AVS › AVS Forum › Display Devices › Direct View (single tube) CRT Displays › THE SONY SERVICE CODES - Articles, Comments, Discoveries
New Posts  All Forums:Forum Nav:

THE SONY SERVICE CODES - Articles, Comments, Discoveries - Page 64

post #1891 of 2959
Quote:
Originally Posted by darcon_adonis View Post

It is as thick as whatever bright image is being displayed.

To clarify, the streak of darker black is going across the screen horizontally in both directions of say, a small white cube being displayed in the center of a black screen, and the streak is as thick as the hieght of the cube.

It's also worth noting that I can see this even if I don't use a source and have the TV's white interface appear in the upper corner, to rule out faulty cables or sources.

Bumping my issue.
post #1892 of 2959
Thread Starter 
Quote:
Originally Posted by bast525 View Post

About the color settings, I think I saw the same thing you are seeing, as I mentioned above, I had tried 14 instead of 13 for the first setting and found it a little better, but reverted back to 13 for the time being. I will try 14-14 when I get home instead of 13 or 14-15. A great side effect of this change in the SM settings, IMO, was that I was able to adjust the color slider in the regular user adjustments back to 50, whereas before I had turned it down a bit to try to lessen that way-too-overbearing red.

Note that you are wandering without direction unless you have a decent calibration DVD. There are color "pushes" in any number of broadcast programs (try the Food Channel!), so one needs a standard way of calibrating color, both in the decoding and the overall amount.
Quote:


So yeah, I'm having issues with vertical convergence in the upper left corner of the screen.

Some of these "defects," if tolerable, are best left alone if you don't much see them. I have an upper-right vertical convergence problem that's visible on crosshatch patterns but hardly noticable on real programming. Those magnets are a real bear to adjust and they interact with each other: you fix convergence in one area, and it's now screwed up somewhere else!
Quote:


is there a difference in what we are talking about with this linearity, as far as, one side seeming more horizontally stretched than the other one, and how you would describe increased stretching as you get away from center on BOTH sides?

Yeah, that's easy. That one is known as S-linearity. The vertical correction is 2170D-1/#4-VSCO, and the horizintal equivalent (what you mentioned) is 2170D-2/#3-SLIN.

Be sure to write down the original settings before you WRITE new ones!
post #1893 of 2959
Thanks for the additional info KenTech. I agree with you that it's not really worth worrying about the convergence issues in the corner, it's rarely noticeable and not worth trying voodoo magic with magnets to fix, IMO.


I also agree it's not worth really trying to fine tune the colors w/o the DVE DVD.... again I hadn't even planned on TOUCHING anything other than the geometry adjustments (only because I did have a video game disc that has a crosshatch test grid displayed when it first starts up). But after reading how EVERYONE had that 'red push' on their Sony sets and me noticing mine had the same thing, and then reading what you said: "The first try I got 13-15-5-4, and others have reported those settings, too. Big improvement! You could do a lot worse than simply trying those settings", I figured there'd be no harm in just trying those numbers even w/o DVE. Sure enough, the picture looked MUCH better to my eyes after the adjustment. I have no idea how close to 'perfect' it really is, but I'm happy with it.

Only problem is that after more extended viewing it now seems to have just a slight bit too much greenish tinge to yellows/oranges and very light beige/browns. But anyways... yeah definately not going to try to really tweak the colors, and I'm not even going to TOUCH the black levels/gamma etc., w/o the DVE disc. And don't worry I am taking all the warnings to heart, I'm not adjusting anything w/o first writing down the original value.

Now my next issue... I spent a little more time tinkering with convergence yesterday with the crosshatch pattern, and got it to where it looked PERFECT. Except for that one upper left corner, it really looked like the convergence was spot-on horizontally and vertically.

