AVS › AVS Forum › Display Devices › Display Calibration › Basic Guide to Color Calibration using a CMS (updated and enhanced)
New Posts  All Forums:Forum Nav:

Basic Guide to Color Calibration using a CMS (updated and enhanced) - Page 7

post #181 of 1810
7/28/07 at 3:25pm
OK anyway here is my files.

 

pioneer cms.zip 1.0771484375k . file
Reply
Reply

AVS Top Picks

post #182 of 1810
7/28/07 at 4:54pm
Quote:
Originally Posted by TomHuffman View Post

I don't think ColorFacts even has a measure of dE for the primaries/secondaries. I believe that it reports dE for gray scale only.


It has a raw data window which provides many colorspace coordinates- the dE it reports is w.r.t. to white. So you can read a color and get a dE - but unless you know the dE for the target color w.r.t. to white and can do vector math - a bit meaningless. A Proper dE would use the desired color target, a proper calibration tool would also give you chart instruments you can see from across the room. The reality is almost all calibration software on the market is designed for grayscale tuning - not color decoding and gamut tuning - pretty sad state of affairs considering grayscale has become a minor part of calibration work for decent digital displays - it is the color decoding and gamut displays intentionally are screwing up!
Reply
Reply
post #183 of 1810
7/28/07 at 5:48pm
Thread Starter 
Quote:
Originally Posted by spudbudy View Post

Hello gang, I have read here and on other topics that the CMS on the Pioneer elite's don't work. I would have believed that before but not now. I have calibrated my 1130 using the CMS controls and get what I consider a very good picture. But it only seems to work well in one of the many different picture modes. not sure why. Anyway my question for Tom is I have done the 75% white/red method for setting color. Tint is a little different because of the CMS it moves. Not a problem because you can move it back. Why can't you use a software program to set the secondary colors? Could you use equal parts of red and blue to get magenta, and equal parts of blue and green to get cyan? Or is it not a question of equal parts of the two colors but there location on the CIE chart. When I get a moment I will include my grayscale for the Pioneer.

I am baffled by this one. I spent several hours on the 1130. I could get the CMS to move the color points, but doing so screwed up the grayscale. I couldn't get them both right at the same time. What mode did it work in and what occurred in the modes where it didn't work?

Yes, some sets have color decoding adjustments that allow you to adjust the hues of the secondaries--Sony CRTs for example. The tint control adjusts the amount of blue and green in cyan. That's why changing the tint control moves cyan towards either of its associated primaries.
Reply
Reply
post #184 of 1810
8/8/07 at 9:06am
Quote:
Originally Posted by TomHuffman
I am baffled by this one. I spent several hours on the 1130. I could get the CMS to move the color points, but doing so screwed up the grayscale. I couldn't get them both right at the same time. What mode did it work in and what occurred in the modes where it didn't work?

Yes, some sets have color decoding adjustments that allow you to adjust the hues of the secondaries--Sony CRTs for example. The tint control adjusts the amount of blue and green in cyan. That's why changing the tint control moves cyan towards either of its associated primaries.
Tom it is possible that your instrument is too inaccurate to do CMS such that it will not wreck gray scale for that display. When you skew CMS too far it can really make a mess. The following is some data for my EyeOne Pro versus my narrow band PR-670 for 10 displays of the same model. The variance is even more problematic when you start looking a different models. I believe a good CMS requires a very high end instrument to avoid problems.

 

Eye One Pro vs PR-670.pdf 2.357421875k . file
Reply
Reply
post #185 of 1810
8/8/07 at 9:57am
Thread Starter 
Quote:
Originally Posted by umr View Post

Tom it is possible that your instrument is too inaccurate to do CMS such that it will not wreck gray scale for that display. When you skew CMS too far it can really make a mess. The following is some data for my EyeOne Pro versus my narrow band PR-670 for 10 displays of the same model. The variance is even more problematic when you start looking a different models. I believe a good CMS requires a very high end instrument to avoid problems.

