THE SONY SERVICE CODES - Articles, Comments, Discoveries [Text View] - Page 9

KenTech

09-07-06, 02:16 PM

Hmm. . . I'm confused what are the MID menus for? I thought all the geometry settings were grouped together under some of the 2107 menus.The MID (Multi-Image Display) chip digitally sets up what is displayed onscreen, forming frames to hold reduced-size views (TwinView), establishing a 4:3 "window" on 16:9 sets, and generally digitally mapping and scaling to the screen. Because some of the ratios in MID1~3 have to be computed exactly (example: vertical pixels in the image vs. scan lines in the raster), adjusting some of the scaling parameters can have disasterous effects on resolution. In particular, none of the vertical-size-related parameters should ever be changed, and I would avoid any horizintal-size parameters, as well. I haven't seen problems with tweaking the positioning parameters, however, but I would resist the urge.

None of the parameters in MID1-3 are marked as final-assembly adjustments but are ID's as "fixed data" (green). On the other hand, most parameters in 2170D1 and 2 are color-coded as "final-assembly" adjustments (yellow).

I would make all geometry adjustments in 2170D if at all possible, and tweak positioning parameters in MID1~3 only as a last resort for small tweaks. Avoid the MID1~3 parameters that imply that they adjust size, height, width, or area (ending in W, R, or S, for example).

If parameters are tentitavely changed in MID1~3, I would check the AVIA Sharpness or Resolution screens before and after to make sure nothing has been compromised. After I tinkered a bit, it's the AVIA Sharpness screen that alerted me that I had screwed up the vertical pixel-mapping badly: horizontal dark lines had turned to mush! (I'm sure one of the sharpness-test screens in DVE would work, too.) Returning the values to the data-chart's numbers fixed everything.

KenTech

09-07-06, 02:36 PM

Based on the list posted by Oliver above (thanks!), I'd put money on the following equivalents in the older service manuals' MID3 group: (The latest I have is revision 7/2005.) The abbreviated identification phrases are subject to interpretation, but digital timings seem to be prominent -- pulse widths, clip levels, start and stop points for a phase-locked loop, etc. "clp" may mean "control pulse" or similar, not "clip." "cl" may not mean "control," either.

0 YCPO [7/2005 has only one: 5 VCPO = vdo clp pos.]
1 CCPO [See above. Are Y and C now separate?]
2 PRPB
3 DOSA
4 YCWD [6 VCWD = vdo clp wdt]
5 MYCD [7 VYCD = vdo yc delay]
6 PSTP [8 VSTP = vdo pll (phase-locked loop) stop]
7 PSTT [9 VSTT = vdo pll strt]
8 VHSC [10 VHSC = vdo hsync cyc]
9 VHSL
10 PLHC
11 PLHL
12 MDTC
13 MFRV [11 VFRV = vdo flv rev]

Dr. Spankenstein

09-07-06, 04:31 PM

My MID3 menu is the same as the one Oliver posted. Along with that, I'll ad my MID1 and MID2 menus for your perusal:

MID1:
0=DHPH
1=DVPH
2=DHAR
3=DVAR
4=DHPW
5=DVPW
6=DYCD
7=DYSD
8=MDHP
9=MDVP
10=MDHS
11=MDVS
12=DGSB
13=DGSR
14=DPSW
15=MDLO
16=BCOL
17=DYSS

MID2:
0=DHHP (HORIZ. POS.)
1=DHHS (HORIZ. SIZE)
2=DHVP (VERT. POS.)
3=DHVS (VERT. SIZE)
4=DHVL (?)

Again my set is a KD-34XS955. Build date: July 2005.

Bryan

Istari1

09-07-06, 06:05 PM

Good info as always, thanks Ken. Attached is a simple pic of what my geometry "problem" is. Really its only a problem when watching something with boxes on the sides - its unnoticeable during DVD or HDTV watching. Any suggestions? I tried PIN and PINO but since the bugle is off center they dont seem to do much. The left side is slightly more bowed then the right side also.

corlay

09-07-06, 08:52 PM

Does anyone know what the following geometry settings do?

2170D-1:
(10) MHTZ

2170D-2:
(5) PINO
(9) UXCG
(10) LXCG
(11) UXCP
(12) LXCP
(13) XCPP

2170D-3:
(10) AKBT

I've changed the values, a bit, to experiment;
but changes to my display seem imperceptible...

--
corlay

stuart81

09-07-06, 09:01 PM

hello to everyone. im new to this, so please pardon me if this has already been discussed.
i own the 960 and i just finished adjusting the overscan on 1080i sources. my problem is that after i reduced the overscan on 1080i the 480i image becomes just a tad small. my question would be; is there anyway to adjust the 480i horizontaland vertical parameters independently of others?

thanks in adance for any help.

KenTech

09-08-06, 01:17 PM

my question would be; is there anyway to adjust the 480i horizontaland vertical parameters independently of others?Welcome to the forum, Stuart!

If you'll refer to the service-data chart for your TV*, you'll note that there is but one setting for raster width for normal viewing, 2170D-2, #2-HSIZ. But the height can be adjusted with 2170D-1, #13-ASPT, and there are separate settings for the various video-display modes. Bottom line: You're stuck with one adjustment for overall width, but can tweak ASPT (aspect ratio) for the different display modes (vertically compressed and not for 4:3 set owners).

If the 1080 frame is not horizontally centered relative to 480, use 2170D-2, #1-HPOS to adjust. There are two settings, one for 1080 and another for "others."

*Use the Contents listing in post #1 to find it; it's been uploaded.

KenTech

09-08-06, 01:36 PM

And here are the 7/2005 manual's lineups to Dr. Spakenstein's listings (thanks!):

MID1:
0=DHPH -\
1=DVPH ---\
2=DHAR ---|
3=DVAR ---|
4=DHPW ---|
5=DVPW ---| [0-11 Same as 7/2005 manual]
6=DYCD ---|
7=DYSD ---|
8=MDHP ---|
9=MDVP ---|
10=MDHS--/
11=MDVS /
12=DGSB
13=DGSR
14=DPSW [22 in 7/2005 manual]
15=MDLO [23 " ]
16=BCOL [24 " ]
17=DYSS [25 " ]

MID2:
0=DHHP (HORIZ. POS.) [DRHP - drc hactv pos]
1=DHHS (HORIZ. SIZE) [DRHS - drc hactv siz]
2=DHVP (VERT. POS.) [DRVP - drc vactv pos]
3=DHVS (VERT. SIZE) [DRVS - drc vactv siz]
4=DHVL (?) [None. Stops at 3.]

Oliver Deplace

09-09-06, 01:51 AM

If parameters are tentitavely changed in MID1~3, I would check the AVIA Sharpness or Resolution screens before and after to make sure nothing has been compromised. After I tinkered a bit, it's the AVIA Sharpness screen that alerted me that I had screwed up the vertical pixel-mapping badly: horizontal dark lines had turned to mush! (I'm sure one of the sharpness-test screens in DVE would work, too.) Returning the values to the data-chart's numbers fixed everything.

After reading that, I put up the Avia Sharpness pattern and went to town in MID2.
I ran the DHHP/DHHS/DHVP/DHVS settings up and down through a wide range and saw no resolution change. The image got squished and stretched (or lost synch), but the lines remained sharp.

A side effect of this foolin' around was: If I pulled the image vertically, as if I were trying to run it underscanned, a black screen would display glow spots at the edge that was exposed. There was a definite demarcation, where the screen is pure black and one click of a vertical control setting and the spots are there.

So, I guess that could be a concern when messing with the MID2.

KenTech

09-09-06, 03:18 AM

I ran the DHHP/DHHS/DHVP/DHVS settings up and down through a wide range and saw no resolution change. The image got squished and stretched (or lost synch), but the lines remained sharp.My digital camera is unavailable for about ten days; but when it returns, I'll try to take a photo of the blurring effect I saw.

BTW, it took only a 1-click change from the optimum vertical-size setting to compromise the picture quality. (Greater deviations from optimum didn't screw it up more, just differently.) At the correct setting (the one on the chart), the vertical sampling ratio is perfect, and the horizontal lines were perfectly sharp. At a value only one higher or lower, the "fuzz" of antialiasing (or fractional-pixel sampling?) appeared above and below horizontal black lines at the center of the screen, making them mushy. So my warning remains -- I know what I saw. But I will try to document it in a week or so and indicate how to reproduce the effect.

On the other hand, seems to me that a change to the positioning parameters in the MID groups wouldn't screw up the sampling -- only slide the image around on the digital "canvas," keeping the same exact pixel-mapping ratio.

Oliver Deplace

09-09-06, 12:51 PM

KenTech

09-09-06, 01:05 PM

I tried it again and now see what you're saying.
The horizontal lines, which form the cross at the center, get thicker. Interesting thing is, if you continue adjusting, the lines would randomly return to their thin state.
I was looking for mush, but they remained sharp, only thicker.Yes. I think the "look" may depend on how you have DRC set -- interlaced, progressive, etc. I typically examine the AVIA Sharpness screen in Progressive or CineMotion (same thing for this screen), and the deterioration was remarkable. Like you said, as the displayed picture is recomputed to a different pixel-sampling ratio for each setting, the results change, some better, some worse. The only one I found that was fine from top to bottom was the original setting. Thing is, I didn't realize I was screwing up the picture; I thought I had found another clever way of adjusting the geometry! (Hah! See my signature slogan!) The sneaky nature of this is why I'm making a big deal out of it -- so fellow perfectionists (you know who you are!) don't inadvertently screw themselves up.

These particular MID settings are identified as fixed presets, not "adjustments," and I have left them alone, getting scaling tweaks from other, intended, methods -- such as the 2170D settings identified as true adjustments. But I did use one of the MID position settings to perfectly align the centering of 480i with 1080i (or something similar -- it was a few months ago). There seems to be no downside to that, except I seem to remember that if you go to extremes, the picture may break up or "tear" in strange ways.

bazooka

09-11-06, 11:44 AM

Hey All, first post here. I've been reading thru the forums and found a lot of great info. However, I just noticed that if I display an all white screen, the left side of my xs955 is a faint blue-ish white while the right side is very faintly redish white? I've been going thru the settings trying to figure out how to fix it, but came up with nothing.

SurfingMatt27

09-11-06, 11:50 AM

bazooka

09-11-06, 11:55 AM

Yeah actually I do have some speakers next to it. Heh, I'm no geek at stuff like this so I didn't think about that. As for degaussing itself, I turned it off waited a few, and turned it on and it's a little better. But now I notice another problem, when I run a sharpness test pattern, one with verticle black bars, the bars seem to have a very small red outline, moslty on the top-left portion of the screen.

SurfingMatt27

09-11-06, 12:14 PM

Firstly you have to let the tv warm up a bit like say 20 minutes before everything looks fine.

It's normal to have some slight convergance issues at the corners of the screen, they go away though once the tv warms up.

bazooka

09-11-06, 01:31 PM

Yeah it looks good now. I also fooled around with the LANDING and got the colors to be even across the screen.

SurfingMatt27

09-11-06, 05:32 PM

Glad i could help!

I know those settings sure helped me with my sony 34hs420.

atye

09-13-06, 09:41 PM

Did a search, couldn't find the answer.

I'm in serious trouble here. I was changing parameters in the SM, did a
mute/enter to write in my 970, and all of a sudden, I get a grid
pattern like a convergence grid on my Toshiba, accompanied by a very loud
sustained beeping sound. No matter what I do, this won't go back to normal.
When I power off and back on, it shows the previously displayed screen
for a moment, then goes back to this grid/beep screen. It's almost like an
emergency/safe/shutdown type thing.

Does anybody know what I'm talking about? How do I fix this?

Thanks

KenTech

09-14-06, 03:16 PM

Does anybody know what I'm talking about? How do I fix this?Easy. You have inadvertently invoked one of the internal test patterns. You just have to shut it off.

You didn't say if you have turned off the set, but go back to service mode, and step backwards to group QM, then go to #1-PATN. You want to set this parameter to ZERO, which will turn off the pattern generator and restore your set to normal. WRITE the setting to preserve it.

Trouble is, the pattern generator takes over the set, and many functions on the remote don't work, especially volume! Be sure to turn down the volume before you invoke any of the patterns. They're harmless and of considerable vale for certain adjustments, but it doesn't replace a good calibration DVD, such as DVE or AVIA.

See the article here (http://www.avsforum.com/avs-vb/showthread.php?p=6181230&&#post6181230) for some details of how it works on the DA-4-based sets.

atye

09-14-06, 03:22 PM

Thanks Ken. I'll give it a try!

Dr. Spankenstein

09-15-06, 05:01 PM

Hello all,

I am rapidly approaching the end of my warranty period. I have recently lost the ability to adjust some of the vertical corrections within the service menu and was planning on taking the set in for repair or replacement of one of the boards. I have also notice more recently that I think I'm seeing some burn-in from my family's viewing of 4:3 material. I can only describe the burn-in as areas that are a shade darker/dimmer and a slight linearity issue when an image moves past these areas (like a right to left pan or vise-versa.) Does this sound like a form of burn-in or maybe an aperture grill failure? If so, show I look into having the tube replaced since they aren't going to be producing these tubes anymore?
It seems like it would be reasonable timing to have something like this done while the parts costs are being taken by Sony.

Any thoughts?

Bryan

Ectospheno

09-15-06, 08:37 PM

Welcome to the Forum! Yes, they should. Do them in the order blue-red-green. Use the blue to adjust Color and Hue, too before proceeding.I would toss the plastic filter card and, for sure, and use the RGBS controls instead: 7=normal, 4=red, 2=green, 1=blue, and (watch out!) 0=black screen. If you hit 0 accidentally, just go back up to another color. Most folks have compromised on settings near 14-14-6-4. Even when I vacillate among different settings for the first two parameters, the -6-4 part seems solid.

The HS420 uses different phosphors from the newer fine-pitch tubes, but I'll bet that doesn't affect the above color settings.

Hi. I recently purchased the KD-34XBR970. I am very pleased with the purchase. The only thing that bothers me is that after going through AVIA to change the video settings the red and green guns didn't look like they are set high enough. I'm in Pro mode, neutral color model, and set everything else using the AVIA disc. At the color check portion blue is at 0, green is down at -15% and red is 10% to 15% below when set to monitor and a similar level above when set to default.

I've read through what I think are relevant portions of this thread, including the quoted portion above. I have two questions:

1. Am I correct in thinking RGBS is what I need to modify?

2. Having never been in the service mode before, the 7=normal, 4=red, etc portion of the comment above is somewhat confusing. Would someone be kind enough to explain that.

Dr. Spankenstein

09-15-06, 09:25 PM

Ecto,

Actually, RGBS is an adjustment parameter in the Service menu. It will turn on/off any combination of the color guns.
The adjustments you will use to correct the color decoder is: 2170P-4 #7RYR/#8RYB/#9GYR/#10GYB. Most people have found that you can get quite close using the settings 14-14-6-4 for the above parameters in the 2170P-4 menu.
What KenTech was suggesting was to use the settings in the RGBS menu in place of of the supplied color filters. They will provide a pure color source from which to check your decoder settings via the Avia disc.
I more than likely have made this more confusing with my explanation.
I humbly enlist others in this thread to make this more succinct.

Regards,

Bryan

Dr. Spankenstein

09-15-06, 09:28 PM

Ecto,

Actually, RGBS is an adjustment parameter in the Service menu. It will turn on/off any combination of the color guns.
The adjustments you will use to correct the color decoder is: 2170P-4 #7RYR/#8RYB/#9GYR/#10GYB. Most people have found that you can get quite close using the settings 14-14-6-4 for the above parameters in the 2170P-4 menu.
What KenTech was suggesting was to use the settings in the RGBS menu in place of of the supplied color filters. They will provide a pure color source from which to check your decoder settings via the Avia disc. 7=All colors, 4=Red, 2=Green, 1=Blue. 5,6 and 3 are combinations of 2 colors. So when it would ask you to view through a red filter, you would set RGBS to 4.
I more than likely have made this more confusing with my explanation.
I humbly enlist others in this thread to make this more succinct.

Regards,

Bryan

gab2409

09-15-06, 10:13 PM

this thread is a wealth of information, albeit over my head. from what I've gathered, it looks like I would see a huge improvement by changing:

red push: RYR/RYB/GYR/GYB settings: 14-15-6-4

is this correct? I'm a little confused at this point, but this seems to be the bulk of the change.

gab2409

09-16-06, 02:14 PM

Dr. Spankenstein

09-16-06, 05:21 PM

Only two suggestions. Try (if you haven't already) raising the BRIGHTNESS a little, but no so the blacks turn grey/foggy. Maybe a boost to CONTRAST, but check (say with the THX Optimizer) that your peak whites are not "blooming", meaning the scanning beam is smearing/growing beyond it's normal bounds.
Otherwise, you could adjust the 2170P-4 #18 GAMM setting to either 1 or 2, but know that by doing this you are raising the brightness in the midtones and this can only serve to take away "depth" from the picture by reducing contrast.

Hope this helps,

Bryan

DSperber

09-16-06, 05:30 PM

ok, I calibrated it using 14-15-6-4, set the tv to pro, and turned off vm. the colors seem really accurate, but the picture looks pretty dark(on dvd and cable). is this just because I'm not used to a digital set?Did I miss your model? XBR960? Something else?

You should also set SHARPNESS=MIN on all non-SD inputs, in addition to VM=OFF and MODE=PRO. You might need some sharpness help on SD inputs.

I have my "color correction" values set at 13-15-5-4, which were initially recommended prior to being revised to 14-15-6-4 as you've set. I'm happy with the original set of values. These, along with my setting of color temp = COOL, and color axis = DEFAULT, plus picture=35, brightness=32, color=31 and hue=0, produces absolutely perfect overall picture to my eyes on my XBR960 (perhaps because of "cool" color which I much prefer).

But don't expect the picture to look good to you (i.e. not dark) in "broad daylight" in a room with outside light from the sides or rear of the room. The picture is supposed to be optimally viewed in a dark room. And then it won't look dark... it will look gorgeous. You can't even properly and fairly adjust it except in a dark room, like you are going to watch it.

Furthermore, you will probably find that DVD and non-HD sources might require a separate setup from true 720p/1080i sources (e.g. HD cable channels). In my setup, I've got my DVD player going to INPUT5 via component video, my cable DVR going to INPUT6 via component video (for HD channel viewing only), and my "standard" SD source going to INPUT1 (DirecTV) via S-video and INPUT3 (DVR via S-video for non-HD channels which I very rarely watch on this input). That gives me the freedom to have different user menu settings for each type of input.

For example, my INPUT5 settings (for DVD) have picture=34, brightness=36, color=36, and hue=G1 (probably because of my DVD player's settings), along with DRC=Cinemotion.

And INPUT1 has picture=39, bright=33, color=34, hue=0, sharp=21, and DRC=Progressive. INPUT3 has picture=40, brightness=34, color=34, hue=0, sharpness=19, and DRC=Progressive.

But to each his own settings. Whatever looks good to you.

Dr. Spankenstein

09-16-06, 06:13 PM

Quote DSperber:

"But don't expect the picture to look good to you (i.e. not dark) in "broad daylight" in a room with outside light from the sides or rear of the room. The picture is supposed to be optimally viewed in a dark room. And then it won't look dark... it will look gorgeous. You can't even properly and fairly adjust it except in a dark room, like you are going to watch it."

I knew I forgot to mention something... ;)
Good advice to all!

Cheers,

Bryan

ioannisds

09-17-06, 05:52 PM

Hi. I'm kinda new to the forum, and have gleaned quite a bit of information from this thread. I own a 34HS420, and I would like some help with the geometry. Please, please, please forgive me if this has already been asked 100 times already, but is there a step-by-step geometry/overscan procedure somewhere in this thread? I've done some adjustment in the CXA2170D-1 and 2 service menus, and then I start looking over this thread and feel like I might need to go back and tweak some more. I have not touched the MID1-3 menus, as I am still a bit unclear on what they do. And then I read about adjusting raster, and my head about exploded. Tell me if I understand the procedure correctly:

1)Adjust raster.

2)Adjust geometry using CXA2170D-1 and 2.

3)Tweak MID1-3? Based on input? Picture size?

Again, I'm sorry if this kind of thing has been asked before. I've been looking over this thread for a couple of hours, and my brain hurts. I can find bits and pieces of what I'm asking throughout, but no "do this, then do this, then do this, and you're done". Thanks for taking the time to read my post, and feel free to rip me a new one if I deserve it.

Mathesar

09-17-06, 11:17 PM

Ugh.. Tonight I was touching up the geometry in the service menu on my 34XBR960N (using a certain screen in a videogame as a test image) After I was done I decided to check how a news ticker looked so I switched over to the Cable input and proceeded to change channels ...not realising I was still in the service menu(!) Im not sure what I pressed but the TV shut off and came back on with the initial setup screen, it erased all my favorites / picture settings / made me auto program all the channels etc., Any idea what button combo would do this? My edited geometry / convergence settings are still in tact at least, I just hope I didn't reset anything else picture related in the service menu(?)

DSperber

09-17-06, 11:42 PM

is there a step-by-step geometry/overscan procedure somewhere in this thread?

There is this one (http://www.avsforum.com/avs-vb/showthread.php?p=6346890&&#post6346890)

which is also pointed to by this one (http://www.avsforum.com/avs-vb/showthread.php?p=6615783&&#post6615783)

which was pointed to by this one (http://www.avsforum.com/avs-vb/showthread.php?p=6723557&&#post6723557)

which was pointed to by this one (http://www.avsforum.com/avs-vb/showthread.php?p=7591682&&#post7591682)

Note that my posts deal with the XBR960 and I have no specific firsthand experience with the HS420, but I'm expecting the service menu to be at least very similar if not almost identical. If I'm wrong, hopefully some HS420 owner will correct the following.

If the service menus are similar then if you follow the recipe described by my series of posts you will get into the 2170D-1 service menu group to adjust VPOS and VSIZ (vertical position and vertical size of the background raster), the 2170D-2 group to adjust HPOS and HSIZ (horizontal position and horizontal size of the "background raster"), and the MID3 group to adjust VDHP, VDHS, VDVE, and VDVS (horizontal position, horizontal size, vertical position, and vertical size, respectively) of the "foreground image". MID3 is then used to manipulate the "foreground image" on top of the "background raster" for the 720p/1080i inputs... at least on the XBR960.

If MID3 isn't present, or doesn't work, you might look at MID1. This group (which I believe has DDHP, DDHS, DDVE and DDVS instead) was present on some lower-end or older XBR, XS and HS models so MID1 might be what they used on the HS420.

Anyway, if you have a proper test pattern input (either from DVE, Avia, or something like the OVERSCAN.ZIP image attached to my third post referenced above) you should be fully able to adjust your set's overscan without any outside help.

Oliver Deplace

09-17-06, 11:55 PM

Mathesar

09-18-06, 12:22 AM

Sounds like the "8-enter" user settings reset. I tried it once and it cleared the channel memory, so I had to re-scan and that took a ½hr. :(

Yea thats actually what it was doing when I posted the first message lol , now that its done I think everything is Ok (picture looks fine) , I just had to re-do my picture settings per video input etc. I guess it could of been worse!

I was able to improve a pretty noticable geometry error in the lower right corner (been bugging me since I bought the TV in Feb) so im happy ;)

ioannisds

09-18-06, 11:57 AM

There is this one (http://www.avsforum.com/avs-vb/showthread.php?p=6346890&&#post6346890)

which is also pointed to by this one (http://www.avsforum.com/avs-vb/showthread.php?p=6615783&&#post6615783)

which was pointed to by this one (http://www.avsforum.com/avs-vb/showthread.php?p=6723557&&#post6723557)

which was pointed to by this one (http://www.avsforum.com/avs-vb/showthread.php?p=7591682&&#post7591682)

Note that my posts deal with the XBR960 and I have no specific firsthand experience with the HS420, but I'm expecting the service menu to be at least very similar if not almost identical. If I'm wrong, hopefully some HS420 owner will correct the following.

If the service menus are similar then if you follow the recipe described by my series of posts you will get into the 2170D-1 service menu group to adjust VPOS and VSIZ (vertical position and vertical size of the background raster), the 2170D-2 group to adjust HPOS and HSIZ (horizontal position and horizontal size of the "background raster"), and the MID3 group to adjust VDHP, VDHS, VDVE, and VDVS (horizontal position, horizontal size, vertical position, and vertical size, respectively) of the "foreground image". MID3 is then used to manipulate the "foreground image" on top of the "background raster" for the 720p/1080i inputs... at least on the XBR960.

If MID3 isn't present, or doesn't work, you might look at MID1. This group (which I believe has DDHP, DDHS, DDVE and DDVS instead) was present on some lower-end or older XBR, XS and HS models so MID1 might be what they used on the HS420.

Anyway, if you have a proper test pattern input (either from DVE, Avia, or something like the OVERSCAN.ZIP image attached to my third post referenced above) you should be fully able to adjust your set's overscan without any outside help.

Thanks so much for pointing me in the right direction. This looks like exactly what I was looking for. Can you maybe explain to me how adjusting the foreground image over the background raster is different/necessary? I was able to get my geometry and overscan looking pretty good without ever diving into the MID3 menu. I just adjusted HPOS, HSIZ, VPOS, and VSIZ until I got my crosshatch well centered and sized, and then did some touch-ups to correct a little vertical bowing and the like. Reading your posts, it would appear that I did not go about this in the right way. Thanks again for your help.

Mindman

09-18-06, 12:19 PM

KenTech (or anyone else who can help),

Recently SurfingMatt27 helped me narrow down an issue with bleeding colors on my Sony KV-30HS420. Basically, in some parts of the screen, I would see red "outlines" off to the left of certain images. I adjusted the main focus item and fixed some of the problem. Now all that's affected are the corners. The left corner has a bad red bleed, the right has a green bleed. Adjusting the other focus items seem to alleviate the problem in one part of the screen while messing up the another part.

