off. But that is reflective of my settings in the service menu. I have read most of your articles, and since i'm "getting smart slowly" it takes a while for me to understand how some of these codes interact - it's like trying to understand a Matrix movie. (Although your explanations are more sufficient than what the W-Brothers provide.)

You mentioned the PRE, SHFO, PPHA values have no impact on 480P material right? I mostly view 480p material, so i imagine SHAP is the only value that comes into play with 480p.

I will look again at some of your articles. My only usual trouble is dechipering which values you post compliment (or go along with) with SYSM=3, and which deal with SYSM=2. Again, i understand a lot of this has to do with my personal learning curve as well.

Thanks for all of your help Ken!

Nitewatchmen - Thanks as well for your input, i'm sure others will benefit from that too!

Just checking in to say that I tried your new settings, KenTech, (except for your newest one with MHLC) and they look good. I haven't set up for any A-B testing (as of yet) but I did have a friend over last night to view a couple of films. Afterward, he asked if I had been making more adjustments and I told him of your recent findings and how I had implemented them. He said it looked great to him. And that was with the sharpness control all the way off and no SVM.

And now, a grayscale calibration question: I tried this recently (and briefly, I might add) and it didn't seem to work too well. I had a theory that I could set the low end of the scale by using the RGBS function to set the black level of each color individually with a PLUGE pattern. Since the results weren't what I had expected, I have to ask if is this a feasible way of accomplishing this? After more adjustments, my picture looks fairly gray now (in B&W; and thanks for the tip about setting scale while viewing a B&W movie instead of staring at those damn grayscale test patterns for hours) but some films still appear a bit dark and I'm thinking that one or more of the colors being "buried" below black may be causing this.