It wasn't. I started notcing as I was fiddling with menus on my DVD player and trying some video games, that with pure white text (with no black border, just plain white text, regardless if background was dark or light), towards the right side of the screen, there was some red 'spill over' to the right of the text. A very bright, almost orangish red. So I pulled the SM back up, and adjusted the CADJ setting a few notches, and the red 'spill over' went away. Only now, anything onscreen that is red colored has spillover! Not quite the same though... it's still red, but it's not the bright white-orangish red, it's a transparent red with maybe the slightest green tinge to it (like moving the tint slider towards the green). Like a red 'shadow'. It spills over in the same direction, to the right. I fiddled and fiddled but could NOT find a setting with CADJ that completely got rid of both kinds of spillover. So I settled on setting it 'right in the middle'.

The funnier thing is it's happening worse on one side... but only partially! What I mean is... okay if I adjust it so that the white text has the bright orange-red spillover, then THAT will only be noticeable on the right side of the screen. But if I adjust it so that there's the faint red "shadow" spillover but the whites are 'clean', then that will be noticeable on any red object anywhere on the screen.

So I tried adjusting CADJ to where there were no red 'shadows', but I had the spillover on the white text. Then I adjusted the D-Conv settings that only affect the right side and corners of the screen, to get rid of the white spillover... but same thing. I can not get rid of the one w/o increasing the other. So it seems like, at least on the right side of the screen (maybe 1/4's worth from the edge), I have to settle for some bad horizontal convergence in one form or another. Or do I? Is there something I'm missing?


EDIT: one thing I forgot about, before I adjusted the CADJ settings, when I had the bright red spillover to the right on white text on the right side of the screen, I also had dim blue spillover to the left of white text on the left side of the screen. After tweaking CADJ BOTH problems cleared up, but again... then I had the red shadow to the right of red objects...

Just thought I'd mention that now that I remembered.
post #1894 of 2959
Quote:
Originally Posted by KenTech View Post

HSIZ, as RWetmore has said. But I thought I'd mention: Where there are some vertical scaling adjustments for different video modes, the horizintal one is global. For example, there is no way to adjust one of the Zoom modes in width without disturbing the normal width setting.

I wanted to do this so 4:3 broadcast on HD could be proportioned to fit my big 4:3 screen if I choose the Zoom mode. But I regret that this seems not possible. (Love to be proved wrong!) My set doesn't have a WideZoom mode, so if there is are specific settings for that, I can't confirm it.

Sets (like 30HS420) with Widezoom have the following settings in the SM,
that are specific to the 'Widezoom' screen mode:

2170-D1:
VLIN
VSCO
ZOOM
APSW
ASPT
SCRL
UVLN
LVLN

2170-D2:
HSIZ
HLIN
MPIN
PIN
PINO
UCP
LCP
PPHA

HSIZ being one, I assume there *is* a way to optomize this screen mode to proper overscan, etc.? Without affecting the other screen modes?

Does anyone have a tried-and-true process?

--
corlay
post #1895 of 2959
Thread Starter 
Quote:
Originally Posted by bast525 View Post

Only now, anything onscreen that is red colored has spillover! Not quite the same though... it's still red, but it's not the bright white-orangish red, it's a transparent red with maybe the slightest green tinge to it (like moving the tint slider towards the green). Like a red 'shadow'. It spills over in the same direction, to the right. I fiddled and fiddled but could NOT find a setting with CADJ that completely got rid of both kinds of spillover. So I settled on setting it 'right in the middle'.

This is not a good strategy because the "spillover" of red from red-colored objects has nothing to do with convergence! It is a consequence of red-color encoding for much of NTSC programming, more with broadcast, less with DVD, almost none for component video. Red has the lowest resolution of all colors, generally, and a bit of color "outside the lines" is par for the course. It is extreme on some SD broadcast TV, however. This is NOT true of digital TV.

You should evaluate convergence on standard material with your Color slider set all the way to zero so none of the other color defects distract you. All convergence problems can be seen with no color in the video, hence the white/gray crosshatch patters' usefulness.
post #1896 of 2959
Thread Starter 
Quote:
Originally Posted by corlay View Post

HSIZ being one, I assume there *is* a way to optomize this screen mode to proper overscan, etc.? Without affecting the other screen modes?