The other suggestion is a convergence issue. I have tinkered with the D-CONV settings and those don't seem to help at all. So, I have two questions. Is there anything else I can do in the service menu to fix this? And, if not, what do I need to tell a service tech to get the problem fixed correctly? Do magnets need to be adjusted?

Thanks a million!!

Istari1

09-18-06, 04:15 PM

RWetmore

09-18-06, 05:04 PM

Dr. Spankenstein

09-18-06, 05:18 PM

Hello folks,

I am needing a process/test to verify if the image on my screen is being scaled correctly by the set and/or DVD player. The reason I ask is, KenTech has stated many times that if you deviate from the MID menu settings regarding vert. and horiz. size that the resultant image will be compromised. Unfortunately, the Service Menu settings for the MID menus are of diminished use since other's and my own set (34xs955) have quite differing menu options in the MID categories and often the values that are stated in the SM won't work because, as one poster put it, "the picture goes to garbage". To me, it looks like it either loses horz. sync or the picture immediately is reduced to a 1" wide vertical band.

I'm just looking for a way to tell if I'm within the parameters that alow the set to scale the image correctly. KenTech had mentioned some patterns, but I have had a hard time locating that post.

I have been trying to use AVIA to optimize my HDMI input for my HD DVD player. Which leads me to another question, how valuable is this going to be if the DVD player is doing upscaling/deinterlacing? The only resolution options are: 480p, 720p and 1080i.

For my review, can someone explain which MID settings are global or resolution/input dependent? MID1 #1-4, 8-11 and MID2 #1-4. Sorry, some of us don't have any relevant MID3 categories. With this help, I think I can compile an alternate list of settings that could be of use for us rogue 955/420/960 owners.

Thanks in advance,

Bryan

gab2409

09-18-06, 09:12 PM

Furthermore, you will probably find that DVD and non-HD sources might require a separate setup from true 720p/1080i sources (e.g. HD cable channels). In my setup, I've got my DVD player going to INPUT5 via component video, my cable DVR going to INPUT6 via component video (for HD channel viewing only), and my "standard" SD source going to INPUT1 (DirecTV) via S-video and INPUT3 (DVR via S-video for non-HD channels which I very rarely watch on this input). That gives me the freedom to have different user menu settings for each type of input.

I have the 30hs420 btw.

2 things:
1. why does the picture look vastly different on the different picture modes(vivid, etc), when they are set exactly the same as pro? I did the SM tweaks...are they video mode specific? do they just change for the mode I am on?

2. why do you have your dvd player connected to a non-hd input? for a non-hd dvd player, will it matter either way? I have mine connected to 6(an hd input).

3. lastly, my picture still seems too dark. I've played with the brightness and contrast. when I turn either up, it starts to look washed out. here are my settings:

In 2103-1 (change these to)
YLEV 19
CLEV 25
SCON 8
SCOL 6
SHUE 11
YDLY 0
SHAP 4
SHFO 3
PREO 1
BPFO 3
BPFQ 2
BPSW 1
TRAP 0
LPF 1
AFCG 0
CDMD 3
SSMD 2
HMSK 1
HALI 0
PPHA 7
CBO1 26
CRO1 43
CBO2 21
CRO2 39
ATPD 0
DCTR 0

In 2170P-1:
YOSW 0
TCOF 0
YOF 7
CBOF 37
CROF 37
SBRT 29
RDRV 30
GDRV 28
BDRV 23
RCUT 43
GCUT 24
BCUT 22
WBSW 0
SBOF 7
RDOF 30
GDOF 31
BDOF 32
RCOF 30
GCOF 31
BCOF 32
DCOL 0

2170P-2

PICO 1
RGBS 7
BLKB 3
RGBL 2
YLMT 3
AGNG 0
AKBO 0
CLPP 3
CLPG 0
CLPS 0
PPAD 3
SYNP 0
HVBT 0

2170P-3

SYSM 3 (for dvd set to factory default for hd use my settings)
YMLV 0 (for dvd set to factory default for hd use my settings)
VMCR 0 (for dvd set to factory default for hd use my settings)
VMLM 0 (for dvd set to factory default for hd use my settings)
VMFO 0 (for dvd set to factory default for hd use my settings)
VMDL 0 (for dvd set to factory default for hd use my settings)
SHOF 1 (for dvd set to factory default for hd use my settings)
SHFO 1
PROV 3
F1LV 0
LTLV 0
LTMD 1
CTLV 0
UBOF 4 (For HDTV, Varies for each input so adjust accordingly)
UCOF 0
UHOF 0
MIDE 20
VM 0 (for dvd set to factory default for hd use my settings)
VMH 0 (for dvd set to factory default for hd use my settings)
VMM 0 (for dvd set to factory default for hd use my settings)
VML 2 (for dvd set to factory default for hd use my settings)
VGAP 5
VGAS 5
VGAB 5
VGAC 5
VGAV 5

2170P-4:

YCON 1
SPIC 0
SCOL 36
SHUE 30
SPIO 7
SCLO 7
SHUO 7
UPIC 31
UBRT 31
UCOL 31
UHUE 31
USHP 31
UTMP 1
RYR 14
RYB 14
GYR 6
GYB 4
GAMM 0
GAMS 0
GAMR 0
GAMG 0
GAMB 0
BLK 0
DCTR 0
APED 0
DSBO 7
IDSW 0
ABLM 0
ABLT 0
SPOF 0
DPSQ 1
LRGB 3

any thoughts?

corlay

09-18-06, 09:25 PM

MID3 is then used to manipulate the "foreground image" on top of the "background raster" for the 720p/1080i inputs... at least on the XBR960.

IS there a concensus amongst this thread's participants?

Does altering MID1-3 settings degrage overall picture focus/sharpness?
Only *cetain* MID settings?

KenTech reported *horizontal* degradation, but not vertical.
Does this mean that Horizontal MID controls are ok to alter as needed?

--

Also...

I was planning to maybe explore using the 'Zoom' picture mode for displaying 2.35:1 source content on my 30HS420 16:9 screen. This would involve some heavy altering of MID settings for that mode, I suspect. But since I never use 'Zoom' to view 4:3 source content (I either use 'Normal" or 'Wide Zoom' for that...) I thought it might be nice to cater 'Zoom' to be used specifically for 2.35:1.

Has anyone else tried this?

--
corlay

Dr. Spankenstein

09-18-06, 10:09 PM

Quote corlay:
Also...

I was planning to maybe explore using the 'Zoom' picture mode for displaying 2.35:1 source content on my 30HS420 16:9 screen. This would involve some heavy altering of MID settings for that mode, I suspect. But since I never use 'Zoom' to view 4:3 source content (I either use 'Normal" or 'Wide Zoom' for that...) I thought it might be nice to cater 'Zoom' to be used specifically for 2.35:1.

Has anyone else tried this?

Uh oh....you wouldn't happen to be one of those "don't like black bars" type, would you? :rolleyes:

Kidding aside, I would imagine that that ammount of stretching of the image would seriously affect the picture quality (softening quite a bit)!

Regards,

Bryan

Dr. Spankenstein

09-18-06, 10:14 PM

gab2409,

The first thing I would try would be to raise the value in GAMM to 1, maybe 2. This will tend to "flatten" the image by brightening the midtone values. If you find you like something between 0 and 1, you can adjust the parameters of GAMR, GAMG and GAMB to equal levels. For example, KenTech suggested that if you could use GAMM @0 then set GAMR, GAMG,GAMB to 3,3,3.

You could start there and season to taste. :p

Enjoy,

Bryan

gab2409

09-18-06, 10:18 PM

interesting. thanks a lot, I'll let you know how it works

Dr. Spankenstein

09-18-06, 10:32 PM

I think you'll like the effect if you don't lay it on too thick.
Please, let me know.

Cheers,

Bryan

corlay

09-18-06, 10:36 PM

Kidding aside, I would imagine that that ammount of stretching of the image would seriously affect the picture quality (softening quite a bit)!

no, not that...

But just to *optimize* the 'Zoom" picture mode for 2.35:1.

1. approach 0% overscan @ the top & bottom
2. and ~10% overscan @ the sides

this would increase the image porportionately, at the expense of some side image content.

right now, *none* of the picture modes accomodate 2.35:1 very well, IMO.

Oliver Deplace

09-19-06, 03:04 AM

IS there a concensus amongst this thread's participants?

Does altering MID1-3 settings degrage overall picture focus/sharpness?
Only *cetain* MID settings?

KenTech reported *horizontal* degradation, but not vertical.
Does this mean that Horizontal MID controls are ok to alter as needed?

--

Also...

I was planning to maybe explore using the 'Zoom' picture mode for displaying 2.35:1 source content on my 30HS420 16:9 screen. This would involve some heavy altering of MID settings for that mode, I suspect. But since I never use 'Zoom' to view 4:3 source content (I either use 'Normal" or 'Wide Zoom' for that...) I thought it might be nice to cater 'Zoom' to be used specifically for 2.35:1.

Has anyone else tried this?

--
corlay

If you alter Zoom you will alter Wide Zoom and, more importantly, Full. So, you'll have to use a different scan rate for your custom mode. You could use 480p for Full and 480i for Zoom (or vice versa). It might be a pain to keep changing the player back and forth between progressive and interlaced. If you can do it remotely, then that's not so bad.

The MID setting that controls vertical size and position can have a deleterious effect. The degradation is subtle, so you might find the trade-off worthwhile. The resulting image will be far from normal anyway.

Normally, you would do your positioning and sizing entirely with 2170D-1 & 2.
Horizontal work might require some MID adjustment and they didn't thicken the lines in test patterns (apart from what one might expect when stretching the image).

Zoom would only require MID horizontal stretching and centering.

KenTech

09-19-06, 03:15 AM

Does altering MID1-3 settings degrage overall picture focus/sharpness?
Only *cetain* MID settings?Absolutely NO MID1-3 settings affect beam focus or real image sharpness. (However, some settings in MID5 affect the video-frequency response, almost like a graphic equalizer for video.) MID1-3 settings appear to affect digital pixel-mapping of the video onto the raster, etc.KenTech reported *horizontal* degradation, but not vertical. Does this mean that Horizontal MID controls are ok to alter as needed?Terminology problem! Getting those "magic" vertical-size ratios wrong causes vertical degradation, smearing horizintal lines. The AVIA Sharpness screen shows it very well. (I can't say if altering the horizintal size affects horizintal sharpness, blurring vertical lines.) The positioning settings don't seem to affect the mapping (sampling) ratios, and so I don't expect, nor do I see, degradation for small tweaks of position. But, knowing the purpose of those digital-sampling ratios established by the size settings, I would not change them and would search for tweaks for picture size/scaling settings in 2170D-1 and -2.

Again, the crossed set of 4 X 4 dark lines passing thru the center of the AVIA Sharpness pattern is really good for establishing the existence of any problems. With the mode set to Progressive or CineMotion for component-video, the boundaries of the horizintal lines should be absolutely sharp, and even more so thru HDMI.

Oliver Deplace

09-19-06, 03:22 AM

Ken:
MID vertical positioning does degrade the image. It's not as obvious, but it displays the same "cyclic" behavior.

DSperber

09-19-06, 05:13 AM

Can you maybe explain to me how adjusting the foreground image over the background raster is different/necessary? I was able to get my geometry and overscan looking pretty good without ever diving into the MID3 menu. I just adjusted HPOS, HSIZ, VPOS, and VSIZ until I got my crosshatch well centered and sized, and then did some touch-ups to correct a little vertical bowing and the like. Reading your posts, it would appear that I did not go about this in the right way. As I mentioned previously, my comments refer to the XBR960, which has been observed to have a somewhat different Service Menu set of adjustments than the lower-end models and even earlier XBR models. So if what I've said doesn't quite apply to your model, well that may just be true.

But as best as I can describe it, I think the Sony approach to geometry/overscan adjustment is sort of like with Adobe Photoshop where you have a "canvas" in the background (<-> background layer), on which the "image" is overlaid in the one foreground layer (on top of the background layer). The visible portion of the foreground image layer can never exceed the limits set by the dimensions of the background layer underneath it. In Photoshop you can have multiple foreground layers, but on the Sony sets we're limited to one foreground layer... namely the image itself.

So in general terms, your use of the Sony Service Menu items I've described facilitates your attempt to try and achieve "optimal" settings for both background layer and foreground layer geometry and overscan.

First, you try and maximize use of screen real estate by adjusting the background layer (i.e. "raster") to just fill the screen, meaning the foreground image on top of it if enlarged to exactly match the size of the background raster will have 0% overscan. If the background layer is smaller than the entire screen, you will not be able to spread the image to the extreme edges of the screen since the background layer (i.e. "canvas") limits the outer bounds of that foreground image layer. Spreading the background layer to the four edges of the screen guarantees a 16x9 image and if you use a good overscan test pattern you will be able to tweak for uniform overscan around all four edges.

Second, you adjust the foreground image layer to get whatever overscan percentage you want (e.g. 1-3%) uniformly around all four edges, since you now know that the background layer completely fills the screen. Since the foreground image layer can never exceed the limits imposed by the background layer underneath it, if you spread the foreground image WIDER and TALLER than the background layer underneath it, you will in essence be imposing whatever overscan percentage you want on the image, i.e. "cropping" the foreground image by making some extreme outer percentage of it invisible (by exceeding the limits of the background layer).

Again, you need to be using a good stationary overscan test pattern all the while, so that you can see exactly the effect of your tweaks (which, of course, should be applied just one click at a time so that you can see exactly what you are doing).

And this is a delicate process, so take it slow.

Also note that if you adjust for 0% overscan (maximizing visible broadcast image), everything you see will be slightly smaller than if you implement a non-zero overscan (where the overall displayed image will appear slightly larger). This is because if you crop 1-3% of the outside of the image, but still fill 100% of the screen with the 97-99% that's left, the displayed remaining portion of the cropped image will by necessity have to be magnified slightly to fill 100% of the screen. So everying on that image will look a little bigger than if you have 0% overscan and thus display 100% of the broadcast image on 100% of the screen.

DSperber

09-19-06, 05:38 AM

I have the 30hs420 btw.

2 things:
1. why does the picture look vastly different on the different picture modes(vivid, etc), when they are set exactly the same as pro? I did the SM tweaks...are they video mode specific? do they just change for the mode I am on?I honestly don't know. I suspect it might be possible for some tweaks to be video mode specific, and others to be generic for all modes. I guess the only way to know this for sure is to scroll through ALL of the SM values you've written down while in PRO, and then scroll through them again while in VIVID, and see if you see a difference.

And I'm also assuming your set has the Advanced Video -> Mode Memory setting, and that you have selected "ON" so that you can customize EACH video input separately and that your mode and other settings will be memorized and applied to EACH video input separately. I know this lives in the XBR960 and "ON" is how this should be set.

I also know that PRO eliminates essentially all (if not all) of the mode-induced preset/bias service menu values as well as user menu values, and puts everything in your hands and back to "31" neutral settings. From there, I don't know why you'd ever go back to VIVID. That's why you're setting up with PRO, to let you optimize your picture for each video input.

But as to your original question of whether you can sort of re-adjust VIVID (to be used sort of as a second PRO, perhaps with slightly different values so that you can "compare" tweaks before committing them in PRO), I honestly thought that was possible.

2. why do you have your dvd player connected to a non-hd input? for a non-hd dvd player, will it matter either way? I have mine connected to 6(an hd input).INPUT5 and INPUT6 are identical, and are both HD inputs supporting 480p, 720p and 1080i via component video input. Why do you think INPUT5 is not HD?

I have my non-HD DVD player on INPUT5 (at 480p) so that my settings for that input are specific for my DVD player and its video output characteristics at 480p. I have my DVR on INPUT6, and I only watch 720p/1080i HD channels through it. But by being on INPUT6 (separate from my 480p INPUT5) I can again adjust the XBR960 for optimal picture as put out by the DVR.

3. lastly, my picture still seems too dark. I've played with the brightness and contrast. when I turn either up, it starts to look washed out. here are my settings:...Well I really can't say. It's a 30HS420, not a 34XBR960, so I wouldn't expect the picture quality to be the same... even optimally adjusted.

Also, adjustment results are in the eye of the beholder. My brightness/contrast values might be inappropriate for your eyes and your set, and vice versa. I do notice that my "red push" adjustments are slightly different from yours, but that shouldn't affect your "too dark" observation.

Are you watching in a dark room, or broad daylight?

Don't really know what to say, assuming you're wanting to stay with the 30HS420. For example, the KDSR-50XBR1 has AMAZING brightness!

ioannisds

09-19-06, 01:19 PM

Thanks for the information DSperber. I think I've got it down now.

gab2409

09-19-06, 03:30 PM

The first thing I would try would be to raise the value in GAMM to 1, maybe 2. This will tend to "flatten" the image by brightening the midtone values. If you find you like something between 0 and 1, you can adjust the parameters of GAMR, GAMG and GAMB to equal levels. For example, KenTech suggested that if you could use GAMM @0 then set GAMR, GAMG,GAMB to 3,3,3.

gamm=3 worked wonders...that fixed my problems. thanks a lot.

corlay

09-19-06, 05:17 PM

Normally, you would do your positioning and sizing entirely with 2170D-1 & 2.

Yes, I *tried* to do all of my size/position alterations in D-1 & D-2. But, there were "offsets" between different inputs (ie. RF, 4880i; V5/V6480p; etc...) that I needed to correct, to get the image properly centered/sized. I ended up optimizing D-1 & D-2 for V5/V6, 480p, Full as my dvd player quality is the most important to me; and then I tweaked the MID settings for RF, 480i, widezoom and normal to center and size my cable channels. I don't have HD at all at the moment...

Horizontal work might require some MID adjustment and they didn't thicken the lines in test patterns (apart from what one might expect when stretching the image).

Zoom would only require MID horizontal stretching and centering.

So, is the concensus that all MID Vertical size/position settings *do* degrade the image quality (albeit very subtle...) and that the Horizontal settings do not?

--
corlay

Ectospheno

09-21-06, 01:50 AM

Ecto,

Actually, RGBS is an adjustment parameter in the Service menu. It will turn on/off any combination of the color guns.
The adjustments you will use to correct the color decoder is: 2170P-4 #7RYR/#8RYB/#9GYR/#10GYB. Most people have found that you can get quite close using the settings 14-14-6-4 for the above parameters in the 2170P-4 menu.

Thanks for the help. The rest of the family was out of the house tonight so I worked up the courage to explore the service codes. Easier than I thought it would be. 14-14-6-4 is indeed much better than it was. I'm going to leave it here for a week or so to see how I like it.

On a different note, I'll have to disagree with others here on the sharpness settings. I don't see how people lower their sharpness to 0 and turn VM off. I'm pretty sure video game text isn't intended to be blurry. Just my $.02.

GlenC

09-21-06, 02:12 AM

........ I don't see how people lower their sharpness to 0 and turn VM off. I'm pretty sure video game text isn't intended to be blurry. Just my $.02."0" is not always the setting. You need to look closely at dark vertical lines on a medium background and set sharpness to the point the white shadow to the right of it just disappears. Lowering sharpness too much can soften the picture too much.

justsc

09-21-06, 12:25 PM

Thanks for the help. The rest of the family was out of the house tonight so I worked up the courage to explore the service codes. Easier than I thought it would be. 14-14-6-4 is indeed much better than it was. I'm going to leave it here for a week or so to see how I like it.

On a different note, I'll have to disagree with others here on the sharpness settings. I don't see how people lower their sharpness to 0 and turn VM off. I'm pretty sure video game text isn't intended to be blurry. Just my $.02.
I tried to go along with the "no VM" majority but I just could not enjoy my viewing experience that way. I have pretty much settled on ClearEdge at Low. I believe my Sharpness setting is at around 25%.

GlenC

09-21-06, 12:49 PM

I tried to go along with the "no VM" majority but I just could not enjoy my viewing experience that way. I have pretty much settled on ClearEdge at Low. I believe my Sharpness setting is at around 25%.This is consistant with those who want some extra sharpness in the picture, without overly distorting the image. Typical with longer viewing distances, the 34" is nearly 25% the viewing area of a 65", but to see the same detail, you need to sit at 1/2 the distance, not many sit 5-6 feet from the TV. I sit 12' from a 110" screen, don't need or want any edge enhancement.

Mindman

09-21-06, 01:39 PM

Look for the posts on adjusting LANDING I bet this is your issue.

On my TV there actually aren't any LANDING settings in the SM, I checked 3 times to be sure. I don't think that's it though. There isn't any discoloration. Just some bleeding, if you will. I am trying to figure out the best way to describe it. I will have a screenshot of my TV today or tomorrow.

KenTech

09-21-06, 01:45 PM

Why does the picture look vastly different on the different picture modes(vivid, etc), when they are set exactly the same as pro? I did the SM tweaks...are they video mode specific? do they just change for the mode I am on?I honestly don't know. I suspect it might be possible for some tweaks to be video mode specific, and others to be generic for all modes. I guess the only way to know this for sure is to scroll through ALL of the SM values you've written down while in PRO, and then scroll through them again while in VIVID, and see if you see a difference.The relationships between the picture modes and service-mode settings are made very plain in the service-data chart. Where there are multiple columns labeled with picture-mode names, the settings are picture-mode-specific. Don't even think about making changes in service mode without one of these charts at hand!

Also, from article #05:
The various picture-mode settings in late-model HS/XS/XBR Sony sets are not set in stone. They are simply pointers to a large group of presets that can be completely reconfigured in service mode. For example, I have no use for the factory's settings for Vivid and Standard - they're way too garish and have too many “enhancements” for my taste. So I have Left Pro pretty much alone, as a reference. Then I have commandeered Vivid for experimentation, and set up Standard and Movie to be equivalent to Pro, but with higher gamma, to resurrect murky video.
Bottom line: The appearance of each picture mode is overwhelmingly determined by service-mode settings, not by user-menu settings. You get to change them, if you wish. See the service-data chart for your set to see how to do this. (These charts have been posted multiple times. See forum Contents in message #1 to track them down. Use these Contents to identify other postings that will help you; some of these issues have been discussed over and over again.)

Mindman

09-21-06, 01:47 PM

Can you supply us with a picture of the screen?

Yes, shortly. I gotta get to the bottom of this and find out what is going on, lol.

RWetmore

09-21-06, 02:52 PM

Thanks for the help. The rest of the family was out of the house tonight so I worked up the courage to explore the service codes. Easier than I thought it would be. 14-14-6-4 is indeed much better than it was. I'm going to leave it here for a week or so to see how I like it.

Spend some time with different settings. I think I have finally settled on 14-15-6-4. No setting seems perfect...some compromise must be made.

On a different note, I'll have to disagree with others here on the sharpness settings. I don't see how people lower their sharpness to 0 and turn VM off. I'm pretty sure video game text isn't intended to be blurry. Just my $.02.

It seems like it shouldn't be blurry unless it is out of focus or the contrast (picture) is set too high. I have sharpness at minimum and Velcocity SCan Mod. turned off, and nothing except very small text is blurry.

Ectospheno

09-21-06, 09:16 PM

Ectospheno

09-21-06, 09:19 PM

It seems like it shouldn't be blurry unless it is out of focus or the contrast (picture) is set too high. I have sharpness at minimum and Velcocity SCan Mod. turned off, and nothing except very small text is blurry.

My contrast is only a little higher than I'd like it to be because my wife doesn't like it at the setting I prefer. It isn't high enough to be causing that problem though.

SurfingMatt27

09-21-06, 09:24 PM

Personally i think contrast looks best in the middle, i use brightness to brighten the picture.

dlrzrwlf

09-21-06, 09:38 PM

Great thread, very helpful.

I was wondering if there is a specific setting that affects horizontal size while in zoom mode only. I have a kd36xs955 and I have successfully fixed the convergence/picture issues that I had with the set, however I can not seem to adjust the horizontal size of the picture in zoom mode without adjusting all other modes. I have a service manual, and I have searched this thread, but I still can't seem to figure it out. Thanks

KenTech

09-21-06, 10:52 PM

I can not seem to adjust the horizontal size of the picture in zoom mode without adjusting all other modes. I have a service manual, and I have searched this thread, but I still can't seem to figure it out.Nor has anyone else! Bottom line, there's only one legit setting parameter for width (HSIZ), and it is common for normal (Full) and for zoomed modes. I wanted to use Zoom to make a 4:3 picture broadcast in HD (normally pillarboxed) fill the screen. Of course, it overfills the screen, and there's no remedy without screwing up the Full mode.

KenTech

09-21-06, 10:59 PM

Personally i think contrast looks best in the middle, i use brightness to brighten the picture.Huh? No two people can really compare their Contrast settings because what each person experiences depends so heavily on how the TV is set up in service mode. Further, contrast sets the maximum brightness, supposedly white, and different people react differently to this setting, depending on the room illumination.

On the other hand, "Brightness" in the user menu is actually black level, and it can be set correctly for any particular program or DVD. Set too high, and there are no deep blacks; too low, and you've got black "holes" instead of shadow detail. (Depends somewhat on room brightness.)

If there is a discretionary setting for picture "comfort," it would be Contrast, not Brightness. Black should be black, but how bright the whites and light colors are is up to you and your eyes.

SurfingMatt27

09-22-06, 12:49 PM

Huh? No two people can really compare their Contrast settings because what each person experiences depends so heavily on how the TV is set up in service mode. Further, contrast sets the maximum brightness, supposedly white, and different people react differently to this setting, depending on the room illumination.

On the other hand, "Brightness" in the user menu is actually black level, and it can be set correctly for any particular program or DVD. Set too high, and there are no deep blacks; too low, and you've got black "holes" instead of shadow detail. (Depends somewhat on room brightness.)

If there is a discretionary setting for picture "comfort," it would be Contrast, not Brightness. Black should be black, but how bright the whites and light colors are is up to you and your eyes.

I know that.. but thanks for the lecture ;)

net17

09-22-06, 04:59 PM

Has anyone used the HDNet test pattern to test overscan? If so what numbers are you using top, bottom, and side to side...I have heard different ways to go about this...ie 5 top and bottom, and 10 side to side...anyone else?

