or Connect
AVS › AVS Forum › Display Devices › Display Calibration › Issue with calibrating Panasonic ST60 and skin tones
New Posts  All Forums:Forum Nav:

Issue with calibrating Panasonic ST60 and skin tones - Page 4

post #91 of 146
Thread Starter 
Quote:
Originally Posted by NeilPeart View Post

I just installed my Panasonic 60" ST60 and noticed the same greenish hue after my self-performed calibration (set is new and has not broken in). My reference set prior to this one was the Pioneer PRO-101FD calibrated by D-Nice (which is happily sitting in the family room). I hope that with time and another calibration this issue is mitigated - I appreciate the larger size of this set but there is something about the Kuro that is unmatched yet intangible (just looks more natural/organic). I hope I can wait it out until 70"+ OLEDs can be had for under $5K biggrin.gif

I'm hoping I can live with this one until OLED too. Hopefully yours isn't too bad and still watchable. The more time I spend watching content on this set the more I think I can pick out the patterns of where the problem is occurring. I think I've nailed it down to the "area-based ABL" as I'm calling it. It's quite apparent to me on any human skin tones or faces. It's known that ABL effect warms up the color temperature a little bit, but what I've noticed is that any particular spot on the screen gets an ABL effect relative to the sum of bright areas around it. That might not make sense, but an example would be a person's face in certain lighting conditions. I think this is really noticeable on people with very dark hair. The center of their face will be the warmest/reddest, but the top of their head near their hairline (which dark hair won't be emitting very much light) often takes on a greener tint. I don't think it's necessarily overly green, but I think the rest of their face becomes too warm and when viewed together the green is more likely to stand out as unnatural. Similarly, if someone was in a dark room but their face was lit, the "glow" would occur on the outside of their face all around.

An opposite example which I witnessed today: a scene with a dark skinned man turned to the side, showing his profile, everything looked completely natural except his ear (near the center of his head) which had a very green tint. The surrounding area around his head was bright and adding more redness & warmth to the edges of his face while the center was further from this effect so it remained greenish in comparison.

This can even be seen on a full-field white screen. There's a faint red glow coming from the center of the screen and the edges remain whiter. This also emphasizes that things aren't actually too green but in skin tones we might tend to pick the redder shades as looking more pleasant or natural.

Does this sound accurate to how ABL is supposed to work? Because that's what I'm witnessing. confused.gif And I'm betting every problem I have with this set is caused by this effect.

To chunon's point from earlier, I have settled on some settings which are a little overly saturated as it helps take away from the problem slightly... in the end I'm actually using very basic settings without any detail calibration at all, because it's gotten to the point where the more I think about it the less I can enjoy this $1000 I spent 7 months ago. Oy.
Edited by vkamicht - 12/15/13 at 12:13pm
post #92 of 146
I've just accepted it as a limitation of the set, it is very content specific does it still bother me ? Yes but the ST60 does have alot of strengths
post #93 of 146
I understand what you guys are saying. I went from a Sony Super Fine Pitch tube (CRT reference) to the Pioneer Elite Kuro generation 9.5 (plasma reference), so I think I'm just used to a reference picture, but my wife doesn't notice the issue at all and thinks I'm crazy (and she is usually very astute with this type of observation). I think I am being a bit too picky considering I got this 60" TV for less than $1400. I'm breaking it in with D-Nice's slides and will perform a calibration after that is complete - I think a bit of red push/green cut will mitigate this issue. What do you guys think about that? I wonder if the VT has the same "limitation?" tongue.gif
post #94 of 146
For me it isnt a greyscale issue, it is a color decoding issue but as I mentioned it does not show itself all the time just certain content and apl levels. My set has been professionally calibrated and I spent alot of time trying to correct it myself. I can't return the set so I have to accept the issue or drive myself crazy smile.gif I think the VT has a better panel and is more accurate color wise havent heard of anyone with this issue on the VT. I am coming from a VT50 which is the best set I have every owned. I consider myself pretty picky but you can only spend so much time banging your head against the wall.
post #95 of 146
Chunon, have you tried it in HDMI - graphics mode like the 30 series needed to be for better gamma and color decoding.
post #96 of 146
No I haven't but I would assume the CMS would be capable of doing whatever the graphic option does to the color space. It measures fine it just doesn't perform like it measures with some material
post #97 of 146
For ones having issues what about taking a DVD that displays these problems down too the store and on the same model tv with same settings see if it's happening on it.
Edited by Vic12345 - 12/16/13 at 4:09pm
post #98 of 146
Thread Starter 
Quote:
Originally Posted by chunon View Post