Does anyone have a tried-and-true process?

I personally don't for WideZoom, but surely this is an easy experiment. (That's how I determined how hopeless my custom-zoom width attempt was on my 4:3 set!)

Suggestion: Write down your existing settings for HSIZ for WideZoom and Others. Put the TV into WideZoom mode, and change HSIZ's value significantly. Then switch back to a normal mode to see if horizontal size is the same visually, or if it too has changed. This will establish empirically the independence (or not) of the two modes' settings.

You may have to WRITE the settings for them to stick thru a mode change, but you have the former settings written down as backup, so don't worry.

ASPC and SCRL are the most useful parameters for tweaking the vertical size and position for several scan/zoom modes. Note that any parameter with a range listed as "0, 1" is only a switch that controls whether a feature or function is on or off and has no fine control.
post #1897 of 2959
KenTech - thank you AGAIN for some great info and clarification... you probably just saved me quite a bit of time that would have been spent on hopelessly fiddling with adjustments when I got home from work tonite Okay so the red spillover can't really be fixed in the way I was thinking, except for to upgrade to better video cables (my 'reference' was a video game being played on an Xbox thru S-video cables).

I MAY just reset all of the D-CONV settings back to factory and start over with the crosshatch pattern again, since I have a feeling I have 'tweaked' it badly since I was using the reference that I was, and that it probably really was perfect where I had had it.


KenTech - another question for you, what are your thoughts on the "THX Optimizer" which is a feature that is being included on a lot of DVD's these days. I noticed it first when I bought the DVD set of the original Star Wars trilogy (eps 4-6 I mean). I know it only covers the basics (color, tint, bright and contrast, and some basic geometry I think), but I was wondering if this would suit me well for just adjusting the color? It mentions on the THX website that you can either a) use a pair of their blue filter glasses or b) turn off the green and red guns on the set to make the color and tint adjustments, and I believe I CAN turn off the red and green guns on this XBR970 so I wouldn't even need to wait for a pair of the glasses to arrive in the mail.

Do you think the results would be good with this? I do want to get the DVE or Avia (leaning towards Avia since it SEEMS it would be more user friendly for a 'newbie' like me, BUT it's like 2-3x the price!), and will probably order one of them by the end of this weekend. But if I can use the THX optimizer to fix my color, tint, picture and bright, and my crosshatch for geometry and convergence, that's almost everything except grayscale and color temp taken care of.

I won't bother with the Optimizer though if you don't think the results would be 'true' or worth the trouble.


EDIT: correction, apparently the Optimizer feature does cover grayscale, convergence, and possibly even sharpness adjustments (newer versions, not sure if my Star Wars one has this).
post #1898 of 2959
Ken, I was adjusting the overscan on my KV-HS420, when I left the room briefly, and a guest (who knows better) had a mental lapse, AND TRIED TO CHANGE THE CHANNEL. Um sweetheart, there aren't any channels, just a DVD player.

It took three days of googling to find your forum, but you've saved me. Thank you.

I've got eyes good enough to identify the filmstock that was used, but not the competence to change anything other than geometry. Any thoughts on where I might start learning?
post #1899 of 2959
I am trying to get my overscan under 5 percent. However on the bottom, when I go below 5, the image gets distorted...

||||||||
/////////

Is there a way to straighten this out?
post #1900 of 2959
I just picked up the last XBR960 in Virginia at CC...unfortunately, an open box one.
This one has some fairly significant convergence issues, and also what appears to be a beam focus issue.
I'm new to the service menu and haven't gone in there yet. From reading here, it looks like the convergence issues might be fairly easy to address in the menu, but what about focus?
And what is the exact procedure for getting into and navigating the service menu? I guess most know the basics of that here, but it's all new to me. I'm very familiar with broadcast monitors and the theories of convergence and grayscale and all that, just not familiar with these Sony service menu adjustments.
thanks
post #1901 of 2959
[quote=RM23J8G]
And what is the exact procedure for getting into and navigating the service menu?