DSperber

09-22-06, 05:30 PM

I went with Medium and 20%. Too many 480p video games on PS2 and Gamecube had crappy text otherwise.Well this probably explains why I opt for sharpness=MIN and VM=OFF and think it looks stunningly clear and brilliant.

I do not play ANY games, and hardly ever watch DVD's. My XBR960 is used almost exclusively for 720p/1080i from Comcast HD or OTA HD.

The proof-positive for me: US Open Tennis looks like they're playing right inside my set. The white lines are absolutely perfect with no ringing or red edge or anything other than pure white. You can see every zit and whisker on skin, every bead of perspiration on face and arms and legs, all in microscopic detail. Next would be "Rome" on HBO, and then "Fifth Element" in HD.

Shawn Watson

09-22-06, 07:28 PM

How do I fix the overscan on a Sony Bravia 50KDFE50A12U? :D

TheGodfather

09-23-06, 06:43 PM

I've had a little mess-up with my MID3, HSIZ, HPOS, VSIZ, and VPOS settings. Can anyone supply me with the stock ones, so that I may startover? When watching CNN, it isn't taking up the whole screen...

I dont have a 4:3 overscan pattern. If anyone would supply me one, I would really aprreciate it.

Edit: I think I've fixed it, but I'd still appreciate a Overscan pattern :)

corlay

09-24-06, 11:46 AM

could someone explain how the 'Zoom'screen mode functions?

I had thought that it just takes *whatever* source video is input, and just magnifies that image to a pre-determined amount (relative to the screen center); thus increasing the overscan at all four screen edges (top, bottom, left, right).

But yesterday, I was monkeying around with different test images from DVE; and found that certain slides are not zoomed proportionately. For the 4:3 slide with circles and grid pattern, the image Zooms proportionately; so that the circles remain circles. But for the 16:9 equivalent pattern, Zoom does not zoom the image proportionately, and the circles get compressed in the horizontal direction.

Does this make sense to anyone?

my set is 30HS420.
DVD player outputs at the 16:9 setting.

raouliii

09-24-06, 01:40 PM

could someone explain how the 'Zoom'screen mode functions?
.............I was monkeying around with different test images from DVE; and found that certain slides are not zoomed proportionately. For the 4:3 slide with circles and grid pattern, the image Zooms proportionately; so that the circles remain circles. But for the 16:9 equivalent pattern, Zoom does not zoom the image proportionately, and the circles get compressed in the horizontal direction.

Does this make sense to anyone?

my set is 30HS420.
DVD player outputs at the 16:9 setting.
I believe the 16:9 patterns on DVE are anamorphic/enhanced for 16x9. Since you have your DVD player correctly set for 16:9, they are being sent to your set in a format that only displays correctly in FULL mode.

The WIDE ZOOM mode zooms in a non-linear way such that the left and right of the frame is zoomed leaving the center relatively untouched.

corlay

09-24-06, 08:02 PM

.The WIDE ZOOM mode zooms in a non-linear way such that the left and right of the frame is zoomed leaving the center relatively untouched.

just to clarify: I was inquiring about the 'Zoom' screen mode, and not 'Wide Zoom'...

Mindman

09-25-06, 12:09 AM

Okay guys, I have finally taken some pictures of my TV screen that clearly show the problem I am having! The pics are attached. Again I have the KV-30HS420.

The pics show the bottom left hand corner of my screen, and show a nice (ugly) red outline around the character's arm, which is the problem I've been having - red outlines. Since I haven't been able to accurately describe it until now, it's good to be able to show you guys so you can offer me some more advice. ;)

I have tried the focus controls (particularly DQP and DF) and I can get rid of the problem on the left side as shown in the pics at a price of losing focus, which obviously isn't acceptable. I should note that the right side of the screen has slight green outlines, but they don't bother me as they aren't anywhere near as bad as the left side. In both cases, the problem seems to be far more prevelant in the corners.

The D-CONV convergence settings don't seem to help any either. Does it need a magnet adjustment? It really seems like the colors are not aligned right, but again, only really in the corners. If it's not possible to fix via the service menu (which I am thinking is the case, unless there is something else I could try) what do I need to tell a Sony-certified tech to accurately fix my problem??

Thanks so much!!

TheGodfather

09-25-06, 02:56 PM

I'm guessing that that is a massive convergence issue.

Dr. Spankenstein

09-25-06, 07:32 PM

GlenC

09-25-06, 07:48 PM

The D-CONV settings work blue and red together. D-CONV severely off will show distinct Red/Green/Blue with a vertical line. If red is the only problem, call Sony service. If still under warranty, call Sony.

TheGodfather

09-25-06, 09:35 PM

Quick question- does work in the service menu void your warranty? If so, which part? Parts? Labor? Tube?

Mindman

09-25-06, 10:12 PM

It really does look like something that the convergence settings should remedy. The D-CON menu allows for 5 different areas of effect per Left and Right side: Entire side, upper middle upper corner, lower middle and lower corner. Have you tried putting up a crosshatch or dot pattern to see EXACTLY where the problem is? I'm sure it is more noticeable in the corners but sometimes the entire screen is off. With a crosshatch/dot pattern you can check the center. If the center's convergence is off, CADJ will center it up. THEN work out towards the sides.

A couple more questions:
Is the red shifted to the left on both sides of the screen or is blue offset (to the left) on the right side of the screen. This could indicate if an mechanical adjustment to the yoke is neccessary.
Are any of your D-CONV settings maxed (either 0 or 63)? sometimes you have to make compromises in one area to make room to ajdust another.

My corners aren't perfect either, but 95% of my screen is stellar. You can also lessen the severity of the color flare by getting your set out of "torch" mode! :eek: Meaning, Don't view in VIVID mode, turn down the CONTRAST and view in a properly darkened room (please note that I didn't say "dark"! I just recently experienced the "magic" of backlighting my 34SX955. It really takes the strain out of viewing and increases the percieved contrast.
I can tell you've got your TV maxed out since you took your photos in a brightly lit room and your screen was still uber-bright! Do your set and your eyes a favor and turn some settings down.

Hope this has helped,

Bryan

Thanks for the info. First of all, I do NOT have my TV on Vivid, or with high contrast or color or whatever else. ;) I am using Pro, with low contrast and very comfortable viewing conditions. For some reason or another it looks like I am using Vivid in the pics, but that's definitely not the case! :)

I just went and checked the convergence settings again and none of them are maxed out. I used the dot and a crosshatch pattern from this thread and I watched different parts of the screen as I adjusted them, and I didn't see any colors moving, and the red problem definitely didn't get better or worse in the game shown in the screenshots.

With the dot pattern, the dots seem to get a little blurrier/wider on the sides, though, but they too are not affected by the convergence settings. I can see all 3 colors on each dot, but they are more tightly packed in the center. If anything, I can seem some red off to the left side of the dots on the far left of the screen. Not much though.

With the crosshatch pattern, the lines in the center seem perfect. Again, they seem to seperate a little when getting closer to the sides like the dots, and are not affected by the convergence settings.

I don't have a good test to see if blues show off to the left on the right side of the screen, but I haven't seen that yet in my games. I have, however, seen SOME red off to the left on the RIGHT side of the screen, just not anywhere near as bad as the left side. That seems to show up better in games than when using the test patterns, and both the games and the test pattern are being displayed through my Xbox 360.

This is quite perplexing. ;) What else could it be???

Dr. Spankenstein

09-25-06, 10:41 PM

Sorry, I wasn't trying to imply that you would watch your set in "torch" mode, I made a poor assumption based on the photos.

If it is only red that seems to be off, I'll have to concur with GlenC. Time to call the service techs. Boo!

Regards,

Bryan

Bosniac

09-26-06, 06:37 AM

Hello, this message is directed towards anyone that can help me. I stumbled up this thread by looking at the net, and reading a LOT of info in this thread about the Service Menu, e.t.c.

Well, I was trying to adjust my horozontal, and vertical size of the screen, as my KD30XS955 seems to be too "zoomed" in, esp. when connected to an Xbox 360, and while playing some games.

Well, as unluck struck me, I have, entered the "infamous 7,9, ENTER code in sequence (please don't ask why), and all the parimeters have reset themselves. Now my TV is all screwed up.

Is there a way to set it to the way I bought it, or am I looking at some servie repair charges here.

Please, I am in desperate help.

Mindman

09-26-06, 11:14 AM

SurfingMatt27

09-26-06, 01:05 PM

That's what I am thinking as well. KenTech, SurfingMatt27, care to chime in on this issue as well?

Having a high contrast setting can effect convergance as well, But if your contrast is relatively low it should'nt be an issue.

I have some minor convergance issues on the sides but after a while when the tv warms upo they go away. Could you try that and wait untill it warms up then tell me if you still have issues?

SurfingMatt27

09-26-06, 01:11 PM

Mindman

09-26-06, 04:56 PM

Mindman i have the same issue on my sony in the bottom left hand corner, but it's so minor that i would'nt worry about it. On a CRT HDTV you are never going to get perfect convergance in the corners anyways so as long as the picture is fine in the middle and it's only your bottom corner that has the issue, i would'nt worry man. After a while you won't notice it. I've had my sony for over 3 years now going on 4 years this october not a single issue.:)

Good point. In fact, it makes me feel better about this knowing someone else has an issue with it too. You are are da bomb, man. Thanks for being so helpful with this here and in the PMs.

I would still like to hear KenTech's opinion... he has been mysteriously silent. ;)

SurfingMatt27

09-26-06, 08:02 PM

The truth is convergane will never be perfect on a CRT, you can get close but not perfection. having minor issue in the corners is normal.As long as it does'nt happen anywhere else then i would'nt worry.

Enjoy your tv!

Dr. Spankenstein

09-26-06, 08:50 PM

Bosniac,

Is luck would have it, you can download the service manual in .pdf from the first post in the thread. A bit of time invested and things should be o.k with the default settings.

While you are taking the time to systematically restore the defaults, I suggest you scan the other highlighted topics in the first post and apply some more refinements to your set while you are in there.

There are a couple of others who have done the same as you and it would be worthwhile to PM them and see if there were any stumbling blocks. (I purposely neglected to provide their names to encourage you to read this entire thread before proceeding any further. Not a punishment, just an education. It will make more sense when you dive in.)

Best of luck, brother.

Regards,

Bryan

Bosniac

09-27-06, 12:20 AM

KenTech

09-28-06, 01:43 PM

The pics show the bottom left hand corner of my screen, and show a nice (ugly) red outline around the character's arm, which is the problem I've been having - red outlines. . . . The D-CONV convergence settings don't seem to help any either.Agreeing with what others have said, I think this is most certainly a color-convergence problem. The only issue may be whether this amount is outslide the range of adjustment of the usual controls.

The primary service-mode parameter is D-CONV Nos. 11 and 10, LLBW and LLMB (if it is left side only). See section 2-3.4 in the servicve manual. If the issue extends along the bottome of the screen, you should probably fix the center-bottom first with #1, YBLW.

There may be an interaction with LANDING #1, LB. You might check to see if this is cranked to an extreme value, 128 being the center value (no effect).

If there is an anomaly in the deflection system or a badly-placed magnet doing this, you have little choice but to see what's up with magnets on this area. (Requires opening up the set.)

At least it's a horizontal-convergence issue, theoretically adjustable by you. Vertical fringing on horizontal edges is not adjustable in service mode.

I agree with Glen that a red-only issue is a defect. If you adjust convergence for the red, however, the defect will be likely much less visible, as blue fringing is not nearly so visible. When I have adjusted convergence, it was the red that got my attention, and I minimized the red fringing. The blue, never very visible in the first place, snapped into line as well. Hope yours does the same. It may not be a red-only issue -- just that the opposing blue fringe is so trumped by the red.

KenTech

09-28-06, 02:05 PM

Quick question- does work in the service menu void your warranty? If so, which part? Parts? Labor? Tube?If you completely reinitialize all the settings with one of the dangerous three-key combos, a service tech might suspect that you did this, as the evidence is pretty clear, and charge you for restoring the set. But your centering, color, brightness, etc. adjustments in service mode are no different from the factory's adjustments -- that is, they don't leave behind some special trace, like breaking a seal on a box. Every set sold has many settings different from the defaults, since there are manufacturing variations that require final tweaking at the factory (mostly the yellow boxes in the service-data charts).

Further, making sensible adjustments can't possibly void warranty, as Sony would have to prove any subsequent failure was the result of your actions. When a Tech came to my house under the service warranty to adjust vertical color convergence for me (magnets on neck of CRT), he didn't blink when I discussed common service-mode settings with him.

An example of an error on your part that would get you charged a service charge if it persisted might be this: You tinker with one of the essential sweep-timing settings, and the picture is hopelessly scrambled or disappears entirely. If you were to WRITE the settings, you'd really be up the creek! Lesson: If you make a Big Mistake, STOP, and just power down immediately, perhape even unplug and replug in the set. You are mostly protected from big goofs, except for those nasty master-reset three-key combos: Unlike your favorite computer programs, there is no last-chance message, "Reset everything? Are you kidding? [OK, Cancel]

TheGodfather

09-28-06, 03:01 PM

Hehe, thanks. I like the part about discussing the service menu with the tech, smooth!

WizJr

09-29-06, 12:46 PM

Hello,

I know that this has probably been brought up before but I could not find a difinitive answer here. I just recently got my new KD-34XBR970 and have noticed that when watching an image in widescreen (with the black bars at the top and bottom) and also just by looking at the information displayed at the top of the screen the the image tends to bow slightly from one side to another. What I mean to say is that it starts low gets higher in the middle and ends at the same low spot it started from on the other side. The only way I can illustrate this is to do this ___------------___ . It is a very gradual cure and nowhere near as exagerated as the illustration but it is consistant throughout the picture that is to say it bow from side to side completely from the top of the screen to the bottom. It's not a huge issue as I watch most things in HD and with the TV set to FULL screen size at 1080i you can't even tell unless the display info pops up. Anyways, to complete my rambling question, is there a setting in the service code menu that will help with this and if so, on what menu is it? I spent last evening trying to find the correct setting but there are enough to boggle the mind. Any help would be greatly appreciated.

Thanks,

Elliott

Shawn Watson

09-29-06, 05:37 PM

I really need to know how to fix the overscan on a Sony Bravia 50KDFE50A12U. Any help at all, please?

Shawn

rarerequest

09-29-06, 09:18 PM

Are there any internal test patterns for sony wega KD-27FS100 ?
Does anyone know how to adjust the maximum white level?
My tv have weak colors and pretty poor color tone .

KenTech

09-29-06, 10:33 PM

What I mean to say is that it starts low gets higher in the middle and ends at the same low spot it started from on the other side.Do you mean that a horizontal line going thru the center of the screen isn't straight but curved, with the left and right ends turned down a bit and the center high? Same for lines above and below center?

WizJr

09-30-06, 08:10 AM

Do you mean that a horizontal line going thru the center of the screen isn't straight but curved, with the left and right ends turned down a bit and the center high? Same for lines above and below center?

Yes! That is exactly what I mean. It seems as if the line is perfectly straight for most of the screen width but about 2-3 inches from each end it curves down just a little. After I posted this yesterday I saw something that said this was a flaw in all 34X***** model televisions and had something to do with the magnets. IS that the case and if so should I bother to call sony repair to fix it or not bother because they will just become unaligned again?

Thanks,

Elliott

GlenC

09-30-06, 04:31 PM

Yes! That is exactly what I mean. It seems as if the line is perfectly straight for most of the screen width but about 2-3 inches from each end it curves down just a little. After I posted this yesterday I saw something that said this was a flaw in all 34X***** model televisions and had something to do with the magnets. IS that the case and if so should I bother to call sony repair to fix it or not bother because they will just become unaligned again?

Thanks,

ElliottWelcome to the world of CRT. Geometry is one of the main areas you need to compromise. If you want absolute perfect geometry, you need a digital pixel display, but there are other compromises there too. The bending lines are on all CRT displays to some degree, more noticeable on the flat widescreen models. It is just the nature of trying to bend an image from a point source with magnets and make the TV affordable by eliminating a lot of geometry circuits to correct the problem. Many $30K+ CRT projectors have this capability, but not in a consumer display.

KenTech

09-30-06, 07:39 PM

After I posted this yesterday I saw something that said this was a flaw in all 34X***** model televisions and had something to do with the magnets. IS that the case and if so should I bother to call sony repair to fix it or not bother because they will just become unaligned again?No, it won't become "unaligned" again. The flaw is in the initial setup. Cure it, and it's cured forever.

Curvature of horizontal lines can be adjusted only by magnets or by deflection-yoke alignment, and we don't really know the cause in your case. I don't believe for a second that it's a flaw in all 34X*** TVs, because the cause is not something inherent in the design. Sony might have made a bad run of deflection-yoke assemblies that has appeared in a certain production run of TVs, but that would have a beginning and ending point. Same thing if a geometry-setup technician was having a bad week and did a second-rate job on 100 TVs -- who knows? My 36XS955, which has virtually the same technical deflection task as yours, is perfect, and I have seen 34XBR960s that were perfect at screen center. A little gamey in the corners, perhaps, but not in the "meat" of the picture. (The "XBR" designation would have no bearing on this.)

As Glen says, no CRT set is lkely to be perfect in geometry, but we who love them for their virtues will just have to forgive them a few quirks. Flat-panel sets have their own quirks and limitations, albeit perfect geometry. This month, this year, I would still buy another CRT-TV (if available!). Your set, it seems to me, goes a bit beyond "quirk."

If you are really not splitting hairs, and this really is annoying from your normal viewing position, I would complain under the service warranty to see if a tech can come out and improve on it. Call directly a local shop recommended by a store that sells the set; you're likely to get more cooperation than calling Sony, who may stonewall.

Reagan

10-02-06, 12:12 PM

I haven't checked this thread in about a year (it's a great thread). I browsed the table of contents and the .pdf files of the service menu and can't find what I'm looking for. Hopefully, someone here knows how to do what I want and wouldn't mind sharing the solution with me.

I have a two year old 420. In short, I want to remove the internal tuner (NTSC only) from the input list. I know that I can use the channel fix to set it to Video 1, but I don't want that either. Now that I've removed my VCR from the pipeline, I only use the two component inputs and the HDMI input. Just those three. So when I hit input, I don't want to stop at the tuner or video 1. I can't figure out any combination of settings on the user menu (if I set Video 1 to skip, it goes to the tuner) or identify any part of the service menu that allows for this. Does anyone know the answer?

Thanks,
Reagan

kidblast

10-09-06, 02:23 PM

For starters, thanks! This thread helped me dial in my TV and get rid of much of my overscan. It was at almost 7% on all sides and is now down to 2% on all sides. You guys rock!

I do have a question though. The top left and right sides of my screen (in HD mode, 1080i) are a little messed up and there has been nothing I've seen that has fixed it. I'd post a picture but I don't have 5 posts yet. I could email the picture. It's pretty much the top inch of the screen. The left and right 25% of the screen is warped and is skewing in towrds the center. Normal geometry corrections have done nothing.

Do you have any suggestions for me? Any help would be much appreciated. Thanks for all you help, keep it up!

Justin

Twiin

10-10-06, 01:38 AM

I'm having a bit of a problem with my 30XS955, and I'm not sure if it's something that's resolvable via service codes or not, hopefully someone knows if this can be changed.

Recently (and it wasn't always this way), there's been a weird horizontal distortion on my screen. It's most noticeable on vertical lines in letters, but it affects the entire image. A line which would normally look like this:

|
|
|
|
|

...now looks like this:

\
/
\
/
\

The zig-zags are probably a few millimeters tall at most, and the distortion is constantly moving, if that makes any sense. The edges of any crisp letters are constantly wavering with this distortion.

I did some minor adjustments via the service menu when I installed this set, but otherwise I haven't touched anything. AFAIK, service codes are on a per-input basis, and this problem occurs anytime there's an image on the screen, on any input, so I'm a bit concerned it's a problem with the set, and not something I can fix via the service codes.

Any thoughts? Is there a setting to adjust anything like this? I imported my TV, so I effectively have no warranty service.

KenTech

10-10-06, 02:31 PM

I have a two year old 420. In short, I want to remove the internal tuner (NTSC only) from the input list. I know that I can use the channel fix to set it to Video 1, but I don't want that either.It appears to me that the programming of the user menus does not treat the internal tuner(s) as an "input." I understand: That would be confusing to most consumers. I think of it as "V8."

Nothing I see in the service-mode codes or menu options appears to eliminate the tuner from the "input" list. Indeed, the tuner is given special status, as an up- or down-channel button press activates the tuner no matter what other input you are using, a convenience I enjoy.

I don't think you can do it.

KenTech

10-10-06, 02:38 PM

Do you have any suggestions for me? Any help would be much appreciated.Dude, think! You haven't told us what model TV you are talking about! How can anyone help?

(Yes, I know this sounds like scolding. But there are so many folks here who would be willing to help -- if inquiring people would just do the obvious to make it easy, so we don't have to beg for more information. Please be specific; no one can read minds, here -- that I know of.)

kidblast

10-10-06, 03:06 PM

Sorry.

Sony KV-36XBR450 is the TV I have. I'm not sure that I've seen that model come up in this thread anywhere but most of the service menu adjustments have worked for me.

I kinda think the problem I'm having was never an issue before because my TV had so much overscan that the problem areas where being covered up.

Thanks.

KenTech

10-10-06, 10:35 PM

Sony KV-36XBR450 is the TV I have. I'm not sure that I've seen that model come up in this thread anywhere but most of the service menu adjustments have worked for me.The important thing is that you've identified a 4:3 screen, and you're concerned about the visible top and bottom edges of the HD/16:9 reduced-height scan.

I have a 36XS955, and I have the same problem. A rectangle is nearly perfect if it occupies nearly the entire screen -- say 1-2" in from the edges. (I've adjusted it this way.) So a straight line across the screen near the top is displayed quite straight. But if I were to lower this line toward the center, it would go through a "warp" area, where its left and right ends would droop downwards a bit, and then it would be straight again as it reaches the center of the screen. Same thing happens in reverse, upside-down, as I move this hypothetical line further down the screen until it's near the bottom. You can plainly see this effect, of course, with any of the common crosshatch patterns.

I believe it's intractable. I've seen it on every 4:3 big-screen Sony I've ever watched in stores, and on one like mine at a friend's house. The geometry corrections to get a rectangular display on a flat tube are really heroic, but not perfect. I think it would require increased complexity in the deflection system to correct this "defect," and Sony have simply decided not to bother, even though it's obvious whenever viewing 16:9 anamorphic DVDs or HD programming. I find now that I hardly see it unless I look for it -- a good attitude if one can't do anything about it!

Now -- do I have this all wrong, and the above-described "droop" isn't the distortion you're duscussing? (Yes, a picture is always helpful.)

kidblast

10-11-06, 08:50 AM

I tried to PM you a screen mock-up but your message box was fool. No problem though...I just can't post any links here till I have 5 posts. I didn't want to make a few garbage posts just so I have 5 though.

The problem I have can't even be seen when I'm watching HD programing but when I'm playing my Xbox360, it sometimes really stands out just because of the nature of the menus, etc. that are being displayed.

I think what you described is pretty close to what's happening. Once I get to 5 posts I'll post a screen mock-up though that shows the problem.

Thanks!

fallenlordz

10-16-06, 09:29 AM

Does anybody know what to do? I have a KD-34XBR970 an I am trying to fix the Geometry and Overscan problem of this TV.

When I fix the Geometry and Overscan via the Service menu for 1080i using DVE, it seems the data correlates back to 480p and the 720p Geometry and overscan.

What i mean is when i change the data of VPOS,VSIZ,HCNT,HPOS,HSIZ, etc. On the 1080i I get a great picture. But once i change the data on this 1080i, the data for the 720p would also be changed without my doing. Same goes for the 480p.

Is there a way to save seperate data for 480p, 720p, and 1080i? Without having one be changed by the other.

Mr.Bitey

10-16-06, 09:52 PM

Hi All,

Could someone please help me out with the service-menu code for a KV-32LS65AUS ? - I have tried a myriad of codes that ive found thus far on the net, but cannot find the one for this TV :(

EDIT: I'm wanting to make some overscan adjustments and to see if I can disable some 'repeating' AV inputs (e.g. I have to scroll through one AV1 to get to the 'other' AV1 thats connected to my digital tv box)..

Regards,
Bitey

nursing

10-17-06, 11:17 AM

Hi,

Can someone please help with instructions on getting into the service menu for the KV-27HFR. I have tried pressing DISPLAY --> Channel 5 --> Volume + --> POWER with the power off, but all that keeps happening is the TV turns on. I have also tried pressing DISPLAY --> Channel 5 --> POWER...nothing.

Thanks,
Nicholas

NextGen

10-25-06, 03:41 PM

DSperber

10-25-06, 04:14 PM

I'm having a lot of trouble getting FULL mode and NORMAL mode to work correctly. I have my raster centered and the geometry looking good for my 34xs955-N in 16x9. However, I can't see any difference horizontaly when I switch to NORMAL mode. In other words the 4:3 stretches the full screen instead of giving me a 4:3 image with black bars.

2170D-5 2-1 EWDC set to 1 is perfect for 4:3 and when set to 0 is perfect for 16:9 but EWDC doesn't seem to be independent of FULL and NORMAL. So I have to go into the service mode anytime I watch 4:3 material.

Any help?What INPUT are you doing this with... S-video (INPUT1/3), component (INPUT5/6), HDMI (INPUT7), 1394?

What channel are you trying to watch in 4:3... analog or digital? What is your source... OTA, cable, SAT? What is your STB/DVR? If you're watching from a STB/DVR, is it set to deliver 4:3 (with side bars) or to stretch to 16:9 when it delivers SD to the HDTV? Your problem may be the STB/DVR setting, not the Sony.