No I haven't but I would assume the CMS would be capable of doing whatever the graphic option does to the color space. It measures fine it just doesn't perform like it measures with some material
Quote:
Originally Posted by PeterLewis View Post

Chunon, have you tried it in HDMI - graphics mode like the 30 series needed to be for better gamma and color decoding.

I played around with the settings tonight and like he says, they don't really do anything that you can't do manually (and thus doesn't help the issue.) The only one that looked any different was "Photos" which altered the gamut to some point in the middle of the "Native" and "Normal" options (and when you're on this mode it disables that color space option entirely)
post #99 of 146
Quote:
Originally Posted by vkamicht View Post

Hey guys, just looking for some quick advice here. I bought the cheap CalMan + C3 package a few months ago and have tried to calibrate my 50" ST60 multiple times. I always manage to get very good readings, but real world performance tells a different story. I find that the color balance seems to change from scene to scene. I am wondering if this is a byproduct of the brightness limiter, and if there's a way to get around this, or if I've calibrated something wrong. I used the smallest windows available to me to avoid the limiter activating during calibration. My main gripe is when watching anything on it, peoples skin tones go from being reddish to greenish between scenes, sometimes even on the same scene (skin can look "spotty" with very different shades, like patches of yellow on faces) It happens on all modes, cinema/custom, etc. Has anyone here dealt with or resolved this problem on this set? I'm about ready to call in a pro if they know what I'm talking about and are confident that they can fix this. I do notice issues in other areas as well, such as whites appearing too green in certain scenes then perfect in others. Any suggestions? Thanks

Have you updated to the latest Calman? Spectracal updated the tables for the C3 in 5.2.0 specifically to make it more accurate with plasmas. I have a C3 and 50ST60, and what I noticed is that for a calibration that measured flat in 5.1.2, in 5.2.0, green steadily increased from low to high IRE to being up maybe 3% at 100IRE, and red steadily decreased to being down about 8% at 100IRE compared to 5.1.2. (I'm estimating from the RGB Balance chart in the Quick Analysis workflow.) The DEs that were < 1 in 5.1.2 steadily increased to 5 at 100IRE in 5.2.0. So, when I redid the calibration in 5.2.0, the main thing was to increase the red from low to high IREs. Put another way, 5.1.2 measured red significantly higher than 5.2.0, leading me to knock it down in that earlier version.

And it doesn't stop there. biggrin.gif I recently bought a D3, and it reads red even lower than the C3 in 5.2.0. That is, C3/5.1.2 measured red > C3/5.2.0 > D3. For the new D3 calibration, C3/5.1.2 reads red up about 15% at 100IRE, with green down about 3%, and C3/5.2.0 reads red up about 8%, with green basically flat. Again, where different, it's a pretty steady ascent or descent over the whole IRE range, from low to high IRE.

I used 5% windows and the PC Client for all this and verified throughout against Mascior's 6.5% patterns with my Sony S5100 BD player. I found that when I used the Spears and Munsil Equal Energy patterns, I ended up with green-tinted skin tones in some low APL scenes, which is what I wrote about earlier in the thread. However, I've never seen these weird shifty color changes within a single scene you've described.

As for 21 point grayscales, I'm not observing huge spikes in the odd numbered IREs. The largest was still less than 2 dE.