If you bounce back to page 1 of the forum you can find pdf's 'n zips for download. Grab em all. There's something good in every one.
post #1902 of 2959
Thanks....I just found them!
Great stuff here.....thanks Ken.
post #1903 of 2959
Quote:
Originally Posted by RM23J8G View Post

I just picked up the last XBR960 in Virginia at CC...unfortunately, an open box one.
This one has some fairly significant convergence issues, and also what appears to be a beam focus issue.
I'm new to the service menu and haven't gone in there yet. From reading here, it looks like the convergence issues might be fairly easy to address in the menu, but what about focus?
And what is the exact procedure for getting into and navigating the service menu? I guess most know the basics of that here, but it's all new to me. I'm very familiar with broadcast monitors and the theories of convergence and grayscale and all that, just not familiar with these Sony service menu adjustments.
thanks

YOu can try the focus adjustments in the service menu, especially overall focus (I believe QDPC), but the adjustments can only do so much. I just recently got a new 960, and and really lucked out with the focus - it's really excellent even at the edges. Prior to this, I owned the 30XS955, and the focus was not as good even after significant adjustments. This suggests to me that the how the focus is calibrated at manufacturing may be the most critical, so I would call out a service tech if the SM adjustments don't yield significant improvment.
post #1904 of 2959
Hey all. Okay, went home last night and tried using the THX Optimizer feature that is on the Star Wars DVD set I have. While pretty basic, it does cover the most obvious/critical bases pretty well, IMO. With test patterns for color, tint, bright, contrast, sharpness, convergence, grayscale, and geometry, I was able to go in and fine tune what I had already done and get the picture even better.

I found that my convergence adjustments weren't nearly as good as I had originally thought, and REALLY got down and tweaked them to near-perfection this time, much better than I had them, and the red bleeding issue is almost 100% gone now.

At any rate, for anyone who has one of the movies that comes with the THX Optimizer feature, I feel it's definately a great way to do a 'quick' tune up with only a small handful of the service menu settings and the regular user settings, and my picture now is MUCH and EASILY NOTICEABLY better than it was 'out of the box'. A good 'free' tool to use for anyone who has one of these movies, though I do plan on still buying Avia at some point for final fine tuning to get the colors 'perfect'.

The only issue I have left that is worth complaining about, and I'm hoping someone knows a fix for in the SM:

The bottom right corner of the screen has a gradually warmer color temperature. I got most of the screen to display, what seems to me to be a really pure white, but the bottom right corner still has a reddish tinge to it. I tried playing with a few settings in the color section of the SM and also with the Landing section, but found NOTHING that affected this.
post #1905 of 2959
I intend to slog my way through the whole forum from page one onwards but in the meantime, some keywords for searching would be a big help.

The only input my KV-34HS420 will ever get is widescreen DVD through the HDMI interface.

Changing HPOS confirms that I'm losing a significant amount of the picture at either end of the screen, and HSIZ only changes screen real estate, it doesn't give me any more of width available from the DVD output signal.

What should I be running searches for? Overscan and blanking were unproductive.
post #1906 of 2959
Quote:
Originally Posted by gribble View Post


Changing HPOS confirms that I'm losing a significant amount of the picture at either end of the screen, and HSIZ only changes screen real estate, it doesn't give me any more of width available from the DVD output signal.


I don't know how relevant this will be or helpful if at all but I thought it worth mentioning. On my XBR 970 (widescreen HDTV), I noticed when I was using a test patter from the THX Optimizer (similiar to Avia etc.), with a rectangle frame to use for referencing how much overscan there was... If I adjusted HSIZ in full screen (16:9) mode, it actually changed the amount of the overscan and how much of the test pattern was being shown on the screen. But then I switched to 4:3 mode to do the same calibration, and I got the same results you are describing... instead of changing the area of the test pattern being displayed on the screen, HSIZ in this mode only affected the amount of the screen the 4:3 box took up. In other words, I COULD NOT adjust the amount of overscan or how much of the test pattern was actually being displayed while in 4:3 mode. I don't really use 4:3 much anymore so didn't look to see if maybe there was another section of settings with an alternate HSIZ or similiar that would correct this.
post #1907 of 2959
Quote:
Originally Posted by bast525 View Post

I don't know how relevant this will be or helpful if at all but I thought it worth mentioning.