Ironkaw

10-26-06, 12:53 PM

I have a Sony KD-34XBR970. I've figured out how to get into service mode, and fixed a couple geometry issues, but I have one problem that I'm not sure how to fix, or if I can. The problem affects the left side of the display. For example, when I hit the display button on the remote, the info boxes are displayed in a transparent mode, with gray lines making up the boxes. The box in the top lefthand side of the screen which shows the imput source (Video 1 - 6) and the label associated with it shows distortion. The gray horizontal line that makes up the bottom of the box starts out fine from the right side, but as you go to the left it becomes increasingly distored. By distortion, I mean the colors separate, and it becomes three distince lines bowing downward. One blue line, the a green line bowing lightly more, and finally are red line below that bowing even more than the other two. Is this a major problem, or can it be adjusted from the service mode. I will have a tech. come and work on it if need be, but I would much prefer to figure out with some help how to fix it myself if possible. Any help will be greatlly appreciated. I would also like to say thanks to all who post here, because I have learned alot. This is a great forum!

Thanks

Chris Mason

187

10-27-06, 08:29 PM

Hi guys I just bought a new 30HS420 and I wanted to know if someone have some good settings for the service menu cause I dont want to scroll all these 70 pages... I just read the post from KenTech about the color setting and he was talking about clouds and all these things and I was like this is gonna take me 1 month to calibrate lolll. These forums have been a great source of information since when I registered 3 years ago and I hope someone will help me with this thing;).

Tv is set to Pro mode, color temp natural, clear edge off.

I think my convergence is fairly good and the thing that i want to achieve with this tv is to get the very best shadow details without sacrificing my black level. Its very important for me as a gamer with the 2 consoles i want to buy: X360 and ps3.

Thanks!

ntwp13

10-29-06, 04:40 PM

Hello and thank you all for this wonderful thread,

I have started tweaking my KV-34HS510, and have a few geometry questions. I used the table of contents and have tried to read all of the info concerning geometry and under/overscan issues. Please forgive me if I missed something.

In the MID1-3 groups, be sure to get the values for width and especially height exactly right, as these values control (as I have discovered) the precise sampling of the video for display. Get the vertical ones wrong, and the image quality will be compromised! There are exact ratios between the particular video-line count and the internal digital representation of that video that have to be maintained.

The critical values are DHAR, DVAR, MDHS, and MDVS in MID1; in MID2, DRHS and DRVS (if relevant); and in MID3, VDHS and VDVS. The correct numbers are the ones in the data tables.

I would copy all other numbers exactly as listed in the MID1-3 tables, especially MID3 #4-11. The ones associated with "phase" or "pos" (position) can be tweaked a little to balance the centering of a video frame between inputs; but don't mess with these without first getting the positioning right in the 2170D groups.

1. After adjusting the raster, I am still unable to eliminate overscan since the "picture" (MID1-3 settings) seems to be bigger than the raster. Does the above quoted advice mean that I should not resize the "picture", and only use the MID1-3 settings to center "picture" on the raster?

2. I'm having trouble eliminating some severe Horiz bowing at the bottom of the screen. I have eliminated the hump that was in the middle of the bottom lines using VCEN and VPIN. But I cannot seem to get rid of some inward horizontal pinching in the corners. The lower left corner seems to be the worst and is very noticeable in sports/news tickers, or the MCE info box.

Here is what I mean by horizontal pinching. Of course the problem is not nearly this bad:

///-----------------\\\
-----------------------
-----------------------
-----------------------
\\\-----------------///

Thanks for any help.

raouliii

10-29-06, 05:44 PM

Hello and thank you all for this wonderful thread,

I have started tweaking my KV-34HS510, and have a few geometry questions. ...............
2. I'm having trouble eliminating some severe Horiz bowing at the bottom of the screen. I have eliminated the hump that was in the middle of the bottom lines using VCEN and VPIN. But I cannot seem to get rid of some inward horizontal pinching in the corners. The lower left corner seems to be the worst and is very noticeable in sports/news tickers, or the MCE info box.

Here is what I mean by horizontal pinching. Of course the problem is not nearly this bad:

///-----------------\\\
-----------------------
-----------------------
-----------------------
\\\-----------------///

Thanks for any help.You might try these service menu items in the 2170D_2 group. A crosshatch pattern is VERY useful for making geometry adjustments. Record your current settings. The items have entries for WideZoom mode and Others.
MPIN, horizontal outer and inner pincushion, is effective on inner and outer areas of the screen.
PIN, horizontal outer pincushion, is more effective in the outer areas of the screen.
UCP, upper corner pincushion, straightens the vertical lines in the upper corners.
LCP, lower corner pincushion, straightens the vertical lines in the lower corners.
Good Luck

ntwp13

10-29-06, 06:29 PM

You might try these service menu items in the 2170D_2 group. A crosshatch pattern is VERY useful for making geometry adjustments. Record your current settings. The items have entries for WideZoom mode and Others.
MPIN, horizontal outer and inner pincushion, is effective on inner and outer areas of the screen.
PIN, horizontal outer pincushion, is more effective in the outer areas of the screen.
UCP, upper corner pincushion, straightens the vertical lines in the upper corners.
LCP, lower corner pincushion, straightens the vertical lines in the lower corners.
Good Luck

Thanks for the quick reply.

I've been using the various test patters posted here including the overscan.bmp and a basic crosshatch.

I thought the MPIN, PIN, UCP, and LCP work on the vertical lines like this:

\\||||||||//
||||||||||||
//||||||||\\

I suppose that all these settings effect everything else though. I'll give these settings a few tweaks when I get a chance. Thanks.

RWetmore

10-30-06, 12:12 AM

TotallyTubular

10-31-06, 10:13 PM

I purchased an XBR970 a few months ago, and I've been happy with its peformance, but I decided I wanted to tweak on the overscan a bit since I've recently hooked my PC up to it. While browsing through the service menu, and playing around with the various settings, I tripped across "APSW" in 2170D-1. To my surprise, it managed to open up some zones I was previously unable to open even while playing with the shutters. Unfortunately though, there is a problem when using this change. The bottom of the screen is severely distorted, and I've no clue of a way to mitigate it's severity.

http://img66.imageshack.us/img66/6586/sspx0044hl7ys7.th.jpg (http://img66.imageshack.us/my.php?image=sspx0044hl7ys7.jpg)

I would appreciate any insight to this problem, whether directly related to the distortion, or else. Thank you.

TotallyTubular

11-01-06, 07:09 PM

Update: I've managed to find a solution to my own problem above. Most of it stemmed from my using of the 960's service chart. Basically the 970's service menu is different when it comes to MID3, and took some exploring to make appropriate changes to my canvas.

The distortion on the bottom of the screen was fixed by "PSPT" in Mid3, and my overscan was fixed by using various settings throughout mid1-3, most of which were not on the 960 chart. This may be helpful to some as it was an odd problem to me.

MOtvGuy

11-02-06, 08:04 AM

Oh great. :rolleyes:

I'm going to suggest to the tech reps at all the service meetings I attend that they eliminate the ability for anyone to access service menus.

TotallyTubular

11-02-06, 08:32 AM

Oh great. :rolleyes:

I'm going to suggest to the tech reps at all the service meetings I attend that they eliminate the ability for anyone to access service menus.

I take it that's a response to my posts. Why?

raouliii

11-02-06, 09:45 AM

Oh great. :rolleyes:

I'm going to suggest to the tech reps at all the service meetings I attend that they eliminate the ability for anyone to access service menus.How do you propose to do that and why?

justsc

11-02-06, 12:40 PM

Oh great. :rolleyes:

I'm going to suggest to the tech reps at all the service meetings I attend that they eliminate the ability for anyone to access service menus.
That's no solution.

This ability has been there from the start. It's a fact of life in the tv industry as you already know.

Besides, the more folks that get into the SM and screw up their sets the more business for you, right?

MOtvGuy

11-02-06, 08:58 PM

I take it that's a response to my posts. Why?

71 pages of people getting into their service menus and tweaking it.

Some of these are the same people whose VCR's flashed 12:00 for years.

Kinda like telling folks back in the day to get into their autoexec.bat files and start typing even though most of them didn't know what they were doing.

As a servicer the last thing I'd want to do is go through line after line of data settings and try and figure out what the customer has done to screw up his set. Very, very time consuming and un-needed.

TotallyTubular

11-03-06, 08:43 AM

I could see how that would be a problem if someone just went through changing every value and saving it that way. I just nibbed at a value here and there to see what it did. I documented it, and managed to fix near every problem I had. If there was any other way to research those settings I would have; turtling through those settings wasn't fun for me either. Do you know of a place, or a way? Oh, and don't worry, my set won't ever see a professional. ;) I'm a college student.

I see your point though, but I'd say that's a little selfish. My TV wouldv'e never gotten fixed. :(

I'd like to reiterate. Do you know how I could research what settings do before I go changing things on a TV like the 970? A lot of the S.M. is similar to the 960, but there are obviously some differences.

bidger

11-03-06, 07:21 PM

I searched this thread, but couldn't find what I was looking for. I have 2 SONY sets in the same room and am trying to find out if there's any way to change the remote address on one so I don't have to put up with both sets responding to remote commands. TIA.

jpl3447

11-03-06, 07:35 PM

I have two TVs in the same room too. I put a piece of electrical tape over the IR spot on the TV that was least used. Then I manually turn it on and off. That is the only thing I could think of.

I wanted to say thanks to KenTech, DSperber and RWetmore for all their tips. Using their advice and explanations I corrected overscan, a minor geometry thing, and improved on the factory settings. I am a student so thanks to the school's computer lab I was able to print out two service data charts (one for each TV) and pencil in any changes I made. I put it all into a binder along with sections containing print outs of articles listed in this thread. I feel very comfortable going into the service menu now. A few hours of research and tweaking yielded results that really made the image on the screen pop. High definiton broadcasts are just amazing and SD is much better. Thanks again!

NextGen

11-05-06, 11:49 PM

What INPUT are you doing this with... S-video (INPUT1/3), component (INPUT5/6), HDMI (INPUT7), 1394?

What channel are you trying to watch in 4:3... analog or digital? What is your source... OTA, cable, SAT? What is your STB/DVR? If you're watching from a STB/DVR, is it set to deliver 4:3 (with side bars) or to stretch to 16:9 when it delivers SD to the HDTV? Your problem may be the STB/DVR setting, not the Sony.

Input 5, and it does this regardless of the material or source. All 4:3 signals in 480i are slightly wider than they should be, about 2-3 inches on each side need to be gotten rid of. 16:9 sources appear correct in Full mode. As does any other source in any other resolution. Even 480p 4:3 material is perfect, I just can't understand the deal with 480i stuff.

p3Orion

11-06-06, 11:32 AM

I recently purchased a Sony KD-27FS170 and would like to change some settings, specifically color calibration and black level. In section 3, COLOR CALIBRATION AND BLACK LEVEL, you state to leave the red settings alone. As such your RDRV is 42 and RCUT is 31. My settings are much higher. RDRV is 84 and RCUT is 100. Are those too high? Should I leave them unchanged and just adjust GDRV (default75), BDRV (default 73), GCUT (default 79), and BCUT (default 76)? Or would using lower settings for all, like you suggest, result in a better image on my TV?

With regard to the other color settings, GDOF through BCOF, you state that the offsets are zero (no effect) when the codes are set to 31. What is the maximum values on your TV? Mine max out at 127.

Thanks. :)

Compass

11-06-06, 07:56 PM

I have two TVs in the same room too. I put a piece of electrical tape over the IR spot on the TV that was least used. Then I manually turn it on and off. That is the only thing I could think of.

Ha. I also have two sets in the same room, and that remote problem has plagued me as well. It's not such a big deal when I'm only watching one (the remote is on top of the TV so I have to get up to turn it on anyway -- at that point I just point it real close to the sensor to avoid turning on the other set). The main problem comes when I have both TVs on at once and want to turn down the volume on just one. This is very tricky to do without getting up.

Mr.Bitey

11-13-06, 09:20 PM

Ive found the service code for the KV-32LS65AUS - seems its radically different to the other codes in other models! - no wonder I was having a bugger of a time getting into the service menu.

Its a 2 step process. 1 - set the remote control into service mode, then set the tv into service mode (nice heh).. Seems MOtvGuy like thinking is trying to make it more difficult. Now wonder the poor tech I dragged out (under warranty) couldnt get into the service mode!

Step 1: Remote Control to Service Mode
1.1 Press the VCR/TV/DVD button until the TV LED lights
1.2 Press and hold the YELLOW button for approx 5 seconds until the TV Light flashes quickly
1.3 Press 99999 all 3 LEDs should light (on the remote)
[Remote Control now set to service mode]
1.4 To Put the Remote back into NORMAL Mode, repeat 1.1 - 1.2 then enter 00000 (instead of 99999 to enter service mode) - all 3 LED's should light
[Remote now in Normal Mode]

Step 2: Put TV into Service Mode
2.1 Set Remote to Service Mode (as above 1.1 - 1.3)
2.2 Turn TV on with Power Switch
2.3 Press the VIDEO STANDBY button on the REMOTE TWICE (its the one on the top LEFT), TT____ should now be displayed on the TV screen (and some other stuff)
[TV Now set to Service-Mode]

2.4 Press MENU on the remote to display the service-menu
2.5 use the UP/DOWN arrows to navigate up and down in the menu
2.6 Press the RIGHT arrow to enter into the required menu item
2.7 Press the MENU button on the remote to go back
2.8 Note: Turn TV OFF to really QUIT the service mode

Hope this helps!

EDITED: The service manual wasnt quite right - but this is tested on my set and works 100% :)

Cheers,
Bitey

prefontainenike

11-16-06, 03:15 AM

hey guys, ive been searching the forum/this thread for an hour, i cant seem to gt out of the test pattern that loaded while i was tweaking in service mode. any ideas???

thanks for all the help,

Chris

louisepalmer

11-16-06, 03:21 PM

hello guys,

Sorry for my english, i'm french and i never found a forum on this subject.
So i found some very interesting thing here but not on my problem.

I just buy two sony CRT TV, One is KV28FX60 and i discover that i HATE 100 HERTZ !
My old sony 50 hertz has a very beautiful image...

My question is if it's possible to disable 100 hertz and all other digital feature to increase image quality. In a word, i want to switch to standard 50 hertz.

I know how to enter service mode, but it's very difficult to understand all the options.

I hope you have understand me :)

Your help will be very appreciated.

Ironkaw

11-16-06, 08:29 PM

RWetmore,

Thanks for the info. I've read, and learned alot more, and decided to swap my set out for another one. The new one has some mild geometry issues, and I have corrected most of them, but there is one issue that is bugging me. Everything in 16:9 is very acceptable, but when I watch something on regular TV (4:3 480I) it is not wide enough. It's narrow by almost 2 inches. Then if I watch a DVD thats 4:3, it is too wide by almost 2 inches. I can live with it too wide, but the regular TV (4:3 480I) is just annoying. If I change the Horizontal width, it changes in every mode, and to get the (4:3 480I) at the correct width it makes everything else way to wide. Is there a way to adjust the 4:3 width on a 970 independent of the other modes? I saw on the service data chart an option in MID1-2, but these aren't the same on the 970, and the MID1-3 options are different also. Any help would be appreciated.

Thanks ,
Chris

KenTech

11-20-06, 01:33 PM

hey guys, ive been searching the forum/this thread for an hour, i cant seem to gt out of the test pattern that loaded while i was tweaking in service mode. any ideas???Sure. Try the instructions here (http://www.avsforum.com/avs-vb/showthread.php?p=8430148&&#post8430148).

doveman

11-24-06, 09:11 PM

Just in passing, I did my 1080i H/V position/size and overscan tweaks using a 1920x1080i test pattern projected by DisplayMate for Windows Video Edition on my PC, connected to my XBR960 from the DVI-to-component output of my PC's ATI Radeon 9800 Pro video card. The ATI Catalyst video drivers had the XBR960 set as a second monitor running at 1080i, 1920x1080 resolution.

I'd be grateful if you could explain how you did this. I'm using the same program and I've got my Nvidia graphics card setup in Dualview mode, with the monitor at 1024x768 and my HDTV (XBR800) at 1080i. Thing is, when I run the program it runs on the monitor and therefore uses it's resolution, so how do you get it to generate a 1920x1080i test pattern?

DSperber

11-24-06, 11:46 PM

I'd be grateful if you could explain how you did this. I'm using the same program and I've got my Nvidia graphics card setup in Dualview mode, with the monitor at 1024x768 and my HDTV (XBR800) at 1080i. Thing is, when I run the program it runs on the monitor and therefore uses it's resolution, so how do you get it to generate a 1920x1080i test pattern?Two different ways will both work.

(1) Reverse monitor #1 and monitor #2, thus making the HDTV your primary monitor. Then the patterns presented by DMWVE will take on the size of your HDTV. When you boot your PC, the HDTV will be your monitor #1 and all apps (including DMWVE) that present full-screen displays will present at 1920x1080.

Of course this approach genuinely makes your HDTV the primary Windows monitor, so all program windows will launch here, etc. Yes, the desktop can be specified to span the combined real estate of both monitors, but the HDTV will definitely be where most things start and stay (when enlarged to full-screen). Takes a bit of getting used to when your keyboard is on the other side of the room and you have to look over at your HDTV which is 15 feet away.

(2) I had previously full-screen captured all of the series of test patterns produced by DMWVE (from every one of its diagnostic/setup groups), using a screen capture utility named Fullshot and having my standard 4:3 CRT monitor set to a desktop resolution of 1920x1080 so that the captured images had the same resolution and dimensions. I captured them to BMP (so as to not lose anything).

If you want (via Settings), Fullshot is smart enough to determine the color depth to be used so as to minimize the size of the BMP, so that a B/W test pattern does not require color and at 1920x1080 that would be 1036918 bytes per capture. "Primary color" captures that can be done at 16-bit color depth require 2074678 bytes per capture. "True color" captures (e.g. spectrums) require 32-bit color and take 6220854 bytes. Anyway, I have a complete set of 1920x1080 BMPs stored that exactly duplicates all of the test patterns produced by DMWVE.

These full-screen BMPs can be displayed full-screen (on either monitor #1 or Monitor #2) using any image viewer/browser worth its salt (e.g. ACDSee).

The method I actually used for my 34XBR960 adjustments was (2), because I wanted to be able to sit in from of my usual PC monitor #1 (19" IBM P96 4:3 CRT monitor, running at 1152x864) while sending the 1920x1080 BMP test pattern images (from my sets of screen captures) to monitor #2 (the XBR960) via ACDSee. If you will do this, the resulting flexibility and convenience is significant.

Without having the set of 1920x1080 BMP capture equivalents of the DMWVE test patterns, method (2) wouldn't have been available. Only method (1) would have been possible. I had actually tried that first but didn't like the hassle of having to look over at my HDTV to see the primary desktop and location for windows that got opened when programs launched, while sitting in my computer chair.

NOTE: ACDSee as a product took a steep turn south beginning with version 6, some years back. Each subsequent version (they're now up to 9) gets progressively worse and crappier. I continue to use version 5 for most of my needs, and actually use version 4 (fastest and leanest, while still fully-featured) for my most demanding volume image browsing, viewing, and processing applications (backed up by Photoshop, of course, for the real work).

doveman

11-25-06, 04:59 PM

Thanks for the detailed explanation. (1) is pretty obvious but I doubt I would have thought of it. doh!

My TV's right next to my PC but it's still awkward trying to view it whilst sitting at the PC. It would actually be easier if it was further away but at right angles to my PC chair. I can see why you'd want to go with (2). My brother occasionally screws up his Dualview and ends up with his desktop on his TV which is in the opposite far corner and it's a nightmare helping him to fix that over the phone!

I'm going to see if my TV looks like it needs much tweaking using (1) for now. Fingers crossed!

DSperber

11-25-06, 09:18 PM

I'm going to see if my TV looks like it needs much tweaking using (1) for now. Fingers crossed!If, as expected, your set has non-zero and possibly excessive overscan, be prepared for your use of the HDTV as monitor #1 causing some operational problems initially (i.e. until you can adjust your overscan).

Depending on your desktop layout, if you've always only had one monitor then your taskbar will have been at the bottom of monitor #1 and at least some of your desktop shortcut icons will probably be in a vertical line down the left edge of monitor #1. Once you make monitor #1 your HDTV with its overscan, a good portion of the taskbar will appear to be "cropped out". And that left vertical edge of desktop shortcuts may similarly be in at least a somewhat awkward location very near the left edge of your screen.

Furthermore, if your set really requires convergence adjustments and has geometry issues as well, the pixel-precision you're probably used to on your PC's monitor and that is not going to look pretty on your HDTV will no doubt surprise you. Just don't forget it's all about getting the DMWVE patterns to your HDTV so that you can do your own service menu adjustments as possible.

But then that's what this project is really all about... to get test patterns onto your HDTV at 1920x1080. You will find, when you're "done", that the test patterns looks better than the Windows desktop, because any non-zero overscan directly affects the desktop which extends to the very very outer edges of the screen (assuming 0% overscan), whereas the DMWVE patterns are "TV-tolerant" and assume that you will end up with maybe 1-2% overscan which is perfectly acceptable and appropriate for watching broadcast HDTV.

I'm attaching the two primary test patterns (in BMP form), if you want to try them first (or alternatively). One is the cross-hatched overscan/geometry/linearity test pattern, and the other is the SMPTE colorbar pattern with Pluge (to adjust brightness).

doveman

11-27-06, 07:54 PM

Thanks for the images. I was suprised how awkward it is trying to use my TV as the primary display. My taskbar actually stayed on my monitor for some reason, which made it slightly easier to run programs but the flickering on the TV (due to the display being interlaced I guess) makes it hard to look at for any length of time.

I heeded your advice and avoided ACDsee and found a free program called FastStone Image Viewer which allows me to display the images on the secondary display. With the cross-hatch pattern, are the dashed lines around the edges meant to be at the edge of the screen, because they're about 1-1.5" in at the moment. There's actually a non-dashed line that is visible on the right hand side and if I adjust the MID settings I can get the same at the top and bottom of the screen, but not on the left. Apart from that, it looks about right.

I've noticed that when I have the PC in Clone mode, the desktop almost fits on the TV. If I open Windows Explorer full-screen, in the Status bar at the bottom it shows "8 objects" and the 8 is chopped off. There's a slight curve inwards in the bottom left and top right corners as well.

I have to admit I was suprised how unclear text is on the TV. It's considerably better in 720p mode but still compares rather badly to my monitor.

I'm mainly going to be using the TV to watch videos from my PC (connected by DVI) or with my Xbox (composite at the moment but I'll be getting a component lead). I do intend to get a Freeview box with a component ouput but I'm waiting to see if a HD one comes out, so at the moment I'm stuck with using my VCR's tuner into a PAL->NTSC converter, which does look rather crappy but I can live with it.

jsimon9633

11-27-06, 11:38 PM

I have a sony kv-32hs500

I was trying to do the convergence and now my screen shows 3 color lines across the top middle, red green and blue

i tried resetting all things in service mode but its still there

also the screen seems like wa too bright like a spotlight shining on it in the midle and corners

can anyone help?

versus21

11-28-06, 12:19 AM

Just hooked up my cpu to my screen and my god I never knew my screen was so jacked up! Text gets blurry when not in the middle of the screen and there is a bulge on one side that makes all the windows distorted.

Went into the service menu but didn't know what the heck i was doing!

So a service manual for the Sony KV-32FS100 would be GREATLY appreciated!!
Or at the least the default service menu settings.

Looked all over the net looking for it but couldn't find if you would have a copy of it pm me please.

DSperber

11-28-06, 01:41 AM

With the cross-hatch pattern, are the dashed lines around the edges meant to be at the edge of the screen, because they're about 1-1.5" in at the moment. There's actually a non-dashed line that is visible on the right hand side and if I adjust the MID settings I can get the same at the top and bottom of the screen, but not on the left. Apart from that, it looks about right.The solid line surrounding the rectangle is the outer edge of a 16:9 pattern at 1920x1080. If you adjust things (first background raster, then foreground image, as has been described in detail in this and other threads) so that the solid outer edge of the rectangle is visible at all four edges of the screen, you now have 0% overscan. Anything where this outer edge is not visible is some non-zero positive overscan. And You certainly don't want "negative overscan", where you've got it so that the outer solid lines of the rectangle are "indented" from the edges of the screen. You want that outer rectangle at the extreme outer edge of the screen (at least to start) for 0% overscan.

As you can tell by counting the horizontal and vertical boxes in the pattern, the interior dotted rectangle defines 10% overscan (since there are 2 out of 20 boxes eliminated in the horizontal dimension by this inner dotted rectangle, and about 1.4 out of the 14 boxes eliminated in the vertical dimension). If you adjust things so that the dotted lines are visible at the four outer edges of the screen, you have just adjusted for 10% overscan.

Anything in between, not able to see the solid outer rectangle but with the dotted lines somewhat "indented" from each edge of the screen, is something less than 10% overscan. Halfway would be 5%, etc. Shooting for 1-2% is reasonable, in my opinion (at least that's where I've gone).

What's important of course is to get all four edges of your screen to display the same thing... some uniform constant overscan (say 1-2%) all around the four edges of the screen. That is your goal. Anything asymmetric is incorrect and your true TV images will be cropped asymmetrically.

The tiny dots in the center of each box are excellent tools to determine if you have a convergence problem (e.g. if they're not pure white, but you can easily see red, green, or blue "spray") and exactly what portion of the screen requires attention (and potentially, a Sony service call for magnet work). Same with the perfectly horizontal and vertical lines of the pattern, which are obviously supposed to be perfectly horizontal and vertical.

If they're not (e.g. if they are non-linear or bow up at the ends near the bottom of the screen, or if they are non-linear or bow down at the ends near the top of the screen, or if they are non-vertical or curved near the left and right edges or in the center) well this is what the pattern is for... to direct your attention and service menu tweaks (or magnet work for the professional) where your screen needs it.