No, it had nothing to do with the instrument. The problem was NOT subtle. This user also saw the same problem with all of the picture modes except one. It never occurred to me to make the adjustments in a different picture mode.
Reply
Reply
post #186 of 1810
8/10/07 at 8:24am
Just got through calibrating the Sharp 12K following Tom's instructions. I had to reduce contrast to eliminate red clipping using a 100IRE 75% window. I also chose a gamma I wanted and got that in the ballpark before I started on grey scale and color. I used a 6K temp setting--6500 setting on Sharp proved to actually be 7K. I got better results just dialing down temp from the outset, rather than pulling it down with just gain and offset controls. My final result was almost identical to Tom's results shown here http://www.avsforum.com/avs-vb/showthread.php?t=862447 Big surprise was that Color setting went from +1 to +14 when I used the red as a % of white method. Overall, color improvement after calibration is dramatic.

My only remaining issue is RGB gamma tracking. My RGB gamma tracks pretty tightly until 80-90IRE. In log mode, at 90IRE, with an average gamma of 2.5, R goes up to 2.6 and B down to 2.35. Is there a rough guide to how much divergence you can have before it is noticeable in the picture?
Reply
Reply
post #187 of 1810
8/10/07 at 8:37am
Thread Starter 
Quote:


Big surprise was that Color setting went from +1 to +14 when I used the red as a % of white method. Overall, color improvement after calibration is dramatic.

You should use the RGBCYM Value control in the CMS to adjust for decoding, rather than the Color control, which affects all colors equally.

Quote:


Is there a rough guide to how much divergence you can have before it is noticeable in the picture?

None that I know of.
Reply
Reply
post #188 of 1810
8/10/07 at 8:57am
Quote:
Originally Posted by TomHuffman View Post

You should use the RGBCYM Value control in the CMS to adjust for decoding, rather than the Color control, which affects all colors equally.

None that I know of.


Thanks. I did use the CMS controls to adjust Green and Blue, but my interpretation of your instructions was that you use the Color control to adjust Red. Is the procedure for Red different because of the Sharp's CMS controls?

Edit: I mean I used the CMS to adjust GBCYM, and color to adjust R
Reply
Reply
post #189 of 1810
8/10/07 at 9:43am
Thread Starter 
Quote:
Originally Posted by Orwellflash View Post

Thanks. I did use the CMS controls to adjust Green and Blue, but my interpretation of your instructions was that you use the Color control to adjust Red. Is the procedure for Red different because of the Sharp's CMS controls?

Edit: I mean I used the CMS to adjust GBCYM, and color to adjust R

Exactly. The Sharp has precise controls for all of the colors, so you shouldn't use the Color control at all. Use Red Value to adjust the level of red. It's the same with Tint. Leave that alone. Use the Cyan Hue control instead.

The Sharp has more controls than most displays. You should be able to dial in nearly perfect color, except green which is shifted towards yellow (at least on the Mk II) to a degree you probably can't fully adjust out.
Reply
Reply
post #190 of 1810
8/10/07 at 11:20am
Thanks Tom, that makes perfect sense. I'll recheck it. I remember now that when I was setting the decoder for the other colors that I checked red too, and it was at 21% of white, so I may have ended up in the right place anyway, but I need to double check that.

Yeah, green is definitely the big outlier, cyan to much lesser extent. Every other color dials in perfectly.
Reply
Reply
post #191 of 1810
8/17/07 at 8:11am
Tom,

I redid my color calibration using only the lightness, chroma, and hue settings in the CMS (though I did leave saturation and tint at +1 for DVI, by oversight) . Of course, for the color decoder adjustments, the only control that mattered was lightness (value). I got the following values: White Y 18.59, Red target 3.9, G 13.2, B 1.49, C 14.64, M 5.3 and Yellow 17.26. I was able to achieve all but Red, which I could only get up to 3.09. This resulted in CMS settings below:


CMS Light Chroma Hue
R +24 , -8 , +4
G +18 , -21 , +1
B +22 , 0 , +12
C +19 , 0, +1
M +28 , -10 , +13
Y +22 , -23 , -3

I was surprised that the lightness adjustments had to be so large for each color. I did a similar adjustment for component input with color and tint set to 0 and got similar results. Are these in line with the CMS settings you ended up with?
Reply
Reply
post #192 of 1810
8/17/07 at 8:48am
Thread Starter 
I don't have the 12K. I have the 20K and its color decoder is spot on requiring virtually no adjustment. I can't speak to proper 12K settings, though I am surprised that you are getting these readings. What instrument are you using?
Reply
Reply
post #193 of 1810
8/17/07 at 9:10am
Quote:
Originally Posted by TomHuffman View Post

I don't have the 12K. I have the 20K and its color decoder is spot on requiring virtually no adjustment. I can't speak to proper 12K settings, though I am surprised that you are getting these readings. What instrument are you using?

A DTP-94. Maybe someone else who owns and has calibrated the 12K can tell me if my values suggest a bad colorimeter or projector. I PMed Bob Sorel to ask what he got. I could post HCFR file, but everything looks great (except green and cyan, of course) as far as grayscale, CIE, temp, gamma. All settings are as follows:

Post Calibration Settings for Sharp XV-Z12000 Mark II
High contrast iris position, economy bulb setting, Da-Lite HP screen

HDMI to DVI from Toshiba XA2 HD DVD Average Gamma 2.37

Contrast -5
Bright +1
Color +1
Tint+1
Sharpness 0
Temp 6000
Gamma C1
CMS 1


R Gamma -3
G Gamma 0
B Gamma +1

R Gain +2
G Gain -22
B Gain 0

R Offset +10
G Offset -1
B Offset -7

CMS Light Chroma Hue
R +24 -8 +4
G +18 -21 +1
B +22 0 +12
C +19 0 +1
M +28 -10 +13
Y +22 -23 -3

Component from Sony B-301 Blu-Ray Average Gamma 2.27 (don't ask)

Contrast 0
Bright 1
Color 0
Tint 0
Sharpness 0
Temp 6000
Gamma C2
CMS 2

R Gamma -1
G Gamma -1
B Gamma -1

R Gain +3
G Gain -27
B Gain -5

R Offset 0
G Offset +1
B Offset -1

CMS Light Chroma Hue
R +21 -8 +5
G +19 -22 0
B +18 +5 +12
C +20 0 +3
M +26 -11 +11
Y +23 -23 +2
Reply
Reply
post #194 of 1810
8/19/07 at 1:26am
I calibrated my 12k with HCFR and a DTP-94. The color decoder was basically perfect, requiring no adjustments to the "Lightness" settings.

Rob
Reply
Reply
post #195 of 1810
8/19/07 at 3:24pm
Thanks Rob,

I am thinking these are the possibilities: 1. DTP-94 needs adjustment or replacement 2. My projector's color decoder is defective 3. I somehow misread Tom's instructions and have incorrectly measured color decoder settings. My lightness settings do seem extreme. I have a Da-Lite HP screen and I am using high contrast mode and low bulb setting, but none of that would affect color decoder, as far as I know. I am using HCFR software and taking Y (luminance) reading from continuous measurments window. Then I calculate % of white (100% stimulation) targets from Tom's table of REC709 values and adjust Y using lightness for each individual color in CMS until I get as close as possible to target. It seems pretty straightforward. I could try averaging a large number of readings for each color to reduce measurment error, but I doubt that is the problem.

Jack
Reply
Reply
post #196 of 1810
8/19/07 at 4:08pm
Are you sure your color patterns are at the same stimulus level as the white pattern?
Reply
Reply
post #197 of 1810
8/19/07 at 4:12pm
Quote:
Originally Posted by zoyd View Post

Are you sure your color patterns are at the same stimulus level as the white pattern?