Well, it adds to the mystery. My problem is in 16:9 full screen, HDMI input, where your HSIZ works fine. Sale of XBR tubes had been discontinued in Canada at the time I bought the set.

Dunno about 4:3. I don't go there either. The set has some blanking issues, I can only hope they didn't lock the hardware lest they become apparent. Been digging through the forum with no luck so far. Anybody have any ideas?

(At least I've figured out how to configure response posts correctly
post #1908 of 2959
[quote=bast525]Does anyone know which setting, if any, can be used to fix this problem:

There is a slight uneven stretching of the image horizontally. I actually sat there with a measuring tape and measured a line of text on the left and right column (exact same text). I found that the text on the left hand side measured about 7", and the text on the right only 6.5". /QUOTE]
HCNT, right before HPOS and HSIZ changes position of the horizontal and alters the proportion of the two areas on either side. Try adjusting that a little to the left, then re-centering with HPOS.

I'm ready to bow to the expertise of the forum, but I didn't see a solution posted and it worked for me
post #1909 of 2959
Quote:
Originally Posted by gribble View Post

HCNT, right before HPOS and HSIZ changes position of the horizontal and alters the proportion of the two areas on either side. Try adjusting that a little to the left, then re-centering with HPOS.

Sorry, meant to say changes position of horizontal CENTER, The whole image shifts, but the adjustment is really for proportion between the two sides. Ten thumbs and blind today
post #1910 of 2959
I'm getting more familiar with the service menu now, and downloaded some of Ken's test signals onto cd to play in my dvd player.
They really don't look all that good through that process...lot's of artifacts... and I'd like to try the memory stick method. I got better looking test signals by recording them from an NTSC test signal generator into a dvd recorder for some reason.
I have zero experience with memory sticks, as I don't have anything that uses them. I see that there are several different varieties and sizes of them. What do I need?
Also, what sort of hardware/software interface do I need to load the test images into them from my computer?
This is pretty cool stuff....especially if I can get this 960 straightened out!
thanks
post #1911 of 2959
Quote:
Originally Posted by RM23J8G View Post

I'm getting more familiar with the service menu now, and downloaded some of Ken's test signals onto cd to play in my dvd player.
They really don't look all that good through that process...lot's of artifacts... and I'd like to try the memory stick method. I got better looking test signals by recording them from an NTSC test signal generator into a dvd recorder for some reason.
I have zero experience with memory sticks, as I don't have anything that uses them. I see that there are several different varieties and sizes of them. What do I need?
Also, what sort of hardware/software interface do I need to load the test images into them from my computer?
This is pretty cool stuff....especially if I can get this 960 straightened out!
thanks

Why not use the internal test patterns..? Read #416, #420 and #428
post #1912 of 2959
Can anybody recommend service codes which might correct horizontal distortion in a 27" Wega flat-CRT? (27FS170)

The effect I see is that objects appear to bulge and ripple as they pass through the middle of the screen.

I've tinkered extensively with PAMP and UPIN/LPIN, but those don't seem to influence this region of the screen. It seems to be slight concavity along a vertical line in the middle of the screen, rather than a fisheye.

i.e.,
|||||||||||||||||)(|||||||||||||||||
|||||||||||||||||)(|||||||||||||||||
|||||||||||||||||)(|||||||||||||||||
|||||||||||||||||)(|||||||||||||||||
|||||||||||||||||)(|||||||||||||||||

Visually, what I see is objects squeeze as they pass horizontally through the middle of the screen and bulge as they leave. It creates an unpleasant ripple. Vertical movement does not exhibit this flaw at all.