NextGen

11-29-06, 02:32 AM

I've been calibrating my xs955 for a while now, and it is finally starting to look pretty remarkable thanks to all the information provided here. However, my TV is being used for games and not for over the air broadcasts, and because of this I'm having some "internal conflicts" as to what the correct settings are for me. I am beginning to realize that some of the "enhancements" recommended here may not be desirable for people like me. I am arriving at this conclusion based on image quality when playing video games when using some suggested settings. Mainly anything that enhances sharpness tends to introduce more pronounced "jaggies" in many PS2/Xbox games. When using a more natural setting, such as completely disabling any form of velocity modulation, the image is more clear to my eyes. Even high quality DVD's look better as there is far less 'grain' in the image.

Does this conclusion make sense or is it just my lack of understanding on how things work? In other words, should I see some noise or grain in high quality image content like movies or video games? My way of thinking is this, even though there are things "lost" in the digital transfer of film to DVD there isn't any loss when it comes to game consoles because they are rendering everything internally on the fly as you are seeing it. Only cut scenes would be using some form of re-compressed video, correct? If so, what are the codes or settings I should use to disable all image enhancements in my set? Perhaps I can start from 0 and work my way up to suit my tastes.

And one last thing, is there a way to "flash" or update the operating system in a Sony TV? I noticed the XS955's, and probably many other sets, have a software version number. I'd like to know if there are any bugs or issues with specific versions as I would think the latest version would be the best to use.

GeminiEntity

11-30-06, 12:15 AM

I have a Sony XBR970 and recently I noticed some problems have gotten worse. I always had a faint outline of any image on the screen, but lately it has gotten a lot more noticeable. Also, usually on faster moving scenes I'm seeing a X like pattern. It's almost like there is a grid behind my screen that is showing up.

I was wondering if anyone could tell me what these two problems might be and the service menu options that would lead me in the right direction.

disco277

11-30-06, 09:31 AM

Hi All,

First post... I need help with my KV-32HS510. I have a Dish 942 hooked up to it and it works great on 480p and on 1080i. However, when I switch to 720p the color is too green. I think I have the whole service manual, but I can't tell where to change the color settings for 720p or where to change the color settings at all for that matter.

Any help would be very much appreciated!

NextGen

12-01-06, 04:38 AM

Here are some of the service code settings I have decided on for now. My 34xs955 is used mainly for playing video games so I am trying to achieve the most accurate and least artificial image from the display of the game console. Based on findings from the members here I've tried to set up my tv the most transparent and least artificial way possible. The settings below either normalize or disable enhancements and offsets. Some offsets are needed, but I've tried to track down and eliminate any need for sharpness and/or enhancements.

I still get noise when watching movies, but real time content like games are amazingly clean and crisp.

Video
Mode = Standard
Picture = Max ( I may go back to 48 since it is the default value)
Brightness = 31
Color = 31
Hue = 0
Sharpness = 29 (default)
Color Temp = Off
Color Axis = Monitor

3D-COMB
VAPG = 0
VAPI = 0

2103-1
SCON, SCOL, SHUE = 7
YDLY = 0
SHAP = 0
SHFO = 0
PREO = 0
AFCG = 0
CDMD = 0
SSMD = 0
ATPD = 0
DCTR = 0

2170P-1
SBRT = 32
RDRV= 63
GDRV = 33
BDRV = 33
RCUT = 63
GCUT = 16
BCUT = 32
DCOL = 0
WBSW = 0

2170P-2
YLMT = 3

2170P-3
SYSM = 3
VMLV, VMCR, VMLM, VMFO, VMCL, SHOF = 0
SHFO = 1
PROV = 0
F1LV = 0
LTLV = 0
LTMD = 0
CTLV = 0
MIDE = 63
VM, VMH, VMM, VML = 0

217OP-4
SPIC = 7
SCOL = 7
SHUE = 31
SPIO, SCLO, SHUO = 7
RYR, RYB = 14
GYR = 6
GYB = 4
GAMM = 0
GAMS = 7
GAMR = 3
GAMG = 3
GAMB = 3

With GAMM = 1 I have GAMS = 7, GAMR = 0, GAMG = 0, GAMG = 0, and GAMB = 0.
I use GAMM = 1 for HD content.

21704-P (continued...)
DCTR = 0
APED = 0
DSBO = 7 (Appears to be some kind of offset value, 7 is middle? I use different settings for HD sources so you will just have to experiment)
ABLT = 0
ABLM = 0
DPSQ = 0
BLK = 0 (so you can manually control DSBO)

MID5
POP = 63
Everything else = 0

3DNR
#34, #45, and #61 @ 15

Thanks to Kentech, NightWatchman, Dsperber, and all the others that posted settings and findings. All of the information has really helped in my quest for great picture quality.

Nitewatchman

12-01-06, 07:09 PM

Thanks to Kentech, NightWatchman, and all the others that posted settings and findings. All of the information has really helped in my quest for great picture quality.

You're welcome!

BTW, if I haven't mentioned it before, add'l kudos to Ken for the excellent index of links added to the first post.

Mathesar

12-01-06, 11:31 PM

Does anyone know of a service menu adjustment that might fix distortion from the built in tv speakers? For example every time I turn on my Nintendo Wii the menu intro sound distorts noticeablely through my XBR960's speakers, Ive tried lowering the bass and treble ,even enabled Steady Sound but nothing works, it distorts even if the volume is very low so it looks like I'm looking for some sort of "Input gain" adjustments, does anything like that exist? I sometimes hear distortion when playing other consoles such as Xbox360 but its not very common, it never happens when watching DVDs or Cable channels , Thanks in advance.

doveman

12-02-06, 08:59 PM

The solid line surrounding the rectangle is the outer edge of a 16:9 pattern at 1920x1080. If you adjust things (first background raster, then foreground image, as has been described in detail in this and other threads) so that the solid outer edge of the rectangle is visible at all four edges of the screen, you now have 0% overscan. Anything where this outer edge is not visible is some non-zero positive overscan. And You certainly don't want "negative overscan", where you've got it so that the outer solid lines of the rectangle are "indented" from the edges of the screen. You want that outer rectangle at the extreme outer edge of the screen (at least to start) for 0% overscan.

Sorry for not replying sooner, I've not had much time to tweak in the last few days. Whilst I think I understand the theory of adjusting the background raster and then the foreground image, I'd be grateful if you could tell me if what I'm proposing makes sense.

As I'm starting from a position where I can only see the outer solid line on the right-hand side of the screen, I'm thinking why not adjust the foreground image with the MID settings first so that I can see this line on all four sides, and then adjust the raster to extend this line to the outer edges of the display.

To complicate matters, no matter what I do (on the TV at least), I'm unable to see the outer solid line on the left-hand side of the screen. Is it possible that the PC output is responsible for this and I'd need to adjust something at that end to fix this?

In the circumstances, all I can think to do is to adjust the raster so that the dashed line is at the edges of the screen and then adjust each side in by the same amount, until I reach the point where the left-hand side no longer benefits from any further adjustment.

Sorry if I'm being dumb :)

redheadguy2001

12-03-06, 08:56 AM

gp-se

12-03-06, 07:17 PM

first off i have a 36xs955
i have been dinking around with fixing the scan and i ran into a problem...

i used the mids to twink the 480i and the everything else, however

when viewing 720p the imagine is small in the middle of the screen and doesnt fill the screen, its like watching it on a 30inch tv on my 36, anyone have any idea how to fix what i did?
I have a similiar issue with my 30" HS420

my digital cable is perfect (1080i\480p) - HDMI
my video 6 DVD player is perfect (480p) - Component
however my ps2 -component video 5 doesn't fill the screen.
I looked at the thread for adjusting overscan however on the hs420 the mid options are different so I can't get it right either.

gp-se

12-03-06, 09:11 PM

Here are some of the service code settings I have decided on for now. My 34xs955 is used mainly for playing video games so I am trying to achieve the most accurate and least artificial image from the display of the game console. I'm very happy with the look of my games now, especially higher resolution ones like GT4 or Xbox360 stuff. 480i and 480p games are a tad 'soft' but I still feel this is more accurate as there are less dot crawl issues, higher detail, and just more clarity overall. I can see details and special effects in games that I never noticed before, and textures are much more believable now. The presence of the dreaded jaggies are GREATLY alleviated to the point that even some of the worst offenders look good now.

Many of these settings should look familiar :)

2103-1
SCON, SCOL, SHUE = 7
YDLY = 0
SHAP = 0
SHFO = 3
PREO = 0

*updated*(Further reading revealed that the "2170P-1" settings below are what RWetmore ended up with, and they work much better than the first settings I posted. I can see the subtle differences in skin tones from person to person, even on regular antenna broadcasts, and games like Gears of War that have subtle use of color really look lifelike and amazing.)

2170P-1
SBRT = 31
RDRV= 63
GDRV = 33
BDRV = 33
RCUT = 63
GCUT = 34
BCUT = 32
DCOL = 0

2170P-3
SYSM = 3
VMLV, VMCR, VMLM, VMFO, VMCL, SHOF = 0
SHFO = 1
PROV = 3
F1LV = 0
LTLV = 3
LTMD = 0
CTLV = 0
MIDE = 63
VM, VMH, VMM, VML = 0

217OP-4
SPIC = 7
SCOL = 7
SHUE = 31
SPIO, SCLO, SHUO = 7
RYR, RYB = 14
GYR = 6
GYB = 4
GAMM = 0
GAMS = 3
GAMR = 4
GAMG = 3
GAMB = 3

MID5
POP = 63
Everything else = 0

I'm still applying some tweaks and adjusting a few things here and there, but the settings above seem to work great on all sources through input 5 component. 480i-1080i are all set up the same in regards to the settings I posted. I'm actually using the exact _DRV, _CUT, and _YR settings RWetmore decided on and they work wonderfuly.

Thanks to Kentech, NightWatchman, Dsperber, and all the others that posted settings and findings. All of the information has really helped in my quest for great picture quality.

*edit*

Ok, one more thing I forgot to mention. Now guys, don't be alarmed here when I post these settings. I know they may seem incorrect, but from my many months of experience with this set this is how everything ended up.

Video
Mode = Standard
Picture = Max
Brightness = 31
Color = 31
Hue = 0
Sharpness = Min
Color Temp = Off
Color Axis = Monitor

My reasoning behind those settings are this:
I want as much signal to come from the video output of the console as possible, to do this I set the Picture setting to Maximum and sure enough, with the 'enhancements' bypassed or disabled, I don't get blooming/bleed from the colors. I may go back and re-confirm my findings with the Picture settings though as it was only recently that I arrived at the correct service mode settings. Maybe I'm not understanding this user control correctly? *Comments or suggestions on this are welcome*

The same logic goes behind the Sharpness settings, I want the most accurate representation of the original source material, in the case of a game console it is rendering the images as you see them in an uncompressed manner; so, I ended up with no sharpness enhancements as they tend to augment the original source. I can confirm this works for me as the bridge fighting stage in DOA4 now allows me to see individual leaves in the far distant background clearly, and textures on clothing and objects look more accurate and real. I can even see individual particle effects from the water splashes and distant waterfall very easily now.

As for Color Temp, I'm not absolutely positive as I have to take an educated guess as to how good my gray scale is. It appears good though, as I don't notice any/few discoloration or patches of mismatched color anywhere. I'm sure it isn't perfect, but it appears pretty close to my eyes. Again, I'm no expert.

I'm still having a few minor issues with geometry, convergence, and focus. They aren't 100% perfect but very very good, and I am pleased with them.

It seems strange that I could use many of the same settings others here are using for their set, as each tv is different. However, my thinking is that since each of us are correcting the geometry, setting the raster, fixing the convergence, and so on, this fixes are making our televisions much more "alike" so it would only make sense to use the same settings in other areas.

I just copied these settings to my 30" HS420 and they work great!

Jedah Doma

12-03-06, 10:26 PM

I haven't heard to much talk about the 970. Anyone have some good settings on the 970 that work well.

P.S. Or are these settings pretty much universal for any newer Sony set?

NextGen

12-04-06, 02:30 AM

I just copied these settings to my 30" HS420 and they work great!

Good to hear they helped out.
I've made some slight changes to some other settings that seem to have helped as well. These are kind of experimental, but seem to help the overall range of white to dark.

I've updated the codes, which mainly involved adding some new ones.
Here is the link to the previous post. (http://www.avsforum.com/avs-vb/showthread.php?p=9042644&&#post9042644) .

My next stop will be experimenting video noise reduction settings (http://www.avsforum.com/avs-vb/showthread.php?p=5497510&&#post5497510) .

justsc

12-04-06, 11:18 AM

I haven't heard to much talk about the 970. Anyone have some good settings on the 970 that work well.

P.S. Or are these settings pretty much universal for any newer Sony set?
The 970 has a very different user menu set-up and I'm not sure how well it's SM dovetails with the other Sony crt SMs.

For example, on the 970 in the user menu, the "ticker" that shows how much one is turning up or down settings like brightness, the ticker goes to 100 ticks. On the other models it goes to 62. The has proven challenging for some who copy values over straight across w/o accounting for the different scale.

I realize this is just for the user menu, but it still matters since when one is done making adjustments in the SM, one still has to deal with the user menu. Because the 970 has such a different User Menu scale, I have to wonder how well its SM dovetails with the other models. As far as I know, the 970 is the only model with the 100 tick ticker.

RWetmore

12-04-06, 04:32 PM

I haven't heard to much talk about the 970. Anyone have some good settings on the 970 that work well.

P.S. Or are these settings pretty much universal for any newer Sony set?

In my experience with these TVs, actual numerical values of settings usually differ significantly from set to set in order to achieve the same levels.

NextGen

12-05-06, 01:31 AM

I have a similiar issue with my 30" HS420

my digital cable is perfect (1080i\480p) - HDMI
my video 6 DVD player is perfect (480p) - Component
however my ps2 -component video 5 doesn't fill the screen.
I looked at the thread for adjusting overscan however on the hs420 the mid options are different so I can't get it right either.

Just a suggestion, but you could try out some different games, something like Valkyrie Profile 2 should fill the screen properly, other games like GT4 may not as the developers used black pixels in the 5% overscan area. I ended up keeping a 5% overscan on my set and most PS2 games 'just' fill the screen.

Jedah Doma

12-05-06, 01:39 AM

Just a suggestion, but you could try out some different games, something like Valkyrie Profile 2 should fill the screen properly, other games like GT4 may not as the developers used black pixels in the 5% overscan area. I ended up keeping a 5% overscan on my set and most PS2 games 'just' fill the screen.

Plus VP:2 is a simply amazing game. Good suggestion! :D

NextGen

12-05-06, 02:57 AM

Plus VP:2 is a simply amazing game. Good suggestion! :D

It's also one of the absolute hardest games to "clean up" as the image quality is far from the best, and is one extremely picky game when it comes to color/greyscale brightness and gamma. You really need to have your set nicely calibrated to make the game look good. Even on my set it's a tad pix-elated and jaggy, otherwise it is simply gorgeous.

TV addict#2

12-06-06, 03:21 PM

wanted to give thanks to KenTech for the thread full of info and tools
got the 1/2" black gap at the right edge of my 2.5 year old kv-34910xbrs' screen lined out,much better
long live AVS Forum

doveman

12-06-06, 05:42 PM

According to the MID1 data chart, (charts are frequently cryptic), the column that applies to 480i seem to be Single - Others - Others, and the default value is 0! I checked my TV with a 480i feed from a DVD on V5, and MDHP is, indeed, set to 0.

At the column level, I don't quite get the "Normal" / "Others" distinction. "Single" does seem to refer to a screen-filling main display, and VGA is for a computer-generated video source connected thru HDMI, as far as I can tell from several charts.

This is rather confusing. I've been tweaking my TV (34XBR800) using my PC's output via DVI @ 1080i, and despite the MID1 settings not referring to 1080i, they definitely affect it.

In the MID1-3 groups, be sure to get the values for width and especially height exactly right, as these values control (as I have discovered) the precise sampling of the video for display. Get the vertical ones wrong, and the image quality will be compromised! There are exact ratios between the particular video-line count and the internal digital representation of that video that have to be maintained.

The critical values are DHAR, DVAR, MDHS, and MDVS in MID1; in MID2, DRHS and DRVS (if relevant); and in MID3, VDHS and VDVS. The correct numbers are the ones in the data tables.

The 34XBR910 service code list shows a value of 240 for MDHS. Mine was on 217 and when I tried to set it to 240 the display went blank at 239!. MDVS is set to 128, which is closer to the list's 135 (or 130) but after what you wrote, I'm rather concerned that my current settings might be adversely affecting the picture quality. although I don't see how I can use the settings in the list on my set.

MID3 #1 VDHS (720p) was set to 161. I've changed that to 117 as per the XBR960 table.

The MID3 #4-11 were already set correctly except for #5 which is set to 72, whilst the table indicates 50, although it does show 72 under Expansion). What is Expansion?

Am I doing the right thing, or do I need specific values for a 34XBR800?

vazel

12-08-06, 06:31 AM

raouliii

12-08-06, 08:24 AM

The D-CONV settings don't seem to be doing anything on my 30HS420. I can have the settings on the lowest or highest and it doesn't have an effect in convergence on the pattern I have on screen. Does anyone know what's up?

And the 30HS420 doesn't seem to have the LANDING group. Are there other settings on the 30HS420 that do the same thing as the settings in LANDING?What pattern are you using? IMHO, a grid of thin white lines on a black field is the best pattern. White has all three colors present in equal intensity and the black background allows you to see the misconverged areas.

Beam landing adjustments are physical, yoke adjustments on sets without LANDING service menu entries.

vazel

12-08-06, 08:42 AM

I am using white lines on black. When I adjust focus in 2170D-4 it effects the convergence but I can't improve anything without not only making convergence in another part of the screen worse but also making things blurry.

This is the pattern I'm using.

http://img89.imageshack.us/img89/7320/convtest16by9kk9.th.jpg (http://img89.imageshack.us/my.php?image=convtest16by9kk9.jpg)

DSperber

12-08-06, 09:21 AM

This is the pattern I'm using.I use this similar pattern, which also has a 10% overscan rectangle inscribed (dotted lines inside of outer solid line rectangle) as well single pixel white dots centered within each square of the cross-hatch pattern. These tiny dots are very helpful in working out a compromise between vertical and horizontal convergence settings. Its very easy to see the colored "glow" around that dot if convergence is off in any direction.

But remember that only HORIZONTAL CONVERGENCE issues can be adjusted through the service menu tweaks. VERTICAL CONVERGENCE problems can ONLY be adjusted by magnets installed on the back of the picture tube... something that only an authorized technician should do.

If you have no way to get either a perfectly converged pattern all over the screen or at least an acceptable compromise all over the screen, it may be necessary to get a service technician to pay you a visit (free from Sony, if your set is still under warranty) to do some permalloy magnet work (standard Sony part kit, assuming your technician knows what he's doing). Even though in theory you should be able to manipulate horizontal convergence through the service menu, if you can only get one side of the screen adjusted properly and the other side goes out then magnets can again be helpful in addressing that problem (to get the entire screen to respond in unison when adjusting horizontal convergence all over the screen through the service menu).

vazel

12-08-06, 09:43 AM

D-CONV just doesn't seem to be working. None of the red/blue shadows move even a pixel from where they are when I go crazy with the D-CONV settings. I wonder if something I did in another setting in the SM is preventing the convergence settings from working. When I first got my TV I screwed up the SM settings, I was one of those idiots that didn't write down the default settings. Took me forever to get things right.

I also used this dot pattern.

http://img170.imageshack.us/img170/4193/convtest16by9dotsst8.th.jpg (http://img170.imageshack.us/my.php?image=convtest16by9dotsst8.jpg)

corlay

12-08-06, 09:36 PM

D-CONV just doesn't seem to be working. None of the red/blue shadows move even a pixel from where they are when I go crazy with the D-CONV settings.

me too.
I also have a 30HS420.

Convergence was one of the very last things about the SM that I attempted to tweak, after getting everything else just right.

I started with the setting for center of screen, and just started cranking it up and then down, just to see the affect on the picture. Except there was no affect - I couldn't perceive any change. I tried this first @ 480p (progressive scan from source), and then @ 480i.I then started crankng other settings in the group up & down. Still No perceptible change. So, I just set everything back to the original settings, and shrugged it off.

Would love to get a definitive answer though.

<MarkingThreadWatched>

--
corlay

doveman

12-09-06, 12:09 AM

@DSperber

I had a tinker with the SMPTE colourbars image you gave me and was rather surprised to find that I had to turn the User Menu colour setting up to Max-2 to tune the colour as suggested in DMW (with only the blue gun on). Any idea why I have to set it so high?

As a reference, I started with the settings Nextgen posted, and after setting the colour I've had to set Picture at about 14 and Black Level at 18, and my hue's on the midpoint default (don't know if that's what he meant by 0. The max for the UM settings on my set is 63).

On a related note, it would seem that whilst using static images to set the B&W and colour levels is useful for getting the desktop/games to look good on the TV, this doesn't necessarily translate into good looking video, as Windows uses different levels for the deskkop and the video overlay. If this is the case, what do people use to set the levels for watching video?

DSperber

12-09-06, 02:31 AM

I had a tinker with the SMPTE colourbars image you gave me and was rather surprised to find that I had to turn the User Menu colour setting up to Max-2 to tune the colour as suggested in DMW (with only the blue gun on). Any idea why I have to set it so high?Don't know.

Actually, the real reason I provided it was for the two Pluge bars on that pattern (in the lower-right black section) used to adjust brightness on the XBR960 (turning down brightness until the two Pluge bars just match the black level of the background and thus become invisible as unique bars).

I saw a comment on this pattern way back when I first posted the two test patterns that talked about the "color space" of the SMPTE color bars in my pattern being intended for a PC monitor rather than for the XBR960. This was in response to my comment on the alternative color bar pattern that had been posted earlier, which on my PC monitor I felt looked pale and washed out and unattractive. I was told that the pattern was intended for the XBR960 and not for a PC monitor, because of its different "color space".

So, although I must confess I don't quite understand the theory, I will admit that the colors in the two different color bar patterns were very different when viewed on the two different display devices. Perhaps that's what is causing your unexpected results, because my SMPTE color bars are truly for the wrong color space (if someone else can actually explain what that means and why this might be the case).

On a related note, it would seem that whilst using static images to set the B&W and colour levels is useful for getting the desktop/games to look good on the TV, this doesn't necessarily translate into good looking video, as Windows uses different levels for the deskkop and the video overlay. If this is the case, what do people use to set the levels for watching video?Definitely true that the output of a video card and Windows to the XBR960 is quite different from true HDTV programming from a STB, OTA, etc. It seems that while you might adjust to a test pattern, you will probably do some additional fine tuning and adjustments based on actual HDTV content (e.g. the Leno show and SNL are pretty much reference examples of perfect human skin color, so watching them and fine tuning to get the skin color absolutely perfect is a good idea).

Personally, I did not use that SMPTE color bar pattern to adjust my user menu color settings. I used DVE (from 480p DVD input), and then fine tuned while in 1080i mode watching Leno.

NOTE: everybody's tastes and settings are bound to be different, so anyone's particular values can really only be considered a possible starting point for your own adjustments. I consider the color and picture appearance of my XBR960 to be absolutely perfect (and my 34XBR960 setting values for 1080i are significantly different than the partial list of values enumerated previously by NextGen, who has a 34XS955). But then I don't play video games at all, almost never watch DVDs, and almost exclusively watch 720p/1080i from my cable DVR or D-VHS. So my objectives are probably different from those who have game devices as a source.

DSperber

12-09-06, 09:51 AM

As a reference, I started with the settings Nextgen posted, and after setting the colour I've had to set Picture at about 14 and Black Level at 18, and my hue's on the midpoint default (don't know if that's what he meant by 0. The max for the UM settings on my set is 63).I decided to reorganize my papers on which I had recorded my user/service menu settings for my 34XBR960. I had numerous full and partial versions of PDF's and XLS's originally posted by KenTech (and possibly others) quite some time back and decided I would make an exhaustive enumeration of what my set looks like and tabulate it in a new spreadsheet (based on an original from KenTech).

So while it is no more useful nor valuable than anybody else's, I'm making my new spreadsheet available for reference.

I went through the complete service menu while tuned to discrete sources at 480i, 480p, 720p and 1080i and recorded the four complete sets of values. The 480i input was through S-video and the other three inputs were through component video.

Then I also sent 720p and 1080i via firewire, and again went through the complete service menu and recorded those two complete sets of values. Turns out the service menu calls firewire "ATSC", which is also what it calls its own internal ATSC tuner, and I'm guessing that's what it might also call HDMI input. In other words the values I gathered for the digital input versions of 720p and 1080i (via i-Link) were different than the component video versions of those same resolutions but the same as the values shown when using the internal ATSC tuner. And I'm guessing the HDMI digital input version of those resolutions will show similar values in the service menu (but I didn't try that experiment since I don't use HDMI input on my set).

Anyway, the six columns of data values in the spreadsheet are titled with a heading that conveys whether the input at that resolution was from "C" (component video) or "D" (digital, i.e. firewire and ATSC tuner in my current effort). The complete set of values for C-1080i is shown in that column. The values in the other five columns are blank if there is no difference from the C-1080i column. Only different values for the other five columns appear as non-blank. I thought that would make reading easier.

NOTE: my set was NOT adjusted by an ISF technician. I had some work done by a Sony factory technician (who didn't do very much), a second local authorized Sony service person (who mostly worked with magnets to eliminate my convergence and curvature problems, and then spent some time in the service menu), and finally myself... using DVE at first and then fine-tuning just by using actual broadcast programs whose color I knew was perfect. As such, the large majority of my values are very likely "untouched as it came from the factory", with only the crucial ones reflecting my own work. And of course the geometry values are unique to my set because of the magnet work done.

Nevertheless... you can use it for reference while working on your own set. I feel that the picture quality and color my XBR960 produces is remarkable.

doveman

12-09-06, 12:25 PM

I saw a comment on this pattern way back when I first posted the two test patterns that talked about the "color space" of the SMPTE color bars in my pattern being intended for a PC monitor rather than for the XBR960. This was in response to my comment on the alternative color bar pattern that had been posted earlier, which on my PC monitor I felt looked pale and washed out and unattractive. I was told that the pattern was intended for the XBR960 and not for a PC monitor, because of its different "color space".