Thanks, I hadn't thought of that, but I am using GetGray. I assume that they are the same stimulus level, and I am using 75% windows for all measurements.

Jack
Reply
Reply
post #198 of 1810
8/20/07 at 11:01pm
Post your HCFR file. Maybe we can figure it out from that.

Rob
Reply
Reply
post #199 of 1810
8/21/07 at 7:56am
Thanks Rob. Here are the HCFR files. I had to cut back on contrast substantially to eliminate a large clipping of red. On DVI, but not component, I have a whiter than white (WTW) showing 3 bars above 235 on a grayscale ramp (same for primary colors). I wonder if that has something to do with my odd color decoder settings. I have been discounting that as a factor because I am not getting WTW with component, and I need similarly large lightness increases in CMS to get its color decoder settings to match standards.

I have the DVI signal set to D video RGB. I will try computer RGB to see what happens.

Jack

 

FinalDVIXA2.zip 1.0810546875k . file

 

Component.zip 0.9677734375k . file
Reply
Reply
post #200 of 1810
8/21/07 at 4:11pm
Hey guys great thread, thanks Tom.

Take a look at my file and tell me what you think. This is with Toms method and only adjusting color and tint in the user settings. This is a Panny 900u with a DTP 94. I can get a perfect chart if I use the panny's CCM function but from what I understand about how it works there is no point in doing it. Correct me if I am wrong. Are these readings good enough as is?

75 white Y=7.67

R Y=1.62
G Y=5.69
B Y=3.40

 

cin1lampHmem1gmacrtc.zip 0.9765625k . file
Reply
Reply
post #201 of 1810
8/21/07 at 5:05pm
Thread Starter 
Regarding color decoding performance, red and green are fine, but blue is much too low. Gray scale and gamma are good, but the chromaticity of red, yellow, and green are considerably off.

Quote:
Originally Posted by audioholicJeffL View Post

Hey guys great thread, thanks Tom.

Take a look at my file and tell me what you think. This is with Toms method and only adjusting color and tint in the user settings. This is a Panny 900u with a DTP 94. I can get a perfect chart if I use the panny's CCM function but from what I understand about how it works there is no point in doing it. Correct me if I am wrong. Are these readings good enough as is?

75 white Y=7.67

R Y=1.62
G Y=5.69
B Y=3.40
Reply
Reply
post #202 of 1810
8/21/07 at 5:22pm
Thread Starter 
Based on your numbers, the 75% white window would be 10.32. If you measured RGB windows also at 75%, this gives

R: 27.4%
G: 110.8%
B: 12.5%

These numbers are way off. Only red is in the ballpark. What's up?

Quote:
Originally Posted by Orwellflash View Post

Thanks Rob. Here are the HCFR files. I had to cut back on contrast substantially to eliminate a large clipping of red. On DVI, but not component, I have a whiter than white (WTW) showing 3 bars above 235 on a grayscale ramp (same for primary colors). I wonder if that has something to do with my odd color decoder settings. I have been discounting that as a factor because I am not getting WTW with component, and I need similarly large lightness increases in CMS to get its color decoder settings to match standards.

I have the DVI signal set to D video RGB. I will try computer RGB to see what happens.

Jack
Reply
Reply
post #203 of 1810
8/21/07 at 5:35pm
Quote:
Originally Posted by TomHuffman View Post

Based on your numbers, the 75% white window would be 10.32. If you measured RGB windows also at 75%, this gives

R: 27.4%
G: 110.8%
B: 12.5%

These numbers are way off. Only red is in the ballpark. What's up?

You hit the nail on the head. Zoyd zeroed in on it right away. I was too much of a blockhead to pick up on it. I used the 100% gray pattern and compared it to 75% amplitude color patterns! Why? Because I thought 75% pattern for color referred to size of the window. Stupid, huh? Even a cursory reading of the GetGray manual would have disabused me of that notion. Sorry for wasting your time.