I'd be extremely grateful to hear some informed opinions about this. It's driving me crazy.
post #1913 of 2959
Thread Starter 
Quote:
Originally Posted by RM23J8G View Post

I'm getting more familiar with the service menu now, and downloaded some of Ken's test signals onto cd to play in my dvd player.
They really don't look all that good through that process...lot's of artifacts... and I'd like to try the memory stick method. I got better looking test signals by recording them from an NTSC test signal generator into a dvd recorder for some reason.

Those test patterns are constructed to display onscreen at 1:1 pixel ratio; nothing else will look correct on the fine-textured patterns.

The Frame buffer on my 36" 4:3 set is exactly 1080h X 1440w, and so many of the patterns are exactly that so the TV does NO rescaling. I can't confirm their value, but I also made some "16x9" patterns that are 1080h x 1920w, in case that is the right thing for widescreen-set owners.

If you simply record these patterns to DVD as stills or play them as jpegs from a CD, the DVD player crunches them to a different pixel ratio, likely 480h x 720w, and it may or may not give you square pixels: If circles aren't round, then it or you are overlooking the fact that DVD pixels have a 9:8, not square, aspect ratio.

The color-based patterns should work fine, but even my excellent Panasonic S97 doesn't handle jpeg-from-CD color very precisely. DVD stills are accurate, and the required rescaling doesn't hurt the colors at all. But you have to have DVD-authoring software to do it. I bought DVE and AVIA, and I don't much care to make any custom DVD with my own patterns.

The TV's manual describes the memory-stick basics. The reader module in the TV isn't especially fast, so you don't have to spend extra $$ on fast MS. It will read almost any form factor MS. I chose a couple 512K sticks on sale at Fry's -- Sony, but I don't think it matters.

You just have to buy a cheap memory-stick adapter for your USB port on your computer. $20 will get you an excellent multi-format (with also compact-flash, secure digital, etc) unit plus cable. This lets you treat the MS like an external disk drive.

Your computer should see the stick as a disk. Follow directions in the TV's manual to copy files onto the stick -- you have to follow a few simple rules. High-quality jpeg looks best. If you have a Mac, there are nuisance invisible "resource" files also generated on the stick that have to be deleted or they get in the way. (Path Finder software makes it easy.)

Hope this helps.

Quote:
Originally Posted by pchristensen View Post

Why not use the internal test patterns..? Read #416, #420 and #428

They're fine, as far as they go, but some are very clumsy to use, the color bars suck, and a valuable pattern may force you to evaluate only part of the screen, when what you need is a repeating pattern that occurs in different places on the screen. In other words, MS test patterns are limited only by your imagination and time, and the built-in patterns, while valuable, are extremely limited.
post #1914 of 2959
Thread Starter 
Quote:
Originally Posted by truesper View Post

Visually, what I see is objects squeeze as they pass horizontally through the middle of the screen and bulge as they leave. It creates an unpleasant ripple. Vertical movement does not exhibit this flaw at all.

I'd be extremely grateful to hear some informed opinions about this. It's driving me crazy.

I see a mild effect like this, too, and big scenic pans in movies make it fairly obvious!

But the adjustments offered in service mode are fairly coarse: You can squeeze or stretch the screen a bit over a third, maybe a quarter, of its width, but those narrow defects are a mystery to me. It could have to do with the cast-glass tube, a deflection-yoke anomaly, or some badly-placed magnets on the back of the tube -- who knows. Mine occurs in alignment with a slight cyan color shift just left of center, and I know ther is a "nest" of magnets (like cockroaches - I've seen 'em!) on the tube-back there. I'm trying ot get up the nerve to open the set and fiddle, but I haven't yet. I don't know what "problem" the magnets' placement was meant to solve in the first place.

If it is an electrical issue, I can't begin to guess what would cause a "glitch" in the huge sawtooth-shaped deflection current required in the deflection yoke to linearly sweep the beam with so little error.
post #1915 of 2959
Quote:
Originally Posted by KenTech View Post

I see a mild effect like this, too, and big scenic pans in movies make it fairly obvious!