So, although I must confess I don't quite understand the theory, I will admit that the colors in the two different color bar patterns were very different when viewed on the two different display devices. Perhaps that's what is causing your unexpected results, because my SMPTE color bars are truly for the wrong color space (if someone else can actually explain what that means and why this might be the case).
That's an interesting suggestion, although I must admit I don't understand it either. I'll try some of the other color patterns from this thread. I think my monitor is somewhat knackered anyway, as video playback is ridiculously dark, but that's what the TV's for and the monitor's fine for PC use.

Definitely true that the output of a video card and Windows to the XBR960 is quite different from true HDTV programming from a STB, OTA, etc.
Sorry if I wasn't clear, but what I was trying to say is that the output from the video card appears to differ when it's outputting overlay (full-screen) video playback, so tuning the TV to the desktop (including applications such as ACDSee) doesn't result in a good image when playing video.

Also, whilst a DVD like DVE should do fine for tuning to PC DVD playback, I'm not sure if it will be suitable for tuning to AVI,WMV, etc playback as the different codecs involved might affect things.

Anyway, this is more suited to the HTPC forum. Thanks for your help adjusting the geometry of my TV, although I still need clarification on KenTech's statement that the MID3 values for horizontal and vertical size MUST stay at the defaults as that throws a spanner in the works for me.

doveman

12-09-06, 12:51 PM

So while it is no more useful nor valuable than anybody else's, I'm making my new spreadsheet available for reference.
Thanks for providing that. I'm looking forward to trying your settings on my TV.

As you said before, gaming and video are quite different, and as I'm going to be using my TV for both, it's great to have your settings as a reference as well as Nextgen's.

Jedah Doma

12-09-06, 01:18 PM

justinblair2

12-09-06, 05:05 PM

Guys i got a bit of a problem, i was tweaking in the service menu, actually i was just inputting some of your guys tweaks for reference, and i completely wrote down everything i changed and its normal values, so something starts to look funny , and i just stop and input all my normal values back in. But now it seems if the tv is way to dull. I upped the color in the normal settings and it seems to correct it but the color didnt use to have to be that high? Any thoughts ?
Also, i was doing this on input 4 with component cables *360*, could i just be able to plug it into input 5 and get the defaults and make corections if i accidently messed something up? One more question, say if i had my video set to pro and tweaked it in the service menu would the values be different if i were in standard mode? Last one i promise, is there anyway to return the set back to the defaults it was given to me?

My set is the xbr 970

Update I found that really the only problem seems to be that the color in the regular menu, when set at its default is way to low *51 is blakc and white* I dont really recall it being likes this until i started tinkering. if that helps lock into the problem.

NextGen

12-09-06, 05:07 PM

first off i have a 36xs955
i have been dinking around with fixing the scan and i ran into a problem...

i used the mids to twink the 480i and the everything else, however

when viewing 720p the imagine is small in the middle of the screen and doesnt fill the screen, its like watching it on a 30inch tv on my 36, anyone have any idea how to fix what i did?

The 955's are completely different from the 960's, at least mine is. All of my settings for horizontal and vertical size are in the MID 2 groups of codes. What I did was adjust ASPT and SCRL of my 480p?, 720p,1080i content to match that of my 480i settings, then I adjusted Vert and Horizontal via the 4 MID 2 codes (DHHP and up) I can't recall the exact name of the other 3 as I just got done painting and the tv is unplugged.

doveman

12-09-06, 07:36 PM

Update: Note(this pertains *only* to 480p/720p+1080i and working with/looking at "color" when working In SM ) ..... *Do* note that when you are working in SM, (well, this is the case on my set(KD34XBR960 manufactured Jan 05 at least), when 480p, 720p or 1080i signals are input, the proper "color matrix" for ATSC/NTSC(or 480p vs 720p/1080i/etc) sources is not selected "automatically" as occurs if you are operating the set normally outside of service mode. Therefore, when looking at "color" from within the SM, you need to, for example set(or verify) 2171CXA/MTRX=0 for 480p NTSC sources(such as from DVD), and 2171CXA/MTRX=1 for 720p/1080i ATSC sources for proper color decoding to occur, otherwise what you'll see color decoding wise while working in SM will be *inaccurate* for 480p/720p/1080i sources. Note that Only 480p/720p+1080i signals are effected, and that MTRX is a temporary setting which is not saved.
:end update
On my 34XBR800, when looking at colour *outside* the SM in a 1080i PC->DVI screen, this auto matrix switch isn't working. I can clearly see the difference when I change CXA2151/MTRX to 1, but outside the SM it stays on the 0 matrix. Any ideas on what I can do to fix this?

jjmilo

12-09-06, 07:58 PM

Has anyone with a 34XBR910/960/970 who also owns a Blu-Ray player used the built in test patterns on the Sony Film bd DVD's ? If so what results have you achieved with the 1920 x1080i monoscope overscan,convergence and crosshatch patterns.

NextGen

12-10-06, 03:27 PM

Just wondering if anyone knows what the ENHA settings are on the 34xs955's?

I stumbled across them the other day and they all seem to be some kind of horizontal enhancements or "settings" that effect sharpness and focus. As of right now I have:

HSHP = 0 and it was previously set to 50. It looks like Horizontal Sharpness and that is also how it acts. I've set it to 0 to disable it for the time being.

HSFO = 15 (Highest Setting) This appears to be either Focus or Focus Offest. When set to 15 my focus is insanely focused, especially after adjusting the 2170P-4 focus paramaters.

Seems like many of my convergence problems are minimized as well.

Nitewatchman

12-10-06, 05:08 PM

On my 34XBR800, when looking at colour *outside* the SM in a 1080i PC->DVI screen, this auto matrix switch isn't working. I can clearly see the difference when I change CXA2151/MTRX to 1, but outside the SM it stays on the 0 matrix. Any ideas on what I can do to fix this?

Interesting ... .... I don't really have any ideas .... Suppose I might start to see if you can control or "fix" the "colorspace" as output from the PC's video card/etc to either match the (presumably) NTSC colorspace apparently being used, or, perhaps a even better solution might be to find a way to ensure the correct "info" on the colorspace is being sent to the display so it can "interpet it correctly ---

I do notice that, according to the Service Code listings from Service Manual for XBR960 perhaps the MTRX parameter may be handled a little differently for its HDMI input than what I experienced with Component video, as I notice there are 2 columns shown for HDMI in that listing -- one column with a value of "0"(in fine print it says : uTiny/$6D_00 Byte1/Bit6=0 ) and another column with a value of "1" (in fine print for this one it says : uTiny/$6D_00
Byte1/Bit6=1 ). I also of course don't know if there are any differences here invloving the Cxa2151 parameters for the XBR800.

I do know on a couple of occasions at first I could have *swore* the same thing happened to me the first time I looked at it outside of SM via component inputs + 480p (NTSC via DVD player/component video+AVIA or DVE color bars), only to find out later it seemed I must have been hallucinating as it turns out upon checking it again it was(and still is) working OK outside of SM ...

FWIW, so far I've also had no problems with this+it using the correct colorspace outside of SM with 1080i(or 720p) via PC/ATI Radeon X1600 Pro via component connections .. I haven't tried it/picked up a DVI to HDMI cable yet to see what might happen there ...

doveman

12-11-06, 07:52 AM

I do notice that, according to the Service Code listings from Service Manual for XBR960 perhaps the MTRX parameter may be handled a little differently for its HDMI input than what I experienced with Component video, as I notice there are 2 columns shown for HDMI in that listing -- one column with a value of "0"(in fine print it says : uTiny/$6D_00 Byte1/Bit6=0 ) and another column with a value of "1" (in fine print for this one it says : uTiny/$6D_00
Byte1/Bit6=1 ). I also of course don't know if there are any differences here invloving the Cxa2151 parameters for the XBR800.
I notice that Dsperber's Excel code list for his 34XBR960 shows the MTRX parameter set to 1, whereas on my XBR800 it defaults to 0. If his 960 actually defaults to 1 when in the SM in 1080i then that appears to be different to your (and my) set. Note that this is the value he recorded when using 1080i via component, although he also tested with Firewire (which he's guessing is the same as HDMI) and didn't record any change in the MTRX value.

I do know on a couple of occasions at first I could have *swore* the same thing happened to me the first time I looked at it outside of SM via component inputs + 480p (NTSC via DVD player/component video+AVIA or DVE color bars), only to find out later it seemed I must have been hallucinating as it turns out upon checking it again it was(and still is) working OK outside of SM ...
I've checked it with Dsperber's colour bar chart and also the colour bars from AVIA and when MTRX is set to 0 the green bar (4th from left, to the left of the purple bar) is a noticeably brighter, garish/fluorescent green than when it is set to 1. If you could check on your set and see whether the green bar is brighter when MTRX is set to 0 or 1 that would be very helpful.

FWIW, so far I've also had no problems with this+it using the correct colorspace outside of SM with 1080i(or 720p) via PC/ATI Radeon X1600 Pro via component connections .. I haven't tried it/picked up a DVI to HDMI cable yet to see what might happen there ...
Unfortunately my PC can't do component out, although it does composite so I'll see what things look like using that. I tested the AVIA colour bars on my Xbox which is connected via composite (Input/Video 3), and the MTRX setting doesn't affect things at all. I'll do some more comparisons this evening.

Nitewatchman

12-11-06, 01:38 PM

Doveman,

Note that At least on my set, I don't think the MTRX parameter holds any signifcance other than as a temporary setting to allow you to manually chose the correct value for the source you are using(On my set at least --- for NTSC 480p the correct value is MTRX=0, for ATSC 720p/1080i the correct value is MTRX=1) when you are working in Service Menu --

The correct colorspace(or color "matrix") for the source you're using should be applied "automatically" when you are outside of SM and operating the set "normally". Perhaps This may have as much to do with the output of your source "device" as the display itself ...

Also on my set at least, I've seen no real rhyme or reason to what the value for MTRX "defaults to" when you first enter SM ... Usually seems to be the correct value for the last source(480p, 720p or 1080i) you were using when you were last outside of SM, but If I recall correctly I may have ran into an exception or two.

I've checked it with Dsperber's colour bar chart and also the colour bars from AVIA and when MTRX is set to 0 the green bar (4th from left, to the left of the purple bar) is a noticeably brighter, garish/fluorescent green than when it is set to 1. If you could check on your set and see whether the green bar is brighter when MTRX is set to 0 or 1 that would be very helpful.

Sending a 480p NTSC source signal(from Avia DVD - Color bars) --- Green bar appears bars "brighter", or more garish/floruescent with MTRX=0 than is the case with MTRX=1, and Magneta appears more garish or "flouresent" with MTRX=1. There is also, for instance noticably more(noticably too much if you've done much of this sort of thing) "blue in the Cyan" with MTRX=1 -- However, MTRX=0 is the "correct" value in this case(with 480p NTSC source signal on my set anyway), and is what results in accurate color balance(including but not limited to AVIA or DVE color tests using RGBS to turn the R/G/B guns on indivdually) for this source given the other color related settings I'm using (such as 14-15-6-4 for 2170P4 RYR~GYB).

If however, I use a 720p/1080i test pattern (such as internal color bars located in the QM section of SM on my set, or a 1080i color bar test pattern via memory stick ), then MTRX=1 is the "correct" value for proper color balance to occur.

Note that I've verified on previous occaisons that the use of the correct Color Matrix for the input source(either NTSC 480p, or ATSC 720p or 1080i) happens "correctly" and automatically "outside" of SM, no matter where I had "MTRX" set the last time I was working in SM.

I have not yet looked at it via a 720p/1080i test pattern via the PC -- It hasn't seemed necessary as I have done comparisions from recordings of MPEG2 transport streams from local broadcast stations(captured via firewire from XBR960's internal ATSC tuner then played back on the PC and output via component video back to the Display), and the color from the "recording" matches what I get from the internal tuner (antenna) "input" perfectly. And, as you may have noticed -- if the incorrect "colorspace" is being used (such as a "wrong" value for MTRX in the SM"), it is quite noticable that the color is well, "wrong" ....

I tested the AVIA colour bars on my Xbox which is connected via composite (Input/Video 3), and the MTRX setting doesn't affect things at all. I'll do some more comparisons this evening.

I can probably save you some time, there. Note that the MTRX parameter HAS NO effect on 480i/NTSC SD signals, so you won't see any change with MTRX=0 or 1 with signals via composite or S-Video inputs. Those signals aren't processed by the 21x1CXA chip, only 480p, 720p+1080i signals are.

TVOD

12-11-06, 03:30 PM

KenTech

12-11-06, 03:31 PM

Just wondering if anyone knows what the ENHA settings are on the 34xs955's? . . . HSHP = 0 and it was previously set to 50. It looks like Horizontal Sharpness and that is also how it acts. I've set it to 0 to disable it for the time being.

HSFO = 15 (Highest Setting) This appears to be either Focus or Focus Offest. When set to 15 my focus is insanely focused, especially after adjusting the 2170P-4 focus paramaters.I'm having trouble locating these parameters in my service-code listings. What code-group did you find them in? In other words, exactly how can one navigate to these parameters in service mode? Theoretically your 34XS955 should be covered by the listing I have, but who knows -- maybe firmware changed in late production.

doveman

12-11-06, 07:52 PM

Nitewatchman

Thanks for the clarification about MTRX. I did understand that it was only a temporary parameter for use in the Service Menu, but I thought that MTRX 1 was used (automatically, in normal operation) for 480p/720p/1080i and that MTRX 0 was used for 480i.

Thanks also for checking the colour bars with the two different MTRX settings. It seems that for some reason, my set is NOT switching to the correct colour space when dealing with 720p/1080i signals via DVI from my PC (outside the SM). I've checked and can confirm that the Magenta and Cyan bars look as they would if MTRX=0.

Probably the easiest thing for me to try is to get a memory stick and see if that causes the set to switch properly. I intend to get a component lead for my Xbox and I know that will allow me to run Xbox Media Center (XBMC) at the higher resolutions but I'm not sure if that will necessarily allow me to display a 720p/1080i image at the correct resolution.

I don't have any experience with a MS and the TV manual indicates that there are certain directory/filename conventions that have to be observed for it to work. When loading the jpegs from my PC to the MS, what names do I need to use?

I presume I can use the 1920x1080 colour bar that DSperber provided and resize it to 480p and 720p resolution and put all 3 copies on the MS for testing purposes.

Nitewatchman

12-11-06, 07:53 PM

The difference with the green and magenta bars is probably due to the difference in GRB weighting of the SD NTSC/PAL/601 Y' (luma) channel and that in HD's 709. It's ~ 59%/ 30% /11% in SD while in 709 it's ~ 71.52% /21.26% /7.22%. This means that green is much more a part of the HD luma channel than the SD.

Yes, I think so too. Excellent info.

Concerning Doveman's reports of what he has observed when not in SM : I suppose what I'm wondering more than anything else is if it's possible his PC video setup may, for some odd reason, be sending(transcoding to) 601 colorspace with HD/1080i via DVI when the display is (presumably) "expecting" a 709 HD colorspace to be used for 1080i(or 720p).

Quickly doing a little googling I can't find anything specific concerning that, but it seems there may be some concern about some upconverting DVD players, and whether or not they are transcoding to the 709 HD "colorspace" values "expected" by HD display.

Nitewatchman

12-11-06, 08:54 PM

Doveman,

Looks like today several of us checked this thread and posted within a minute of each other ;) ..

It seems that for some reason, my set is NOT switching to the correct colour space when dealing with 720p/1080i signals via DVI from my PC (outside the SM). I've checked and can confirm that the Magenta and Cyan bars look as they would if MTRX=0.

Or, As I mentioned in my response to TVOD above ... Perhaps instead its possible that your PC Video via DVI is sending 601 (NTSC) colorspace values to the display for 720p+1080i rather than the expected 709 HD colorspace ....

I'd think one might *assume* that with any given HD display it would seem logical HD video should be sent to the display with the 709 HD "colorspace" values ..... But, then again I wonder if the same assumption is as likely to be true concerning any given PC monitor that may be attached to your DVI output and used to display HD(or SD) video ... If not, again I'm just guessing and speculating and I really have *no* idea but perhaps, in some cases there are good reasons for say, a PC video card to transcode from the HD "colorspace" to NTSC 601 colorspace for output, even though HD is being output .... If so, it seems however there ought to be some way for the user to change it via software (or perhaps via windows Registry key/etc) ...

Also -- It may work that way(I don't know and have no way to tell for sure), and the way MTRX works within SM it's easy to think it does ---- But, I don't know if it's so much the set "switching" to using the correct colorspace as it perhaps is the set "expecting" and therefore "using"(in a sense) 709 colorspace for HD(whether the "correct" values are there from the source or not), and 601 NTSC colorspace for SD in a "hardwired" sort of way when not working in SM.

I don't know, and again - hopefully someone with more experience with this issue can chime in, as I'm mostly just stabbing in the dark here with my comments as I have no experience with this concerning DVI as of yet or how/if it may vary vs. component HD video(or from the MS) sent to the set -- But -- it seems to me that *perhaps* it may be more likely the PC video setup may not be sending the correct "colorspace" (or, if there is any associated metadata missing which might need to be there to "tell" the display what "colorspace" to use) the set is "expecting" rather than the set itself not "switching" to using the "correct" colorspace ...

I don't have any experience with a MS and the TV manual indicates that there are certain directory/filename conventions that have to be observed for it to work. When loading the jpegs from my PC to the MS, what names do I need to use?

Can't recall anything specific, but, i can say, for instance, subdirectory\filenames such as x:\testpattern\1440x1080spokes.jpg, or x:\photoshd\gcany11.jpg work here ...

I presume I can use the 1920x1080 colour bar that DSperber provided and resize it to 480p and 720p resolution and put all 3 copies on the MS for testing purposes.

Well --- The only resolution sent from the MS for JPG's(on XBR960 anyway) is 1920x1080 --- a 1280x720, or 720x480 (or say 1366x900 - AR isn't effected) image would only be "smaller"(although you could zoom it or pan using those functions via the MS viewer), on 16x9 display, 1440x1080 would be a 4x3 image centered with pillar(black) bars on the sides.

It will work for 1080, however, and judging from my results, I fully expect on your set it will use the correct colorspace for HD (in other words I expect you'd get the same results outside of SM via MS as you would using MTRX=1 when you are within SM )... I've been quite busy as of late and haven't had much time to experiment with any recently posted test patterns/etc, but I expect it would work with the pattern you mention, and I can say it worked that way for me with the "color" JPG (DVE "like") test patterns Ken Posted earlier in this thread.

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

In any case, let us know how it goes and what you find out!

doveman

12-11-06, 10:28 PM

I'd think one might *assume* that with any given HD display it would seem logical HD video should be sent to the display with the 709 HD "colorspace" values ..... But, then again I wonder if the same assumption is as likely to be true concerning any given PC monitor that may be attached to your DVI output and used to display HD(or SD) video ... If not, again I'm just guessing and speculating and I really have *no* idea but perhaps, in some cases there are good reasons for say, a PC video card to transcode from the HD "colorspace" to NTSC 601 colorspace for output, even though HD is being output .... If so, it seems however there ought to be some way for the user to change it via software (or perhaps via windows Registry key/etc) ...
You may have a point that if using the PC's DVI port to drive a 2nd PC monitor it would be correct for it to operate in 601 colorspace (I confess to having no idea, I'm just guessing here!). Having said that, the driver for my graphics card has a tick option for the 2nd display/DVI port 'Treat display as HDTV' which I've obviously ticked. When this is ticked, a further option appears offering the choice of 480p/720p/1080i TV format, and I would expect the latter two to cause the card to send the metadata/operate in the 709 colourspace.

Another thought that occurred to me is that, even if the DVI output is set to 720p or 1080i, maybe it only switches to the 709 colourspace when HD video (720p/1080i) is actually being played. Again, this is a complete guess and I don't even know if it's possible but it would account for static test images or AVIA DVD patterns displaying in the "wrong" colourspace. A reason why this might make sense is that that Windows desktop and games are probably not designed to be shown in HD colourspace and so would look wrong.

I'll ask on the Nvidia forum to see if anyone knows anything about this and see if I can find a suitable HD colour bar test pattern.

It's probably completely unrelated, but I should mention that I had to set 2170P-3 #14 UCOF=7 (max) for the DVI input. This was after tuning the TV to AVIA using my Xbox via composite input (at UCOF=2), which resulted in User Menu settings:
Picture: 25
Brightness: 18
Colour: 34
Hue -1/30/+1 Red (not quite sure what the convention for describing this is yet).

Thanks for the MS info. Sounds fairly straightforward and a 1080i image is all I really need to try. Might be a few days before I get a chance to try it as I'll have to visit a friend who's got a MS slot on his PC but I'll keep you all updated.

doveman

12-11-06, 11:32 PM

Guys i got a bit of a problem, i was tweaking in the service menu, actually i was just inputting some of your guys tweaks for reference, and i completely wrote down everything i changed and its normal values, so something starts to look funny , and i just stop and input all my normal values back in. But now it seems if the tv is way to dull. I upped the color in the normal settings and it seems to correct it but the color didnt use to have to be that high? Any thoughts ?
Also, i was doing this on input 4 with component cables *360*, could i just be able to plug it into input 5 and get the defaults and make corections if i accidently messed something up? One more question, say if i had my video set to pro and tweaked it in the service menu would the values be different if i were in standard mode?
It's possible that you lowered UCOF (2170P-3 #14 on my XBR800) so you could check that and try turning it up, although as far as I know UCOF is mode AND input dependent, so if you only changed it for PRO, Standard should still be as it was.

jameasson

12-12-06, 11:26 AM

hi,
i got the xbr970 recently
and adjusted the horizontal and vertical service menu settings so the whole screen fits perfectly. but when i pop on my xbox 360 on the bottom of the screen theres a verry small portion (probally 1/2 inch) that i can see the picture but its blurry when i move my character. is there a setting in the service menu where i can correct this or any recommendations?
thanks in advance
james

Nitewatchman

12-12-06, 02:58 PM

Another thought that occurred to me is that, even if the DVI output is set to 720p or 1080i, maybe it only switches to the 709 colourspace when HD video (720p/1080i) is actually being played. Again, this is a complete guess and I don't even know if it's possible but it would account for static test images or AVIA DVD patterns displaying in the "wrong" colourspace. A reason why this might make sense is that that Windows desktop and games are probably not designed to be shown in HD colourspace and so would look wrong.

I thought about that a little as well .... I don't know with 100% certianity, either -- but, I would hope not, in fact I don't even see how that would "work" very well ..

For instance, you could be displaying Windows Desktop (or any other graphics "originating" from the PC) on a PC monitor, a "plain old" NTSC analog TV, or a HDTV -- and, it seems to me the colors "should" be the same on all those displays, regardless of the native resolution/etc. of the source "material". In other words, in this case I'd think to "fit" any standard required for the video to be displayed properly via any given "standard" or on any given display device (VGA, XVGA, VESA, NTSC, or HD display), any necessary standard/etc required should/would need to be supported handled by the video output device and/or display hardware.

Or, to look at this another way ... A 1920x1080 MPEG2 HD stream(or 720x480 from say, a DVD) displayed on my PC monitor(at whatever resolution) should(and does seem to, pretty much) have the same "colors" as it does on a HD display at 1920x1080i .... Same thing with a JPEG image displayed on PC monitor running at say, 800x600 vs. using the MS on the Sony at 1920x1080i ....

DSperber

12-13-06, 06:01 AM

Thanks for providing that. I'm looking forward to trying your settings on my TV. As you said before, gaming and video are quite different, and as I'm going to be using my TV for both, it's great to have your settings as a reference as well as Nextgen's.I've spent some more time on my spreadsheet for the 34XBR960 (making it prettier, more readable, and most importantly more accurate based on further research with i-Link, HDMI, and ATSC) and am again making it available.

I've also reversed my original notion that showing fewer values (i.e. only those which differ from the C-1080i column) would make it more convenient to reference. I've decided to populate all cells in all columns, but applying color to those cells which differ from the C-1080i column values in the same row. Furthermore, separate highlighting colors are chosen to indicate 1080i, 720p, or 480p/i, making it very easy to see where the differences are.

I've also added a new column to show the RANGE of possible values for each item. In addition, when a value in a cell is other than the default (according to my edition of the Sony Service Manual ) I now show the default value enclosed in parentheses following the actual current setting value. Single numbers in a cell mean that the value is still the Sony default. This makes it very easy to see which items are still sitting at Sony-provided defaults and which items have been changed during tuning/tweaking (and by the two Sony technicians who worked on my set during its first few weeks of life in my home).

As it turns out, the i-Link and ATSC sources DO display the identical service menu settings values. In fact both show up in the service menu with a caption of ATSC, leading one to believe that both inputs are referencing the same single set of values. However the two sources are unique in that they have separate user menu settings memories.

But, as it turns out, the HDMI input is SEPARATE from i-Link/ATSC. It has its own unique set of service menu settings values. In some instances values for a given service menu item are different for HDMI vs. the corresponding i-Link/ATSC value.

Consequently, this new version of my spreadsheet shows two new separate columns for HDMI input at both 1080i and 720p resolutions. The "D-1080i" and "D-720p" columns on my old spreadsheet have been renamed to "i/A-1080i" and "i/A-720p" on the new spreadsheet, and pertain specifically to i-Link and ATSC.

Finally, I did a slight adjustment to geometry while I was gathering all of this data. I'd been bothered for some time by some bowing on the left and right sides at the top and bottom of the screen, shown most obviously when a score bar is at the top of the screen (e.g. in a football game) or when a "crawl" is present at the bottom of the screen. The curvature is also visible when viewing an OAR movie where the black bars on top and bottom are in marked contrast to the movie itself, and the imperfection from totally linear top and bottom edges of the image are distraction.

Anyway, I did come up with a new compromise of adjustments to VPIN, NSCO (in 2170D-1), and some matching adjustments to HPOS, UCP and LCP (in 2170D-2). The results are very satisfying, with even the onscreen graphics from the XBR960 itself (e.g. DISPLAY) which are supposed to be rectangular and linear now appearing much more straight and oriented more rectangularly than they were before. These somewhat different item values are reflected in this new edition of the spreadsheet.