Jack

Edit: BTW, I realized my error about a half hour before you posted your question--that's why my response seemed so quick. I was getting ready to write Zoyd to confess, when I saw your post.
Reply
Reply
post #204 of 1810
8/21/07 at 5:59pm
Quote:
Originally Posted by TomHuffman View Post

Regarding color decoding performance, red and green are fine, but blue is much too low. Gray scale and gamma are good, but the chromaticity of red, yellow, and green are considerably off.

I could nail the CIE right on if I use the CCM function but from what I understand it only affects the selected color. Is it worth doing? or should leave it? How does the chromaticity being off on red, green and yellow affect the picture quality?
Reply
Reply
post #205 of 1810
8/21/07 at 6:41pm
Here it is with the CCM applied.

 

cin1lampHmem1gmacrtc.zip 0.9765625k . file
Reply
Reply
post #206 of 1810
8/21/07 at 10:23pm
Thread Starter 
Since all color is derived from combinations of red, green, and blue, if red and green are significantly off it will have a profound effect.

Regarding the CCM, what do your eyes tell you?

Quote:
Originally Posted by audioholicJeffL View Post

I could nail the CIE right on if I use the CCM function but from what I understand it only affects the selected color. Is it worth doing? or should leave it? How does the chromaticity being off on red, green and yellow affect the picture quality?
Reply
Reply
post #207 of 1810
8/22/07 at 5:21am
Quote:
Originally Posted by TomHuffman View Post

Since all color is derived from combinations of red, green, and blue, if red and green are significantly off it will have a profound effect.

Regarding the CCM, what do your eyes tell you?

Well on some scenes when I switch it on or off I can't see any change at all, but on others I can. Scenes with really bright colors are where it can be noticed the most especially in red areas and animated stuff. So I guess I will leave it on. Thanks for the input
Reply
Reply
post #208 of 1810
8/22/07 at 8:20am
What is interesting is that on the original adjustments before applying the color management settings, if I switch to rec 709 color space it lines up better. I wonder if that is telling me that my Denon dvd player is using rec 709 color space?
Reply
Reply
post #209 of 1810
8/24/07 at 8:14am
Quote:
Originally Posted by audioholicJeffL View Post

What is interesting is that on the original adjustments before applying the color management settings, if I switch to rec 709 color space it lines up better. I wonder if that is telling me that my Denon dvd player is using rec 709 color space?

That is the conclusion I would draw.
Reply
Reply
post #210 of 1810
9/2/07 at 3:00pm
Thread Starter 
Yes, the Samsung LED DLPs have a CMS in the service menu, but it is not easy to use. Selecting the Movie preset gets you close.

Quote:
Originally Posted by davidjschenk View Post

Hello.

I am entirely new to the project of calibrating one's display, but I have what is, for me, an important question in this matter. Does anyone know if the new LED locally-backlit Samsung xx81F LCD tvs have proper Color Management Systems in them? I read Tom's very informative guide on page one and easily see the importance of having a well-functioning CMS. I just don't know if these rather quirky new Samsung displays have them.

Also, as I will be using a 4081F almost exclusively as an all-purpose/HTPC computer monitor, and as I have individual color controls in my ATI (X800) control panel, is it possible for me to circumvent color management limitations in the display via that control panel? I assume there are at least some potential down-sides to such a move, yes?? What would they be?

Any help or advice is greatly appreciated, especially since I have not yet purchased any calibration equipment at all, but am ready to go with a Display2 colorimeter, GetGray, and HCFR, if indeed the color controls in my display and/or graphics card support the sort of fine-tuning these tools offer.

Yours,

David Schenk
Reply
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Display Calibration

AVS Top Picks

AVS › AVS Forum › Display Devices › Display Calibration › Basic Guide to Color Calibration using a CMS (updated and enhanced)