But the adjustments offered in service mode are fairly coarse: You can squeeze or stretch the screen a bit over a third, maybe a quarter, of its width, but those narrow defects are a mystery to me. It could have to do with the cast-glass tube, a deflection-yoke anomaly, or some badly-placed magnets on the back of the tube -- who knows. Mine occurs in alignment with a slight cyan color shift just left of center, and I know ther is a "nest" of magnets (like cockroaches - I've seen 'em!) on the tube-back there. I'm trying ot get up the nerve to open the set and fiddle, but I haven't yet. I don't know what "problem" the magnets' placement was meant to solve in the first place.

If it is an electrical issue, I can't begin to guess what would cause a "glitch" in the huge sawtooth-shaped deflection current required in the deflection yoke to linearly sweep the beam with so little error.

Somebody suggested to me that it may be a horizontal linearity issue. Tweaking SLIN (stretch linearity) and/or HCNT (centering/linearity) was a suggested resolution, but this 27FS170 doesn't have those adjustments.

I'm planning to check out the 24FS120 at a store to see if it has the same problem. If I can't find a Wega model free of this defect, I'll have to go for some inferior brand (cry). Curved tube TVs have lots of distortion around the edges, but at least it's not right in the middle of the picture!
post #1916 of 2959
Thread Starter 
Quote:
Originally Posted by truesper View Post

Somebody suggested to me that it may be a horizontal linearity issue. Tweaking SLIN (stretch linearity) and/or HCNT (centering/linearity) was a suggested resolution, but this 27FS170 doesn't have those adjustments.

It is an h-linearity issue by definition: Any glitch in the smoothness of a horizintal pan indicates a departure from linearity in the sweep current. But the two adjustments you mention are really much too coarse to affect it. SLIN operates over the middle third of the screen, like this:

oooooSSSSSooooo

Looks like what you need is this:

oooooooSooooooo
post #1917 of 2959
Quote:
Originally Posted by KenTech View Post

It is an h-linearity issue by definition: Any glitch in the smoothness of a horizintal pan indicates a departure from linearity in the sweep current. But the two adjustments you mention are really much too coarse to affect it. SLIN operates over the middle third of the screen, like this:

oooooSSSSSooooo

Looks like what you need is this:

oooooooSooooooo

I'm surprised not to see more people complaining about this. I've seen the effect in every Sony TV I've looked at so far (32FS170, 27FS170, 34XBR970).

I thought I was going nuts or just too picky, but it's SO obvious every time the screen pans sideways. My old 20" fishbowl tv wasn't exactly the height of geometric correctness, but at least the distortion was at the edges rather than the center of the screen like this.
post #1918 of 2959
Quote:
Originally Posted by KenTech View Post

They're fine, as far as they go, but some are very clumsy to use, the color bars suck, and a valuable pattern may force you to evaluate only part of the screen, when what you need is a repeating pattern that occurs in different places on the screen. In other words, MS test patterns are limited only by your imagination and time, and the built-in patterns, while valuable, are extremely limited.

By the way, I found XBMC's picture viewer to be pretty good for displaying test patterns. It's a homebrew multimedia application that runs on a (mod-chipped) xbox game console. It's convenient because you can browse images over the network. You can also zoom/pan around on them which is useful for detecting some anomalies that are hard to see in a static image (such as the problem I've been complaining about).
post #1919 of 2959
Quote:
Originally Posted by truesper View Post

The effect I see is that objects appear to bulge and ripple as they pass through the middle of the screen.

I've tinkered extensively with PAMP and UPIN/LPIN, but those don't seem to influence this region of the screen. It seems to be slight concavity along a vertical line in the middle of the screen, rather than a fisheye.

i.e.,
|||||||||||||||||)(|||||||||||||||||
|||||||||||||||||)(|||||||||||||||||
|||||||||||||||||)(|||||||||||||||||
|||||||||||||||||)(|||||||||||||||||
|||||||||||||||||)(|||||||||||||||||

Visually, what I see is objects squeeze as they pass horizontally through the middle of the screen and bulge as they leave. It creates an unpleasant ripple. Vertical movement does not exhibit this flaw at all.