Again, you're welcome to do whatever you want with this spreadsheet... including copying it and then editing it with your own values, if for no other reason than that it is in a very useful format and presentation.

justsc

12-13-06, 11:14 AM

You have done an outstanding job with this spreadsheet!

Thank You.

KenTech

12-13-06, 02:37 PM

I've spent some more time on my spreadsheet for the 34XBR960 (making it prettier, more readable, and most importantly more accurate based on further research with i-Link, HDMI, and ATSC) and am again making it available.This is an excellent, excellent resource! THANK YOU for the trouble you've gone to and for posting it.

Psst . . . Do you know what NextGen is talking about, quoted in my message a few posts above? (ENHA, HSHP, etc. for his 34XS955)

DSperber

12-13-06, 03:49 PM

Psst . . . Do you know what NextGen is talking about, quoted in my message a few posts above? (ENHA, HSHP, etc. for his 34XS955)No idea. But then my experience is confined to the XBR960 and the Service Manual I purchased (along with the combined knowledge of this forum and specifically this thread).

From what I've seen, according to all of the previously available spreadsheets and PDFs (from which my current version was obviously derived) for XS models and XBR models these new items have never been mentioned before. Of course it's possible his set is either older or newer than the publication date of the service data sheets we've been looking at for reference.

For example, in my edition of the DA-4 Chassis manual (version 9-965-965-05, latest revision date 2/2005) I show a group of items for the XBR960 in the 2170P-4 group (UPIC, UBRT, UCOL, UHUE, USHP, UTMP) that theoretically appear between SHUO and RYR. And yet, my set's actual firmware does not have those items. In contrast, my set contains the new group HDMI (with three items) right after the DCRV group, but this group is not documented at all (although it appears that the INP item in this group is a 0,1 item: 0=no HDMI cable connected, 1=HDMI cable connected).

So who knows?

doveman

12-13-06, 09:02 PM

DSperber,

Many thanks for making your spreadsheet even more excellent.

doveman

12-15-06, 06:28 AM

Or, to look at this another way ... A 1920x1080 MPEG2 HD stream(or 720x480 from say, a DVD) displayed on my PC monitor(at whatever resolution) should(and does seem to, pretty much) have the same "colors" as it does on a HD display at 1920x1080i .... Same thing with a JPEG image displayed on PC monitor running at say, 800x600 vs. using the MS on the Sony at 1920x1080i ....

I stumbled across this thread just next door :)

http://www.avsforum.com/avs-vb/archive/index.php/t-675384.html

There's some links to some interesting pages in this thread. At the bottom of this page, it demonstrates what a picture looks like when it's filtered through the wrong matrix

http://www.sigmadesigns.com/public/Support/chromaticity.html

and this one shows the results of filtering a colour bar pattern through the wrong matrix.

http://www.avsforum.com/avs-vb/attachment.php?attachmentid=32137

I get the impression that what happens is the source (eg HD DVD player) is meant to look at the material and see whether it's SD or HD. Then it takes the material and converts it, using the relevant matrix, from whatever format it's in on the disc, to whatever format is used by the chosen output connecter. When the TV receives this, it then has to convert it to it's "native" format (but it may go through other conversions before this, depending on what processing is being used on the TV).

Probably it's only with internal (tuner/MS) sources that the TV has to select and use the relevant matrix.

I still want to find out what hardware/software on the PC is responsible for doing this. To be consistent with other displays, I would have thought it also has to filter HD material through the 709 matrix and then code the output into RGB values to send to the monitor.

Unfortunately, the friend who I thought might have a MS slot on his PC doesn't, but I think another friend might so hopefully I'll still get a chance to test it sometime.

Oliver Deplace

12-15-06, 05:29 PM

This is an excellent, excellent resource! THANK YOU for the trouble you've gone to and for posting it.

Psst . . . Do you know what NextGen is talking about, quoted in my message a few posts above? (ENHA, HSHP, etc. for his 34XS955)
From my notes:
34XS

VERSION = 1st SM screen
3D-COMB
2103-1

2170P-1
2170P-2
2170P-3
2170P-4

2170D-1
2170D-2
2170D-3
2170D-4
2170D-5
D-CONV
LANDING
CXA2171
MID 1
MID 2
MID 3
MID 5
DRCMFV1
ENHA------->0-HSHP, 1-HSFO,2-HPOR,3-HLTL,4-HLTM,5-HAPL,6-HAPA,7-HCTL,8-HCTM
AUDIO
SNNR
3DNR
DRCV
HDMI
OP
QM (blue)
QT (blue)
ID

NextGen

12-16-06, 04:27 AM

I'm having trouble locating these parameters in my service-code listings. What code-group did you find them in? In other words, exactly how can one navigate to these parameters in service mode? Theoretically your 34XS955 should be covered by the listing I have, but who knows -- maybe firmware changed in late production.

I wondered that too, about the firmware changes. There are quite a few settings that are different on mine from what you and others have posted. For instance I cannot find the correct service codes for adjusting the size of a 4:3 pattern for the normal mode. It just doesn't seem to exist. I do not have the MID2 DR values anywhere on my set. And I sure would like to know what all the different SNNR values are for.

As Oliver posted the ENHA settings are right before you get to AUDIO on my set, up there past the MID5 settings. I would highly recommend others to try adjusting the HSFO value before attempting a hands on focus pot adjustment. I have absolutely no need for any form of sharpness after doing the adjustment.

shaddix

12-18-06, 07:27 PM

as far as overscan adjustments go, how exactly am i supposed to get the entire image on the screen if i am not allowed to adjust the values in MID3

TomPiltoff

12-18-06, 09:48 PM

GlenC

12-19-06, 12:00 AM

ptchristensen

12-19-06, 12:02 AM

Hmm....I need some help. I own the KD30XS955 and I'd like to improve my picture quality as much as possible, but I'm terrible, terrible with numbers. Just looking at those charts makes my brain hurt. Could anybody provide some very simple instructions to make even very basic improvements on my set via the service menu?

That's easy...
Stay out of the service menu, buy a calibration dvd like "Digital Video Essentials" and tweak your TV from the menu settings. If your TV has bad basic settings you will be amazed by what can be done her.

If that is not enough - tough it up and read the numbers, there is no way around those, if you want to tweak from the service menu.

jpl3447

12-19-06, 01:07 AM

There is a learning curve involved. All the info is in here at the beginning. It took me about a month of studying before I went into the service menu.
Don't feel bad... as the time goes by and if one keeps reading and printing out stuff to compile into a notebook (service menu charts and parameters, tips and tricks, other user discoveries) then it will become easy.

KenTech

12-19-06, 03:37 PM

What if you goof making a change?Well, that's really helpful, Glen.

This is a help forum--a do-it-yourself forum--not an instill-fear forum. There are other forums for those other purposes. Anyone who reads here knows already that one can screw up big-time, and there are folks here who can encourage and help. Fostering paranoia is not welcome. Informative cautionary instruction is.

DSperber

12-19-06, 03:42 PM

as far as overscan adjustments go, how exactly am i supposed to get the entire image on the screen if i am not allowed to adjust the values in MID3Need more info. First, what model is your set?

Then, assuming you have one of the CRTs we typically discuss here, why would you not be able to get into MID3? Is the MID3 group not present?

If you have an older model XBR (e.g. XBR800) you might need to use the MID1 items rather than MID3. The item names are slightly different but they accomplish pretty much the same thing, insofar as moving the image around and resizing it in the horizontal and vertical dimensions.

GlenC

12-19-06, 03:53 PM

Well, that's really helpful, Glen.

This is a help forum--a do-it-yourself forum--not an instill-fear forum. There are other forums for those other purposes. Anyone who reads here knows already that one can screw up big-time, and there are folks here who can encourage and help. Fostering paranoia is not welcome. Informative cautionary instruction is.It is still a valid point, Caution, the wrong adjustment in service menus can damage a TV. Being well informed is the goal, everyone needs to understand the risk. There are those who, no matter how much information they have, should avoid DIY. Not everyone is a DIY'er, no matter how hard they might try. However, like in the stock-market, if you can afford the risk, go for it.

shaddix

12-19-06, 11:07 PM

I have an 36" hs420

Dsperber I have used what you have written earlier in the thread to adjust my set, however I have also read in this thread in a post by kentech to not adjust the settings in MID1-3 because it screws something up. I would link to the post but i can't post links since I have fewer than 5 posts.

Using only the geometry the 2170D-1 or w/e the groups are that have the geometry adjustments i am unable to get the overscan where I want it, I find the image is too large and the only way to resize it is with mid1-3

So I suppose my post is more directed towards kentech in how do you go about adjusting overscan without modifying the mid3 values?

DSperber

12-20-06, 09:56 AM

I have an 36" hs420

Dsperber I have used what you have written earlier in the thread to adjust my set, however I have also read in this thread in a post by kentech to not adjust the settings in MID1-3 because it screws something up. So I suppose my post is more directed towards kentech in how do you go about adjusting overscan without modifying the mid3 values?I don't recall anything being posted previously by Kentech or anyone else regarding not using MID3 as part of the overall 2-part adjustment process for geometry and overscan for fear of "screwing something up".

And that seems impossible to me anyway, since it is definitely a 2-part process... the 2170D sections for adjustment of the vertical and horizontal components of the background raster (similar to the "canvas" in Photoshop), and then the MID3 adjustments (for 720p/1080i as I recall) for the vertical and horizontal components of the foreground image (similar to the "image" or "layers" in Photoshop) on top of the background raster.

If your HS420 has a working MID3, you should utilize it to complete the whole geometry/overscan adjustment sequence. If the HS420 works differently and it's actually MID1 that you need to tweak (I have no firsthand experience with the HS420 in any size), well then use that group. But on the XBR960 use of MID3 is definitely part of the scenario.

Nitewatchman

12-20-06, 10:32 AM

I don't recall anything being posted previously by Kentech or anyone else regarding not using MID3 as part of the overall 2-part adjustment process for geometry and overscan for fear of "screwing something up".

I suspect what he is concerned about concerning the MID3 adjustments involves Ken's observation that adjustment of the MID1-MID3 Vertial Size (such as MID3 VDVS) controls from the factory(or service code chart) values reduces resolution/image quality.

Ken posted some of his observations on this here :

Post # 1730 - Paragraph #3~5 :

http://www.avsforum.com/avs-vb/showthread.php?p=7909446&&#post7909446

Post # 1823 :

http://www.avsforum.com/avs-vb/showthread.php?p=8036820&highlight=VDVS#post8036820

justinblair2

12-20-06, 04:14 PM

Hey Guys i have the xbr 970 and im gettin some issues with vertical and horizontal placement through the xbox 360. One example is Ghost Recon, there is suppose to be a HUD all around the screen and its almost entirely cut off vertically and horizontally. I've had the vertical adjustment where i wanted it but when watching videos on my 360 there is like a blurring and strechting of the picture toward the bottom, so i set the values back to default. I guess really what im asking here is where could i find the values to set the vertical and horizontal positioning of the screen on a per input basis. I've read through DSperber spreadsheet, but i just cant seem to dial in what im looking for. I feel like i missing something when im gaming because of this loss of picture. Any help would be greatly appreciated.

Thanks a bunch

doveman

12-20-06, 11:23 PM

I suspect what he is concerned about concerning the MID3 adjustments involves Ken's observation that adjustment of the MID1-MID3 Vertial Size (such as MID3 VDVS) controls from the factory(or service code chart) values reduces resolution/image quality.

Ken posted some of his observations on this here :

Post # 1730 - Paragraph #3~5 :

http://www.avsforum.com/avs-vb/showthread.php?p=7909446&&#post7909446
Actually, in that post Ken states that we should set the horizontal, as well as the vertical, size settings in MID1-3 to their defaults. He then goes on to suggest doing the same for almost all the other MID1-3 settings.

This also confused me and I too asked for clarification about it recently to no avail, so I'm still none the wiser.

GlenC

12-21-06, 12:56 AM

Ken can correct me on this (I haven’t been into a XBR960 for a while), but as I remember horizontal and vertical size adjustments:
H/V size adjusted in 2170D-1/2 is for inputs 1-5, 480i Wide Zoom (Base size adjustment, global starting point).

MID2-1,3 - corrects 480i (normal & other modes) for inputs 5/6, YC (S-video), ATSC/HDMI and 480i expand. If I remember correctly, the phase will center the normal 480i image within the 16:9.

MID3-1,3 - corrects for 408p (normal & other modes), 720p and 1080i in the component, HDMI, ATSC inputs.

doveman

12-21-06, 06:49 AM

Could someone point me in the direction of a service manual for the XBR800 please.

DSperber

12-21-06, 11:03 AM

H/V size adjusted in 2170D-1/2 is for inputs 1-5, 480i Wide Zoom (Base size adjustment, global starting point).

MID2-1,3 - corrects 480i (normal & other modes) for inputs 5/6, YC (S-video), ATSC/HDMI and 480i expand. If I remember correctly, the phase will center the normal 480i image within the 16:9.

MID3-1,3 - corrects for 480p (normal & other modes), 720p and 1080i in the component, HDMI, ATSC inputs.I'm still unclear as to what the "fear factor" is relating to adjusting MID3.

Starting from either service manual default values or whatever the set is currently sitting at (i.e. from the factory adjusted settings or previous in-home adjustments by you or a Sony technician) small tweaks (i.e. one click at a time) to just four relevant items (VDHP, VDHS, VDVE, VDVS) are seemingly absolutely necessary in order to get the foreground image sized and positioned correctly. And of course the prerequisite steps are adjustments to the relevant background raster size and position using 2170D-1 and 2170D-2, with the MID3 tweaks complementing things.

In my own case, I did initial adjustments using DVE (480p on INPUT5). Subsequently when I had the DVI second head of my PC's ATI Radeon video card connected (1080i on INPUT6, through ATI's DVI-to-component adapter) I did further adjustments using the overscan test pattern from DMWVE (as previously described, and test pattern posted) presented by ACDSee to Monitor 2 (which was the XBR960). I did not do adjustments while selected to any other input than INPUT5 and INPUT6.

As can be seen in my XLS spreadsheet, the current settings for the above four items as shown in the C-1080i column are way different from the "default" values (shown in parentheses), clearly reflecting my own personal adjustments [which obviously are dependent on my own personal corresponding 2170D-1 and 2170D-2 values], and yet I have no harm but rather what I feel to be a perfectly acceptable picture:

VDHP 0-255 80 (107)
VDHS 0-255 252 (240)
VDVE 0-63 9 (19)
VDVS 0-255 140 (135)

It's not clear why my values for 480p (INPUT5) in particular, as well as all other resolutions and inputs do not reflect these same adjusted values if MID3 is supposed to apply to them all, but rather appear to be the same as the service manual defaults. I'd like to get to the bottom of that mystery.

I also have a 1080i D-VHS version of DVE that I've never used, but over the holidays I'm going to do some experimenting with it. Since my JVC DT100U can feed the XBR960 using component, HDMI, and firewire, and can be set to force 720p output or put out native 1080i, I'm going to see how the DVE overscan test pattern looks at other than the 480p/INPUT5 I have used it with. I can also set up my ATI Radeon to put out 720p instead of 1080i, so I can also try my DMWVE overscan test pattern (via ACDSee and Monitor 2) at 720p/INPUT6 and see how that looks.

Hopefully this will give me some understanding as to why only my four crucial C-1080i values in MID3 appear to be non-default.

But the bottom line here is that I don't see how one can't also touch these specific four items in MID3 if you're going to adjust 2170D-1 and 2170D-2 as part of a general geometry/overscan adjustment project.

DSperber

12-21-06, 11:09 AM

Could someone point me in the direction of a service manual for the XBR800 please.I don't believe service manuals are available online from Sony. This isn't the User Guide or other public domain document.

I bought my service manual for the XBR960 for about $38 plus shipping by calling Sony Parts: 1-800-488-SONY (7669). At least I believe that was the number I called.

You'll likely have to buy a service manual for the XBR800 (if still available) direct from Sony, as I did.

G-Bull

12-21-06, 11:27 AM

I'm still unclear as to what the "fear factor" is relating to adjusting MID3.

I have a 34XBR970. I can attest to the problem with adjusting vertical size in MID3. It's been a month since I last messed with it, and I didn't take photos, but next time I get around to lugging my PC upstairs to the TV and messing around in the SM, I'll take some photos so you all can see what the problem is.

If you adjust the vertical size in MID3 (at least on a XBR970, since that's the only TV I have and that's the only one I can say I've seen), it seems to actually ADD extra lines within the image that it draws on the screen, to make the overall image taller. So let's say (and these numbers I'm using are purely as an example) you've got an image that's ten pixels tall. You use MID3 to stretch it vertically so it's now eleven pixels tall. The TV does this by duplicating one of the lines in the image. It can't stretch ten rows into eleven, it can only add another row to the ten rows, by duplicating one of the rows that's already there.

I've seen it on a test pattern with lots of horizontal lines. Again, I'll take some photos next time I've got my PC hooked up and I'll post them here. But this is essentially how a test pattern with horizontal lines appears:

Before vertical size adjustment in MID3:
---------line---------------------------
---------line---------------------------
---------line---------------------------
---------line---------------------------
---------line---------------------------
---------line---------------------------
---------line---------------------------
---------line---------------------------
---------line---------------------------
---------line---------------------------
---------line---------------------------
---------line---------------------------
---------line---------------------------
---------line---------------------------

After vertical size adjustment in MID3:
---------line---------------------------
---------line---------------------------
---------line---------------------------
======brighter line=====================
---------line---------------------------
---------line---------------------------
---------line---------------------------
======brighter line=====================
---------line---------------------------
---------line---------------------------
---------line---------------------------
======brighter line=====================
---------line---------------------------
---------line---------------------------

<< EDIT: it's actually MID2. Everything I said was in "MID3" in this post, I meant "MID2." The setting is DHVS in MID2 on my 34XBR970. I don't know what I was smoking earlier. I meant MID2. Sorry for the confusion >>

ptchristensen

12-21-06, 11:39 AM

I'm still unclear as to what the "fear factor" is relating to adjusting MID3.But the bottom line here is that I don't see how one can't also touch these specific four items in MID3 if you're going to adjust 2170D-1 and 2170D-2 as part of a general geometry/overscan adjustment project.

You are right, it's impossible to center and adjust overscan without the MID2 and MID3 settings. The danger that KenTech has pointed out is that the MID2 and MID3 settings stretches the picture and thereby the scan lines. In the 2170D settings the signal and scan lines are moved as a whole - in the MID settings the signal is anchored in the upper and left side and stretched from the left and from the top.

If you have to do extensive changes in the MID settings you can very easily stretch the picture and the scan lines and thereby get an uneven flow of scan lines acreoss the screen.

When you do use the MID settings, which you will have to do, make sure that it's minimal. And always use a crosshatch signal to make sure that the squares are evenly sized and spaced troughout the screen, and none of the lines are brighter or darker than others.

Nitewatchman

12-21-06, 12:03 PM

Actually, in that post Ken states that we should set the horizontal, as well as the vertical, size settings in MID1-3 to their defaults. He then goes on to suggest doing the same for almost all the other MID1-3 settings.

This also confused me and I too asked for clarification about it recently to no avail, so I'm still none the wiser.

Sorry, perhaps someone else can do better job of it, but at this point I don't quite understand why additional clarifcation is required, and I don't know how it could be more clearly stated than Ken stated(as quoted below) in his post # 1730, paragraph #3 :

In the MID1-3 groups, be sure to get the values for width and especially height exactly right, as these values control (as I have discovered) the precise sampling of the video for display. Get the vertical ones wrong, and the image quality will be compromised! There are exact ratios between the particular video-line count and the internal digital representation of that video that have to be maintained.

Now, I will say I don't recall if anyone has observed whether or not the horizontal size adjustments in MID1~3 will compromise resolution/image quality, however it seems to me to be a safe bet it probably would if vertical size adjustments do.

Also, It's not that "stretching" the picture is the issue[update: although I admit, I suppose that is "one way" of thinking of it which isn' necessarily wrong], as I understand it the issue with adjustment from the proper values as Ken explains it would likely result in reduction of resolution.

However, also, obviously, adjusting horizontal size contols in Mid1~3 and not Vertical size controls(without other adjustments being made elsewhere - such as 2170D1 ASPT or VSIZ) would result in improper aspect ratio occuring.

I don't believe anyone has observed any such issues so far with the MID1~3 posistioning controls, however, nevertheless where this issue is concerned I think it might be wise to also use those sparingly, and only if absolutely necessary for proper centering with a specific scan rate/input device/etc.

I'm still unclear as to what the "fear factor" is relating to adjusting MID3.

It's not as if you're "not allowed" to adjust these settings, I don't think anyone is saying it is going to "damage your set" if you do so ... It's up to *you* if you want to compromise resolution/image quality in favor of a little less overscan/etc ... Isn't it better to be aware of the issue as Ken has observed than not?

GlenC

12-21-06, 12:21 PM

...........In my own case, I did initial adjustments using DVE (480p on INPUT5). Subsequently when I had the DVI second head of my PC's ATI Radeon video card connected (1080i on INPUT6, through ATI's DVI-to-component adapter) I did further adjustments using the overscan test pattern from DMWVE (as previously described, and test pattern posted) presented by ACDSee to Monitor 2 (which was the XBR960). I did not do adjustments while selected to any other input than INPUT5 and INPUT6.Since adjustments for 1080i, 720p and 480p are input specific (not so much with 480p) you skipped a step by not doing the initial adjustments at 480i. There is also the consideration of what inputs are to be used. You may get a couple inputs looking great, but if you go to another input or scan-rate, there may be issues. There is definitely a hierarchy in procedures when doing a full setup, preferable to severely tweaking one input/scan-rate.

G-Bull, your bright line example shows the results of the image being squeezed beyond the resolution of the screen/raster where image lines are blending together.

Nitewatchman

12-21-06, 01:02 PM

But the bottom line here is that I don't see how one can't also touch these specific four items in MID3 if you're going to adjust 2170D-1 and 2170D-2 as part of a general geometry/overscan adjustment project.

You are right, it's impossible to center and adjust overscan without the MID2 and MID3 settings.

Perhaps I'm missing something, but I don't understand this, other than perhaps in the context of GlenC's last post. For example for me/my set, It was only necessary to adjust 2170D1/2 controls (and occasional use during adjustments of the blanking shutters in 2170D3) to have :

#1). approx 4% overscan(and I can go to about 3% without any issues involving screen centering or picture size- i.e. any "black" showing up on the edges so to speak) --

#2). Proper aspect ratio,

#3). Very close to proper centering for all used scan rates/inputs - - 480i, 480p, 720p, 1080i, component inputs/etc.

#4). #1-~3 for all screen "modes" - full/Zoom/etc. (of course, Widezoom is a nonlinear "stretch mode", so that's not exactly quite "proper" aspect ratio). --

All WITHOUT the need to adjust any MID settings ...

Although, I would like an *independant* adjustment of horizontal/vertical size for each of the HD ZOOM, Zoom and "normal" screen "modes".(for KD34XBR960) -- Most importantly and signficantly for HD ZOOM, but although I've looked for it I can't seem to find those, including in the MID controls.

The real, main issue I have with reducing overscan further isn't one which would be improved by adjustments to the MID settings, instead, it involves such issues as the curving of vertial lines at the edges I can't fully compensate for via the various geometry settings, and "artifacts" on the edge of the raster.

However, If I want exactly "perfect" picture centering for all used inputs+devices/scan rates, I would need to make the following slight adjustments to the MID posistining controls(default values on my set in parentheses) :

MID2:

Centering changes :

DRHP (DRC Horizontal posistion) :

480i V5/V6

"Normal" - 155 (153)
"Others" - 119 (120)

480i ATSC :

"Normal" - 154 (145)
"Others" - 117 (108)

CV/YC AND RF :

"Normal" - 151 (154)
"Others" - 115 (117)

DRVP (DRC Vertical posistion) :

480i V5/V6

"Normal" - 35 (37)
"Others" - 35 (37)

480i ATSC :

"Normal" - 35 (37)
"Others" - 35 (37)

CV/YC AND RF :

"Normal" - 36 (37)
"Others" - 36 (37)

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

MID3 :

VDHP (Video display Horizontal Posistion) :

480p V5/V6 :

"Normal" - 199 (200)
"Others" - 151 (152)

480p ATSC :

"Normal" (left at default may need slight adjustment)
"Others: - 153 (152)

720p ATSC - 119 (117)

1080i ATSC/MS - 87 (85)

G-Bull

12-21-06, 01:06 PM

G-Bull, your bright line example shows the results of the image being squeezed beyond the resolution of the screen/raster where image lines are blending together.
Thanks for the explanation. Alls I knows is when I saw the effect, it just screamed to me "keep these at the factory settings..."

Which sucked a bit because I had used those MID3 settings to trial-and-error my way to a beautiful 3% overscan on all my resolutions (at least through HDMI, which is how I have my HD DVR cable box connected). So I had to pretty much start over without using the MID3 vertical size... c'est la vie, I guess.

But I did make some changes to MID3 horizontal settings. I didn't notice an effect like I did with the vertical settings, but does anybody know for sure whether the MID3 horizontal settings has the same negative effect?

Nitewatchman

12-21-06, 01:21 PM

But I did make some changes to MID3 horizontal settings. I didn't notice an effect like I did with the vertical settings, but does anybody know for sure whether the MID3 horizontal settings has the same negative effect?

G-Bull, first off thank you for posting your observations/findings on this ...

I'm not sure whether or not anyone has been able to observe whether or not the effect is present for adjustment of the MID3 horizontal size controls, but I would speculate that may be difficult to achieve with any certianity from an "observation" standpoint using test patterns/resolution wedges/etc ....

As I noted in an earlier post today, however, it seems to me it may seem logical to assume it may involve the same negative effect, and therefore perhaps it may be wise to avoid changing those if possible ..... But, that's just my opinion of course ...