Subjectively, I liked my C3/5.2.0 results better than 5.1.2 even before it was confirmed the change to Calman was intentional; I'm able to A/B two different inputs with about a second to switch between them by using the ST60's discrete input commands, which let me press a single button on my remote to select any input on the TV. I haven't gotten around to A/B'ing the D3 and C3/5.2.0 calibrations, but the D3 results look fine, and the meter measures faster and more reliably, especially at 0%.

The above is all about my 50ST60. For calibrating my NEC 20WMGX2 CCFL monitor, the C3 and D3 agree more closely, each measuring the other's auto-calibration at about 1.7 dE average. It's very hard to tell the ICCs apart.

Now I want a spectrophotometer. (And I sincerely hope that's the end game. smile.gif) While I know it doesn't necessarily mean anything, I found it interesting that the D3 gets my blue filter significantly closer to "zeroing out" than the C3.
post #100 of 146
Quote:
Originally Posted by vkamicht View Post

The more time I spend watching content on this set the more I think I can pick out the patterns of where the problem is occurring. I think I've nailed it down to the "area-based ABL" as I'm calling it. It's quite apparent to me on any human skin tones or faces. It's known that ABL effect warms up the color temperature a little bit, but what I've noticed is that any particular spot on the screen gets an ABL effect relative to the sum of bright areas around it. That might not make sense, but an example would be a person's face in certain lighting conditions. I think this is really noticeable on people with very dark hair. The center of their face will be the warmest/reddest, but the top of their head near their hairline (which dark hair won't be emitting very much light) often takes on a greener tint. I don't think it's necessarily overly green, but I think the rest of their face becomes too warm and when viewed together the green is more likely to stand out as unnatural. Similarly, if someone was in a dark room but their face was lit, the "glow" would occur on the outside of their face all around.

OK, I think I see what you're talking about now, and I've got a good repro case for it.

  1. I took a screenshot of a face close-up in XBMC, and I resized it to 50%, so it takes up 1/4 of the TV's display area.
  2. I load it into Windows 7 Paint and put the window in the upper left corner of my 50ST60. Make sure your wallpaper is pure black, so the window sits on a black screen.
  3. Now I drag the sizing rectangle at the bottom right corner to resize the window. The image stays the same size and at the same position within the Paint window, but the window's light blue background causes the APL to change as I resize the window. As I do this, there are several sudden shifts in the skin tone, with red coming and going a couple of times. That is, it will start out red, lose it, get it back, lose it again, then get it back, etc. These are small but definitely noticeable changes.

Here's the image I'm using. It's one of the Pioneer Kuro demo disc models, and she's got very light skin with a hint of rosiness in her cheeks.



I'd be interested in hearing if other plasmas do this. It's not something I've really noticed when watching shows, I guess because lingering close-ups with slow but extreme APL variation are pretty rare. My LCDs don't do it at all.
post #101 of 146
Sawfish,

The blue windows desktop might have something to do with it.

You might want to try a neutral gray background for the desktop and see it you get the same color shift.

I've heard it said about the VT50s that the color was non linear and varies with APL. That might also apply to the 60s. Your test sounds like a good way to test it..
post #102 of 146
Quote:
Originally Posted by JimP View Post

Sawfish,

The blue windows desktop might have something to do with it.

You might want to try a neutral gray background for the desktop and see it you get the same color shift.

It's not the desktop that is blue. Like I said, make sure the wallpaper is black, so the window sets on a black background. The whole screen should be black, with not even the taskbar showing. I extend the desktop to the TV, and that's the way it is for me. I guess you could also configure the taskbar to auto-hide to take it (mostly?) out of the equation.

The light blue is the background of the Paint window, and I can't change that, though I was able to achieve the same effect. The important thing is for the image to stay fixed in size and position while the color or brightness varies around it, so I loaded the image into Paint.net (not Paint), increased the Canvas size to 1920x1080, keeping the image at the upper left and unscaled, and then filled the new area with gray. Then I loaded the new image into Paint and carried out the same procedure. I got the same results, and I can see the gray changing color. It also makes it easy to see the "area bleed", which is like line bleed but covers the whole area to the right of the image. That, is the gray to the right of her face is a different color than the gray below the image.