I'd be extremely grateful to hear some informed opinions about this. It's driving me crazy.


This is EXACTLY what is going on with my XBR970 as well! This is the problem I was trying to explain in the last paragraph of my post #1891, you described it a bit better than I, though . It seems to affect a larger portion of my screen though, almost 1/4 of the screen is noticeably 'pinched' in the center. I WAS able to use a combination of HCNT and SLIN to improve it a little, and also to fix the 'off center' nature of the stretching that I described a few posts back as well (one side stretched noticably more than the other). Now the screen is equally stretched to either side.

This concave center problem is VERY annoying though, as far as how anything moving horizontally across the screen bulges when at the sides and then squeeeezes as it passes thru the center. It makes it impossible to get a perfectly round circle when using the circle/grid test pattern. I have tried every setting in the SM that has any effect on geometry and have not found one that was able to fix this issue.

I was just about to call a Sony service place today and see if they'd send someone out to fix it, but hearing how many others are having the same problem, I wonder if it CAN be fixed. Hopefully someone here can figure out a setting or combination of settings that can fix this.


On another note, I picked up Avia on Friday and went thru all my settings again. I was already VERRRRY close or spot on with most of them. The THX Optimizer does cover many of the same bases and will get you like 80-90% of the results that Avia will as far as the picture adjustment goes (especially if you get the proper blue filter or glasses from THX, which I hadn't had when I used it hence why Avia did help me zero in on a few of the color adjustments). Avia does have a lot more advanced stuff that seems great but also seems like a lot of people wouldn't ever use it. I would suggest that anyone who has a DVD with the Optimizer on it already do a run thru with that and see if the results are satisfying enough before going out and plunking down $50 for Avia. All that said, I am glad that I bought it and even though the store I bought it from has an open return policy I will hold onto it anyways, as it has a lot of useful info on there that did make it worth the purchase price for me.

About the contrast adjustment on Avia though... I found that as some other people have commented, the test for adjusting contrast is kind of vague and I"m not sure the setting I arrived at is the 'optimal' setting. Even at the highest setting the white square never bloomed (seems newer TV's don't tend to do this). I just turned it down to where I thought the white area still looked white and not gray but w/o a reference I feel like I'm just 'guessing' even if I am close.


KenTech - did a service guy ever pay you a visit? Did you get any results?



EDIT: question, is there an agreed upon 'ideal' amount of overscan to shoot for? Seems the Avia DVD favors 5% overscan since many of it's geometry correction patterns have corner and side markers all placed at 5%. So that's where I adjusted to... just wondering what everyone else does.
post #1920 of 2959
bast525:

Contrast setting with Avia. The screen has some grey boxes toped by white on the upper half and a white section on the lower half. There is a line on either side to help you as well (needle pulse pattern).

For these sets, raise the Picture control until the lines on either side start to bend, they should be straight. Any Picture setting with bent lines is too high, back off until the lines are straight up and down. Your set may allow Picture to go all the way up without bending the lines, mine did not.

Now get right up in front of your set and look closely at the top portion of the screen. You are looking at the area between scan lines on the gray boxes. The size of the area between scan lines on the white box should be no smaller than the size of the area between scan lines of the grey boxes below it. This is not a ruler kind of measurement, but your eye will be good enough. If the Picture level is too high, the area between scan lines of the top two boxes will be reduced and possibly be missing.

The setting you come up with will probably be a whole lot lower than Sony's defaults and may take some getting used to. But the upside is that your display will be able to show you all of the signal without distortion.

I will leave the other questions to those better informed members of the forum.

Good luck
New Posts  All Forums:Forum Nav:
  Return Home
AVS › AVS Forum › Display Devices › Direct View (single tube) CRT Displays › THE SONY SERVICE CODES - Articles, Comments, Discoveries