DSperber

12-21-06, 01:23 PM

It's not as if you're "not allowed" to adjust these settings, I don't think anyone is saying it is going to "damage your set" if you do so ... It's up to *you* if you want to compromise resolution/image quality in favor of a little less overscan/etc ... Isn't it better to be aware of the issue as Ken has observed than not?Overall I certainly agree. But much of what is being "warned" is also very much common sense and visually apparent once you begin to adjust things (one click at a time is clearly the right approach, so you can see very clearly what you are doing to the picture).

There's absolutely no question that adjustments of the image in the horizontal and vertical dimensions open the door to screwing up aspect ratio if you are sloppy, which will make people's heads squat and fat, or tall and thin when live content is displayed. But this should really be quite obvious even when you're doing it with test patterns on the screen.

Also, if you "reveal" too much visual image beyond the scan line resolution something's got to give... and it will obviously be picture quality. No question this whole thing is a delicate process, best addressed one click at a time in any dimension for any item being tweaked. And it's highly likely you will revisit previously adjusted items until you get the "balance" of all of them together just right (at least by your eyes).

Of course that's why using a stationary and appropriate cross-hatched overscan test pattern is also the proper way to go about these tweaks. That way you can attempt to accomplish a uniform amount of overscan at all four edges, thus keeping the 16x9 aspect ratio correct (for 16x9 source) as you make your small adjustments while at the same time not exposing anything beyond the absolute outer edges of the test pattern image. I submit that when you succeed in these adjustments people's heads and bodies will look "human" and in perfect proportion when you actually watch real world content instead of a test pattern, and you will also see uniformly a great deal more real estate around the entire periphery of a scene (e.g. to the right and below the channel logo in the the lower right corner of most shows now) than was likely the case before beginning your adjustments.

Applying common sense will stand you in good stead. A bad picture when you're done says you've done something wrong.

Nitewatchman

12-21-06, 02:15 PM

Of course that's why using a stationary and appropriate cross-hatched overscan test pattern is also the proper way to go about these tweaks.

Of course it is. A cross-hatch pattern with Circles is also a good tool to use for the purposes you mention, as well as for checking/adjustment for linearity issues.
[/QUOTE]

I submit that when you succeed in these adjustments people's heads and bodies will look "human" and in perfect proportion when you actually watch real world content instead of a test pattern, and you will also see uniformly a great deal more real estate around the entire periphery of a scene (e.g. to the right and below the channel logo in the the lower right corner of most shows now) than was likely the case before beginning your adjustments.

Well, I certianly have a different view of this issue than you do. Perhaps I should put this another way.

If you use the MID1~3 vertical size controls during your adjustments, per KenTech's findings image quality/resolution will be diminished ..... And, again, I see nothing wrong with informing folks about this issue - In fact, If it were me, and I was posting any info about these adjustments, I would surely mention Ken's findings on this.

If you can "see" such an issue with any various tests or test patterns(such as resolution wedges/etc), certianly the issue will be there when you're viewing "regular" programming, or output from a PC/etc, as well, whether or not it is particuarly "noticable".

If such a reduction of image quality/resolution is "OK" for you merely because you don't necessarily notice a significant enough issue with it "directly" when you are making such adjustments, or anyone else who decides to utilize those controls, fine, but I must say for my purposes, it isn't "OK" for me - including from a "common sense" perspective, which is why, KNOWING this, I would generally not KNOWINGLY advocate their use as I also know these sets are not necessarily designed for best results with such low overscan as say, 1~3%.

Also, I would point out there is no reason whatsoever to need to adjust the MID controls to achieve proper aspect ratio, or even in most cases acceptable overscan(I can go as low as 3-4% without needing to adjust any MID settings for instance), as everything needed is in 2170D1/2.

G-Bull

12-21-06, 02:55 PM

I've been meaning to revisit my overscan/size/center settings, anyway, and Nitewatchman: you've convinced me to see if I can't do it all without touching the size controls in "the MIDs."

Of course I wrote down all my factory settings before doing anything, anyway, so it won't be hard to put everything back.

But, tell me, this self-imposed moratorium on making size adjustments with MID, that doesn't include the MID position settings, does it? Because I HAD to use one of the MID vertical position settings to move the image down (I don't remember which, I think it was in MID3). When I was first making adjustments, I couldn't figure out why I wasn't able to see the top 4% or 5% of the image, even when I used 2170D1&2 size adjustments to make the whole raster small and in the middle of the screen. Turns out the vertical position setting in MID3 was set so that the top part of the image was completely cut off and wasn't even being drawn on the raster! From the factory! I don't know what's up with Sony's quality control, but I think there was some major corner-cutting going on. Don't get me wrong, I do love the PQ of my 34XBR970, but I wouldn't recommend it to anyone who doesn't have the knowledge/ability/time/patience to go into the SM and set it up right.

<< EDIT: The vertical adjustment I mentioned is not in MID3, it's DHVS and it's in MID2 on the 34XBR970 >>

Nitewatchman

12-21-06, 03:22 PM

But, tell me, this self-imposed moratorium on making size adjustments with MID, that doesn't include the MID position settings, does it?

Not that I'm aware of ... In other words, I don't believe anyone so far has discovered any potential loss of resolution issues via adjustment of the MID posistioning settings .. FWIW, In my case, I currently have made mostly slight adjustments to those, and currently have them set as noted in my earlier post, and I haven't noticed any issues(I can't say I've looked extremely closely for any potential losses of image quality/resolution, either, however), as I like "perfect as possible centering<g>" .....

Because I HAD to use one of the MID vertical position settings to move the image down (I don't remember which, I think it was in MID3). When I was first making adjustments, I couldn't figure out why I wasn't able to see the top 4% or 5% of the image, even when I used 2170D1&2 size adjustments to make the whole raster small and in the middle of the screen. Turns out the vertical position setting in MID3 was set so that the top part of the image was completely cut off and wasn't even being drawn on the raster! From the factory! I don't know what's up with Sony's quality control, but I think there was some major corner-cutting going on. Don't get me wrong, I do love the PQ of my 34XBR970, but I wouldn't recommend it to anyone who doesn't have the knowledge/ability/time/patience to go into the SM and set it up right.

Well, perhaps one thing that might be involved with that --- with a NTSC signal, a number of the top lines are used to send information along with the signal, such as Line 21 (601) Closed captioning information which looks like "junk" if you can see to the entire top edge of the "raster" ... ... This info is also sometimes(but not allways) sent via atsc digital when a broadcaster "converts" the analog video to MPEG2 transport stream and sends it as 480i SD, or upconverts their NTSC SD signal to say, 720p and 1080i .... In such cases that "junk" at top of screen is still "embedded" in the ATSC/digital video that results, even though "digital" captioning information is/can also be sent via EIA-708 captions which doesn't embed the info directly into the video signal(which is of course just "bits" in a MPEG2 stream).

I don't recall "exactly" how this looked/works out on my sony, as it's been a while since I've made those adjustments/etc. For example, I think some equipment may automatically "block" that junk at the top of the screen and moves the first line with "no junk" to the top, with no service menu/etc. options available to let you "see what's actually above it" --- In which cases as you might imagine, this issue can be related to how proper screen centering is accomplished for various signal sources as well, especially when it comes to signals originating as NTSC signals with the line 21 captions/etc present ...

However, do Keep in mind, you need to "retract" the "shutters"(which keep the electron beam from striking the sides of the tube - there are both vertical+horizontal "shutters") in 2170D3 in order to be able to see the true edges of the raster. This procedure+info related to it is detailed elsewhere in this thread, and in the service manual as it relates to centering the raster horizontally.

shaddix

12-21-06, 03:24 PM

Hmmmmmmm perhaps you guys don't understand how bad the overscan is with MID values on my hs420 set to default.

If I get the 2170D values set with as little overscan as i find acceptable, and then I go change the MID values to default, The overscan is terrible, at least 15% on the top, I always have to move the picture way way down, Even with MID3 setting the picture as low as possible there is still some overscan at the top.

But also one thing,(I may not know what i'm talking about here). Adjusting the 2170D values changes position of the scan lines on the screen. Assuming I'm correct about this, shouldn't adjusting the MID values to center the image properly on the scan lines be optimal?

With my current line of thinking, the only way i can see adjusting MID would be bad, is if you failed to adjust the 2170D first, and left that with 15% overscan on all sides, then set MID to have 4% overscan, You would have too much image crammed into too small a space.

Please correct me on anything that I am misunderstanding :)

Nitewatchman

12-21-06, 04:02 PM

Sorry for the double post on this ...

When I was first making adjustments, I couldn't figure out why I wasn't able to see the top 4% or 5% of the image, even when I used 2170D1&2 size adjustments to make the whole raster small and in the middle of the screen. Turns out the vertical position setting in MID3 was set so that the top part of the image was completely cut off and wasn't even being drawn on the raster! From the factory! I don't know what's up with Sony's quality control, but I think there was some major corner-cutting going on. Don't get me wrong, I do love the PQ of my 34XBR970, but I wouldn't recommend it to anyone who doesn't have the knowledge/ability/time/patience to go into the SM and set it up right.

Another thing ... I should say I've noticed an issue with this that needed correcting on several models(from different manufactuers) of "flat screen" Direct-View CRT ..

All of those sets (including my Sony) shipped with overscan at a whopping high 8~10% or so !!!!! Why would they do that? Well, I suspect in some cases given the vast array of different video sources out there via "consumer devices", and the probably "quick" calibration these sets receive on the factory floor you'd be surprised at how easy it is to get to the "edge" of the picture with say a SD satellite signal and a screen mode that manuipulates aspect ratio in ways similar to "wide-zoom" ...... They probably don't want customers calling and complaining about wavy "black" edges showing up on the edge of the screen .....

Or, perhaps more importantly, they probably don't want folks calling and complaining that the Newschannel/etc "tickers" are being "cut off" at the bottom of the screen ..... If I recall correctly, generally the "safe zone" for broadcasters/cablenets/etc to follow in this regard is(at the least) 5% overscan, but there is no set "rule" on that, including for manufactuers implementation on their sets ....

G-Bull

12-21-06, 04:23 PM

However, do Keep in mind, you need to "retract" the "shutters"(which keep the electron beam from striking the sides of the tube - there are both vertical+horizontal "shutters") in 2170D3 in order to be able to see the true edges of the raster. This procedure+info related to it is detailed elsewhere in this thread, and in the service manual as it relates to centering the raster horizontally.
I accounted for that when I made the adjustments and disabled those shutters first, but the top of the screen still wouldn't appear when I was using an overscan pattern. It took me quite a while to find the correct setting to adjust, mainly because the MID settings were different in my 34XBR970 from what's been documented here about the 960 and HS420...

I didn't know any of that you mentioned about closed captioning, etc. Good to know. But Sony's method of "sliding the image clear off the screen" by 5% or so seems excessive. They could have just cut it off with the 2170D-3 shutters or something. Shoddy final set-up at the factory, if you ask me. Like I said, I really do like my 970, but it really didn't look good when I pulled it out of the box. I've still got some pretty bad vertical convergence issues in the corners that I need to call a service tech to come out (it's off by at least 1/8" to 1/4" in the corners) so I think that Sony is just trying to make the TVs "good enough for most people" and ship them out the door.

shaddix - That does seem unacceptable. You're saying that the factory MID size settings have the image that much larger than the viewable area, so there's that much overscan? Even when you reduce the size with 2170D-1&2, you still can't see the edges of an overscan pattern? Are you reducing (or turning off) the 2170D-3 shutters?

shaddix

12-21-06, 05:15 PM

shaddix - That does seem unacceptable. You're saying that the factory MID size settings have the image that much larger than the viewable area, so there's that much overscan? Even when you reduce the size with 2170D-1&2, you still can't see the edges of an overscan pattern? Are you reducing (or turning off) the 2170D-3 shutters?

Yes you are correct, with MID values all set to defaults, when I set the 2170D values to show the entire image by setting the height and width to minimum with whatever those hblk and vblk(i forget the exact name) set to 0 instead of 1 you cannot see the entire image, the only way is to *greatly* modify some of the MID numbers

maybe my set is defective ;_;

shaddix

12-21-06, 05:25 PM

when i do the 2170D adjustments, i modify the MID values so that the image is massive, to make sure I can see the edges of the raster(? i don't really know what this means >_<) change 2170D to my liking, record the new numbers, 0>enter to read then change the values and write them

ahhhhhh sorry for the double post i th ought someone posted after me but i was mistaken D: however this post still does have some merit so i'll leave it ^_^

btw thank you guys for the help, it's a nice christmas present :>

Nitewatchman

12-21-06, 07:38 PM

ahhhhhh sorry for the double post i th ought someone posted after me but i was mistaken

No need to apologize, that was my fault ... I posted something not realizing you had made another post, and after reading it(your first "most recent" post), on 2nd thought decided it probably wasn't of much help so I deleted the post ....

Nitewatchman

12-21-06, 08:09 PM

Shaddix,

Ok, I'll try to take a stab at this this way, although keep in mind you will likely be able to find better info on some of this by using the thread "index or "TOC o' links" Ken has provided in First post, or by searching this thread for some key words related to your comments :

Yes you are correct, with MID values all set to defaults, when I set the 2170D values to show the entire image by setting the height and width to minimum with whatever those hblk and vblk(i forget the exact name) set to 0 instead of 1 you cannot see the entire image, the only way is to *greatly* modify some of the MID numbers

2170D1 VSIZ (vertical size) and 2170D2 HSIZ (horizontal size) are the appropriate adjustments for vertical and horizontal size of the image(*and/or the raster*) on my set - including if you want to reduce the size of the image, although you may also need to "retract those 2170D3 shutters" to see absoluetely 100% of the "image".

And, IMO unless there is something vastly different about your model set, even if you *do* decide to use the MID controls to adjust vertical/horizontal size despite the cautions expressed here, it should *still* only be as a *last resort* IMO, and only in a Input/scan rate specific manner if say for instance, the VSIZ/HSIZ adjusment works fine for a 480i signal, but you need something slightly "different" for say, 1080i that you can't "fix" with *any* 2170D1/2 controls ..... Again, that's just My opinion however, FWIW ...

[Updated section follows] :

Also, perhaps I should mention 2170D1 VPOS (vertical posistion) and 2170D2 HPOS (horizontal posistion) are the appropriate, basic image *posistioning* controls, although, at least on my set I do note adjustment of VPOS seems to also slightly effect geometry -- In a manner somewhat similar to 2170D2 PPHA.

2170D2 HCNT is the control used to center the raster Horizontally, there is no equivilent for centering the raster vertically, AFAIK -- It's either "static" vertically or effected by VPOS as well, I don't recall ...

There are also several other controls in 2170D1/D2 which effect vertical size, or vertical posistion, or horizontal size -- Some of these are "geometry/linearity related controls which *just happen* to effect Vertical/Horizontal size as sort of a "by product" of what they do --- Controls such as VLIN, UVLN, LVLN(effects vertical linearity as well as vertical size), or SLIN (effects horizontal size, H Linearity as well as curvature of Vertical lines).

However, one potentially useful control in *some* circumstances that effects Vertical size(and therefore also the "aspect ratio"), but not vertical posistion or geometry -- is 2170D1 ASPT, which is *screen mode specific* (480full, 1080full, Zoom/Widezoom/etc) ... Another effects vertical *posistion* but not vertical size or geometry, also in a *screen mode specific* manner --- that control is "SCRL" ....

Other than the MID controls, I don't think there is a control that allows for input/scan rate or most "screen mode" specific adjustment of horizontal Size or posistion, other than for Widezoom and "others" in the case of HSIZ(horizontal size), and 1080full and "others" in the case of HPOS(horizontal posistion). And, I don't believe any of the MID Hsize/H posistioning controls are "screen mode" specific, only "scan rate" or input specific ...

[end update]

when i do the 2170D adjustments, i modify the MID values so that the image is massive, to make sure I can see the edges of the raster(? i don't really know what this means >_<) change 2170D to my liking, record the new numbers, 0>enter to read then change the values and write them

Hmmm .... Basically, the "raster" is like a canvas if you're painting a picture, and generally(but not allways), the picture you're painting (the image or actual "picture area") will pretty much "fill" the *raster* or the "canvas".

To see the edge of the raster(search this thread for "raster" to find better explanation of what raster is and/or this thread or service manual for how to center it Horizontally or how to "view the edge of it")... in addition to appropriately and temporarily adjusting the 2170P1 AGNG control, and "retracting the shutters" in 2170D3, that would also involve reducing HSIZ (or VSIZ) to see the "edges" of the raster, not adjustment of MID values or making the image "massive" --- in fact, you temporarily need to REDUCE the size of the image(*and the raster*) to "less than" 0% overscan in order to see the edge of the raster .....

*do* keep in mind these sets are really designed to view "TV" programming, with a certian amount of overscan involved, generally 4% or more -- they aren't designed for "0% overscan" in order so you can see every little bit of image on the edges, such as on the sides or top .... Forutnetly, some PC video cards(my ATI Radeon X1600 Pro can do this in a "adjustable" fashion via a function in it's "Catylst control center" software) do have the option of sending an "underscanned" image to a HDTV display so you can set it up to see every bit of "edge" of the image as may be necessary in some cases for gaming or "desktop applications" use .... Although generally speaking, I don't think I'd personally want to use one of these sets as a PC monitor for non-gaming or other than to display graphics on a bigger screen/etc..

If none of that helps .... Another thing I perhaps might want to wonder about/ think about if your having these issues only from output from a PC vs. what your getting from "regular" TV or HD sources or DVD player hooked to the TV, is perhaps you might want to make sure your video card is sending a "properly formatted" signal to the TV, and that the TV is recoginizing it properly .... both aspect ratio wise and say, "screen centering wise", and "resolution format" wise ... For instance, for 1080i HD signal, that would be 1920x1080i @ 30HZ interlaced ...

btw thank you guys for the help, it's a nice christmas present :>

I hope some of it actually ends up helping you out+Merry Christmas!

shaddix

12-21-06, 11:53 PM

thank you for clearing up the input specific/screen mode specific settings, i hadn't understood that before and was just hoping that a change wasn't global to everything when i was making it lol, i have gotten lucky though i think :D

as far as pc output goes, i have a 9800 pro with a dvi to hdmi cable but i'm not even going to get into the problems that show up when i try that whole bit D:

alright well let me see if i can clarify what i'm thinking here and maybe this will do me good.

the raster controls the scan lines, how many different lines of pixels the electron guns are shooting onto the screen correct? and MID controls the position and size of the actual image being sent to the set in relation to the raster.

so lets say the image being sent to the set is a 1080 image, 1920x1080 pixels, the raster is going to be composed of 1920 horizontal pixels and 1080 vertical ones being shot out by the gun, the actual image is also going to be 1920 horizontal pixels and 1080 veritcal ones. Now hopefully what this accomplishes is that for each line that the electron guns are shooting, there's a line of image to go along with it. Am i correct with this?

now assuming I am correct with this, if the MID default values have the image way off center or distorted, then the default values are what are compromising the image quality? and I need to modify them sot hat the image lines up properly with the raster

now this may all be incorrect but i'm hoping it isn't >_<

doveman

12-22-06, 05:47 AM

I don't believe service manuals are available online from Sony. This isn't the User Guide or other public domain document.

I bought my service manual for the XBR960 for about $38 plus shipping by calling Sony Parts: 1-800-488-SONY (7669). At least I believe that was the number I called.

You'll likely have to buy a service manual for the XBR800 (if still available) direct from Sony, as I did.

Thanks for the info. I'll try that if I need to.

I've got a PDF for the KD-34XBR2 (which I think I downloaded from here) and I've found a website with some service manuals (http://electromaniacs.com/) and downloaded one for the DA4 chassis, 510 models. Am I right in thinking the XBR800 is a DA4 chassis model as well? If so, how similar are the circuits likely to be?

Nitewatchman

12-22-06, 03:52 PM

if the MID default values have the image way off center or distorted, then the default values are what are compromising the image quality?

No -- assuming your set has the "correct" MID default values it's supposed to have(which *should* be the case), CHANGING those default values(for vertical size particularly per Ken's findings) is what will compromise image quality - (disclaimer) --- unless perhaps newer models of sets work differently than KD34XBR960 or ken's XS955, although I wouldn't expect that would not be the case).

Therefore, it's best(IMO) to Use 2170D1/D2 controls I mentioned in last post to adjust the size/posistioning of the image(not *just* the raster), not the MID values - I think it is best to Leave the MID values at the defaults.

I think you're reading too much into this thing about "the raster" and "the image" being somehow "seperate" things ... They really aren't "seperate" things except in terms of some adjustments that are *possible* which involve moving the image around "on top" of the raster, without moving the raster itself ... such(If I recal correctly, such as 2170D2 HPOS or the MID3 adjustments), and seperating them in terms of "pixels" or "scan lines" isn't really accurate, I don't think --

Lets see if I can put it this way ... The image is "part" of the raster, but the raster can be LARGER than the image(for a simple explanation the video from a 4x3 signal wouldn't fill the raster entirely on a 16x9 set - It's doesn't necessarily have anything to do with the resolution/"pixels" involved in the source image), just like a picture painted on a artist's canvas could fill the entire canvas, or, there could be a bit of a border around it ....

Again, Hope this helps ....

Nitewatchman

12-22-06, 04:29 PM

the raster controls the scan lines, how many different lines of pixels the electron guns are shooting onto the screen correct? and MID controls the position and size of the actual image being sent to the set in relation to the raster.

so lets say the image being sent to the set is a 1080 image, 1920x1080 pixels, the raster is going to be composed of 1920 horizontal pixels and 1080 vertical ones being shot out by the gun, the actual image is also going to be 1920 horizontal pixels and 1080 veritcal ones. Now hopefully what this accomplishes is that for each line that the electron guns are shooting, there's a line of image to go along with it. Am i correct with this?

If I understand where you are coming from here, CRT's aren't "digital" displays, what results on screen isn't "1920x1080" pixels, that sort of thing is only accurate for fixed pixel displays. As it is a topic beyond the scope of this post, I would suggest if you want to understand how CRT displays work, You'll need to do some researching and reading on the subject - Here may be a good place to start :

http://www.repairfaq.org/sam/crtfaq.htm

But, for instance for the purposes of this post -- dot pitch and, in the case of these sony sets, the apeture grille are important functions of how much "resolution" you get to see "on screen", not how "big" or small/etc. the "Raster" is ...

This may not be entirely accurate as I would think the change in sampling that would occur in the MID processors would perhaps not be a "simple" "Flat" reduction of resolution of the input source signal as my following example illustrates ... but perhaps a good way to look at this might be .... When you adjust the MID Vertical Size to different values from the defaults -- and potentially MID horizontal size controls as well, I'll just assume that's the case for the following --- Per Ken's findings, it is probably *similiar* to if you were "stretching" or "squeezing", say a 1920x1080i source image to some *other* resolution(say, 1600x900 - although you hopefully wouldn't be making that "severe" of an adjustment, and assuming you are perserving proper aspect ratio when adjusting both H+V size. -- That loss of resolution occuring to digital video *before* it even reaches the actual hardware that creates the picture, similar to if you resampled a photo image on your PC with Photo editing software ... Whearas, Adjusting the 2170D1/D2 controls, you aren't *changing* the resolution of the source signal *or* what is displayed via the "scanning lines", in which case the actual resolution being dispalyed is a matter involving dot pitch/etc ....

Oliver Deplace

12-23-06, 12:09 AM

Ideally, you'd want to do all the sizing/positioning with 2170D, but it appears that many cannot get by on 2170D alone.

MID vertical size and position changes can subtly degrade the image. The degradation using MID vertical size can easily be seen with a test pattern. Normal program material isn't so revealing. IOW, if you had one with changes and others without, I doubt you'd be able to pick it out of a line-up. But, if your going to lose sleep over it, don't do it.

The degradation using MID vertical position is so subtle as to make it insignificant.

MID horizontal doesn't seem to cause degradation at all.

So, if the image layer is much larger than the raster, you're pretty much stuck with making some changes in MID.

Pick your poison:
1) A subtle loss of resolution that you may not be able to detect.
2) Lose picture info to cropping.

You gotta do what you gotta do.

DSperber

12-23-06, 12:38 AM

Pick your poison:
1) A subtle loss of resolution that you may not be able to detect.
2) Lose picture info to cropping.A picture is worth a thousand words...

There's nothing to be lost by actually trying some adjustment temporarily by simply tweaking in the service menu and not writing it permanently. Just put it back to what you had (or some other adjusted value) and you're no worse off.

Having nothing to do with any other changes for linearity, convergence, rotation, whatever, why not just see what these do for you when selected to INPUT5/6 while watching 1080i. Ideally a test pattern at 1080i would be ideal, but live content with a crawl at the bottom or scorebar near the top, and logo in the lower right, would also be a reasonable alternative:

VPOS 0-63 28 (26) <-- 2170D-1
VSIZ 0-63 32 (39)

HPOS 0-63 32 (31) <-- 2170D-2
HSIZ 0-63 41 (50)

VDHP 0-255 80 (107) <-- MID3
VDHS 0-255 252 (240)
VDVE 0-63 9 (19)
VDVS 0-255 140 (135)

Just try them and see for yourself. If you don't like it, put back what was there before. If it looks not half-bad, make your own additional adjustments until you're satisfied.

I'm not trying to say my settings are right or wrong, or the only show in town. For all I know they may be uniquely applicable to my particular XBR960 and/or the rest of my customizations (as shown in my spreadsheet), or may not present what someone else wants to see. But certainly experimenting for yourself to see what happens will not do any harm and you can always back out to what you have if you don't like what you see, or just start from where I've put you and finish things up for yourself.

Quite frankly, I do not think I've lost any resolution whatsoever by what I've done. In fact if anything the fact that the overscan has been reduced (by increasing the horizontal and vertical sizes, thereby "shrinking" the effective size of whatever is seen onscreen since a somewhat larger total image area is being presented in the same size 34" screen rectangle) actually makes things much sharper and crisper, and more filmlike in its appearance.