Quote:
I've heard it said about the VT50s that the color was non linear and varies with APL. That might also apply to the 60s. Your test sounds like a good way to test it..

I guess the good news is, I rarely notice these color shifts and line bleed in real video material. I calibrate against 5% windows where I think these effects are minimal, and I like the overall results across the APL range much better than when I tried the S&M Equal Energy patterns, which got me green-tinged skin tones in certain low APL scenes.
post #103 of 146
My Pioneer PRO-101FD (Kuro Elite plasma monitor) never exhibited such behavior, nor did my Sony KD-34XS955 (Sony Super Fine Pitch CRT). I still have both sets and they look great. I don't recall seeing it on my friend's Samsung plasma either.
post #104 of 146
I did a calibration in Cinema with a 2.3 gamma target this weekend, used 6.5% standard windows. The end result is excellent with great color saturation and pop. The display actually performs as it measures now and the green tint has been greatly reduced in low apl scenes. Might want to give that combo a try and let me know what you think.
post #105 of 146
chunon

What I'd like to ask you to do is to compare your results to that in standard picture mode with just the user level adjustments that are available there and see if low APL scenes have the same green color shift. Use normal color temperature.
post #106 of 146
I'll give it a try let you know what I see, why standard mode ?
post #107 of 146
I'm speculating that Panasonic did more than we're giving them credit for to get standard picture mode correct.
post #108 of 146
Okay will see what it looks like, I think the same might hold true for Cinema. Alot of assumptions are carried forward from model year to model year and I don't think they are all necessarily true.
post #109 of 146
Quote:
Originally Posted by chunon View Post

I did a calibration in Cinema with a 2.3 gamma target this weekend, used 6.5% standard windows. The end result is excellent with great color saturation and pop. The display actually performs as it measures now and the green tint has been greatly reduced in low apl scenes. Might want to give that combo a try and let me know what you think.

About the small windows, told ya. biggrin.gif

Any particular reason to use Cinema? I find that it has some white and red clipping that is absent from Custom, and the CMS also doesn't do as well. For example, the 60% green is oversaturated and the 80% green is undersaturated, whereas in Custom, all the colors line up better. (Using Normal gamut for both, of course.)

I really liked the results with BT.1886 and specifying .005 for the black level. This gives a gamma around 2.2 at 10% IRE rising to 2.35 at 100%. I think your 2.3 should be very close, but I definitely found 2.4 to be too dark, so I always used 2.2 before the recent Calman updates defaulted to BT.1886.

It's interesting to measure 2.6 gamma as set in the TV with the gamma and WB otherwise at their defaults. Like BT.1886, it starts off lower and increases with higher IREs, though compared to BT.1886, it's about .05 lower at 10% IRE, and it's more of a straight line up to 70%. This demonstrates why all the posted calibrations targeting 2.4 gamma have those huge negative numbers for low IREs and the overall pattern of the 10 point gamma settings.
post #110 of 146
I like the look of Cinema better less noise and dithering imo, have always favored the smaller windows but always used APL windows prior to this set. I didn't see an saturation issues in CMS but maybe I am missing something. I do agree tho the 2.6 just seems to be a more natural preset for whatever reason.
post #111 of 146
Quote:
Originally Posted by chunon View Post

I like the look of Cinema better less noise and dithering imo

I can't really tell any difference there with each mode calibrated for the same light output.

Quote:
have always favored the smaller windows but always used APL windows prior to this set.

Yes, my suggestion a few weeks ago was to use the non-APL 5% windows. Glad to hear it worked out.
post #112 of 146
Quote:
Originally Posted by sawfish View Post

Quote:
Originally Posted by chunon View Post

I did a calibration in Cinema with a 2.3 gamma target this weekend, used 6.5% standard windows. The end result is excellent with great color saturation and pop. The display actually performs as it measures now and the green tint has been greatly reduced in low apl scenes. Might want to give that combo a try and let me know what you think.

About the small windows, told ya. biggrin.gif

Any particular reason to use Cinema? I find that it has some white and red clipping that is absent from Custom, and the CMS also doesn't do as well. For example, the 60% green is oversaturated and the 80% green is undersaturated, whereas in Custom, all the colors line up better. (Using Normal gamut for both, of course.)

I really liked the results with BT.1886 and specifying .005 for the black level. This gives a gamma around 2.2 at 10% IRE rising to 2.35 at 100%. I think your 2.3 should be very close, but I definitely found 2.4 to be too dark, so I always used 2.2 before the recent Calman updates defaulted to BT.1886.

It's interesting to measure 2.6 gamma as set in the TV with the gamma and WB otherwise at their defaults. Like BT.1886, it starts off lower and increases with higher IREs, though compared to BT.1886, it's about .05 lower at 10% IRE, and it's more of a straight line up to 70%. This demonstrates why all the posted calibrations targeting 2.4 gamma have those huge negative numbers for low IREs and the overall pattern of the 10 point gamma settings.
Why .005? My GT50 measures down to .003.
post #113 of 146
I've noticed similar things with mine. I was watching Bill Maher a few weeks ago, and when a grey bar with a guests name would pop up on the bottom of the screen, you could see the skin tons lose red and shift green slightly.

After seeing the TV do that kind of thing, one has to wonder how on earth anyone could reach settings that one could call "calibrated". I mean for the most part, the picture looks great, but I occasionally notice stuff like that, and it bothers me. While I do love the picture a plasma reproduces in comparison to an LCD, I have to say that I don't really care for plasma technology.
post #114 of 146
Quote:
Originally Posted by HD-Master View Post

Why .005? My GT50 measures down to .003.

I always liked the shadow detail for gamma 2.2 so went with .005, which gives me about a 2.2 target at 10% IRE. The .005 is basically just a nice, round number not too far off from what others have reported for the ST60. I've recently bought a D3 which doesn't error out at 0% like my C3 did, and it measures my 50ST60 between .0025 and .003, which ISTR is what people with reference level equipment reported. Plugging this in for the black level makes a small difference in the gamma curve, raising the gamma at 10% to somewhere around 2.25, but I haven't gotten around to testing it with the "real" value.

Quote:
Originally Posted by ttnuagmada View Post

I've noticed similar things with mine. I was watching Bill Maher a few weeks ago, and when a grey bar with a guests name would pop up on the bottom of the screen, you could see the skin tons lose red and shift green slightly.

After seeing the TV do that kind of thing, one has to wonder how on earth anyone could reach settings that one could call "calibrated". I mean for the most part, the picture looks great, but I occasionally notice stuff like that, and it bothers me. While I do love the picture a plasma reproduces in comparison to an LCD, I have to say that I don't really care for plasma technology.

The plasma technology has its flaws, that's for sure. However, I'm continually amazed how infrequently I notice the effects of ABL and line bleed, the two things I mentioned earlier. Outside of animation and the brief time I spend on the computer desktop, it's quite rare for me to notice them. As far as calibration goes, you can be correct at one point but are at the mercy of the ABL everywhere else. The challenge is to find the best "one point". I settled on the 5% windows for my 50ST60 after trying the S&M Equal Energy patterns, a type of constant APL pattern, whose higher APL engages the ABL and is subject to red push, which led me to turn the red down, and I think this is what caused some low APL scenes to yield green-tinted skin tones. OTOH, using the 5% windows does result in elevated reds in high APL scenes, but it doesn't look unnatural, and in fact, I think it still looks great. About the only time I question it is in Family Guy, because Peter Griffin looks a little pinker than he probably should.
post #115 of 146
Quote:
Originally Posted by ttnuagmada View Post

I've noticed similar things with mine. I was watching Bill Maher a few weeks ago, and when a grey bar with a guests name would pop up on the bottom of the screen, you could see the skin tons lose red and shift green slightly.

After seeing the TV do that kind of thing, one has to wonder how on earth anyone could reach settings that one could call "calibrated". I mean for the most part, the picture looks great, but I occasionally notice stuff like that, and it bothers me. While I do love the picture a plasma reproduces in comparison to an LCD, I have to say that I don't really care for plasma technology.

I agree calibrating a plasma can be a frustrating experience and I think the ST60 is quirkier than sets in past years. My VT50 was much more consistent, what is responsible for that I don't know but color shift is definitely an issue with the 60 series imo.
post #116 of 146
 
Quote:
Originally Posted by chunon View Post


I agree calibrating a plasma can be a frustrating experience and I think the ST60 is quirkier than sets in past years. My VT50 was much more consistent, what is responsible for that I don't know but color shift is definitely an issue with the 60 series imo.

 

Quote:
Originally Posted by sawfish View Post


I can't really tell any difference there with each mode calibrated for the same light output.
Yes, my suggestion a few weeks ago was to use the non-APL 5% windows. Glad to hear it worked out.

 

I just bought a floor model 65st60 and it doesn't seem to have the PQ my S60's have. Since I don't have any of this calibration stuff, would you two guys mind sharing your settings? I have it set on cinema now with the suggested stuff shut off. I, and I'm sure others would really appreciate it!

 

Merry Christmas to all of you!


Edited by CatchersDad - 12/24/13 at 8:42pm
post #117 of 146
Sharing the settings is not the best solution because all tv are different (composants variability, tolérance...)
The same settings on different tv (same model, same brand, same size...) make most of the time the image never look the same.

And with sharing settings the tv have more chance to have a worst looking than a stock settings.

My advise : choose true cinema mode, gamma 2.4, all others stuff off, choose à colorspace in your bd player like YCbCr 4:2:2 or 4:4:4.

Download a free calibration disc (avshd709, gcd, mcd...) To check brightness (reference black at video level 16 with the black clipping pattern) check contrast (reference white at video level at 235 with the apl clipping pattern - don't use white clipping pattern to not activate abl of the plasma - Check the white clipping, check for eye strain, check if white is not tinted - pink) and check color clipping.
Check sharpness and overscan with the sharpness/overscan pattern.

That's the best solution for users without calibration gear wink.gif

Hope this help
post #118 of 146
Quote:
Originally Posted by CatchersDad View Post


I just bought a floor model 65st60 and it doesn't seem to have the PQ my S60's have. Since I don't have any of this calibration stuff, would you two guys mind sharing your settings? I have it set on cinema now with the suggested stuff shut off. I, and I'm sure others would really appreciate it!

Merry Christmas to all of you!

You can find settings to try on CNET, Sound and Vision magazine and also in the ST60 series thread. Putting in someone else's settings is not calibration.
post #119 of 146
I calibrated today using my limited gear and limited experience. I started by using the DVE HD Basic disc to set brightness/contrast and then used Calman5 along with the AVS HD 709 disc and C3 meter. Finally I used the S&M disc to verify 24Hz operation and other miscellaneous things. I performed two passes with the meter and calibrated the Cinema and Custom modes. The skin tone issues seem to be resolved (the TV has ~200 hrs), but no matter what I do I have to set the Color setting to 60 in the Cinema mode. The Custom mode is correct after calibration of grayscale and CMS (color set to 50) but the Cinema mode seems to have a decoder error with my particular TV. How important is the decoder/color setting? Also, no matter what I do CInema seems a bit brighter than Custom - is that normal?
post #120 of 146
I was able to calibrate Cinema with the default color of 50 and CMS controls no color tracking issues that I am aware of. Are you using the a blue filter to set color ? That is not an accurate way to do it,
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Display Calibration
AVS › AVS Forum › Display Devices › Display Calibration › Issue with calibrating Panasonic ST60 and skin tones