Westinghouse color...is it really this bad? - AVS Forum
Forum Jump: 
Reply
 
Thread Tools
post #1 of 67 Old 01-05-2010, 08:06 AM - Thread Starter
AVS Special Member
 
angryht's Avatar
 
Join Date: Aug 2004
Location: Omaha, NE
Posts: 1,321
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 11
I've been trying to get the colors right on my Westinghouse LCD (LTV-40w1 HDC) using HCFR and a I1 LT colorimeter. It doesn't have CMS so I am limited to saturation and hue controls. Grayscale has been adjusted to fairly close to D65. Based on Tom Huffman's guide, I adjusted the saturation so that the red brightness is very close to the target (adjusted slightly based on my specific gamut). I then adjusted the hue control to get the cyan aligned with the target. This adjustment rotated all of the secondaries counterclockwise. Prior to the hue adjustment, the magenta was very close to the target but the yellow was already on the green side. After the hue adjustment, the cyan was spot on but yellow was even closer to green and magenta was shifted closer to red. And when I put on some actual content with faces (skintones) - it looked horrendous!! Everybody looked sunburn (red push?) with a little bit of a greenish (sickly) look. I am not able to post my HCFR until later tonight or tomorrow night.

Two questions:
1. I thought that by adjusting the red brightness to the correct target, red push would not be this bad. It appears that I was incorrect. Can someone educate me a little? I guess I don't quite understand it. Isn't red push an increase in red brightness to compensate for a blueish (higher color temperature)? I'm guessing red push also screws with all the other colors.
2. Anybody have any experience with these Westinghouse LCD's? Are they really this bad? It seems like I need to find a compromise in terms of adjusting my secondaries. I remember reading that errors in yellow were especially problematic. Any suggestions?

-Greg
angryht is offline  
Sponsored Links
Advertisement
 
post #2 of 67 Old 01-05-2010, 08:55 AM
AVS Club Gold
 
Bear5k's Avatar
 
Join Date: Feb 2006
Location: Houston, TX
Posts: 2,239
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by angryht View Post

Two questions:
1. I thought that by adjusting the red brightness to the correct target, red push would not be this bad. It appears that I was incorrect. Can someone educate me a little? I guess I don't quite understand it. Isn't red push an increase in red brightness to compensate for a blueish (higher color temperature)? I'm guessing red push also screws with all the other colors.

Setting aside Tom's method, was the color decoder correct before you started changing the controls? In other words, using your actual gamut, did your primaries and secondaries line-up with where they were supposed to be given whatever color temperature your set was then calibrated to display?

If the decoder was accurate, then you don't have red push as it is traditionally defined (an emphasis on red in the color decoder itself). Note that gamut issues are not necessarily the same as red push.

Quote:


2. Anybody have any experience with these Westinghouse LCD's? Are they really this bad? It seems like I need to find a compromise in terms of adjusting my secondaries. I remember reading that errors in yellow were especially problematic. Any suggestions?

Lots of errors can be problematic. The question is which ones and of what type. We don't see hue well in green, but a deviation in brightness can be very visible. A counterpoint to this is blue, where we see lots of different hues of blue, but we are less sensitive to changes in brightness. Yellow becomes important as a component of flesh tones, but not necessarily for "pure" yellows.

Color accuracy evangelist and CalMAN insider
Bear5k is offline  
post #3 of 67 Old 01-05-2010, 10:28 AM - Thread Starter
AVS Special Member
 
angryht's Avatar
 
Join Date: Aug 2004
Location: Omaha, NE
Posts: 1,321
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 11
Quote:
Originally Posted by Bear5k View Post

In other words, using your actual gamut, did your primaries and secondaries line-up with where they were supposed to be given whatever color temperature your set was then calibrated to display?

That's a good question. I guess I'm not sure. I measured the grayscale initially, adjusted it to close to D65, then moved on to the saturation and hue.

Quote:
Originally Posted by Bear5k View Post

If the decoder was accurate, then you don't have red push as it is traditionally defined (an emphasis on red in the color decoder itself). Note that gamut issues are not necessarily the same as red push.

By 'accurate', do you mean the secondaries line up with the targets, from my specific gamut, when the saturation and hue controls are in the neutral position with no color or tint adjustment and prior to my grayscale adjustment? I guess I don't know the answer to that, yet. I have noticed that the lines between my primaries and the secondaries do not intersect my new grayscale, when adjusted. In order to adjust the saturation control to get the red brightness correct, I increased it from neutral of 25 to 29. To get cyan on the target, I adjusted if from the neutral of 25 to 21. So maybe I need to start from scratch.


Quote:
Originally Posted by Bear5k View Post

Lots of errors can be problematic. The question is which ones and of what type. We don't see hue well in green, but a deviation in brightness can be very visible. A counterpoint to this is blue, where we see lots of different hues of blue, but we are less sensitive to changes in brightness. Yellow becomes important as a component of flesh tones, but not necessarily for "pure" yellows.

Thanks, this is exactly what I was wondering about. I know that I'm not happy with the results so far so I'll have to keep adjusting. Hopefully, I can get the HCFR files posted tonight.

-Greg
angryht is offline  
post #4 of 67 Old 01-05-2010, 11:38 AM
AVS Special Member
 
Geof's Avatar
 
Join Date: Feb 1999
Location: Eden NY
Posts: 6,012
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 5 Post(s)
Liked: 33
Quote:
Originally Posted by angryht View Post

I've been trying to get the colors right on my Westinghouse LCD (LTV-40w1 HDC) using HCFR and a I1 LT colorimeter. It doesn't have CMS so I am limited to saturation and hue controls. Grayscale has been adjusted to fairly close to D65. Based on Tom Huffman's guide, I adjusted the saturation so that the red brightness is very close to the target (adjusted slightly based on my specific gamut). I then adjusted the hue control to get the cyan aligned with the target. This adjustment rotated all of the secondaries counterclockwise. Prior to the hue adjustment, the magenta was very close to the target but the yellow was already on the green side. After the hue adjustment, the cyan was spot on but yellow was even closer to green and magenta was shifted closer to red. And when I put on some actual content with faces (skintones) - it looked horrendous!! Everybody looked sunburn (red push?) with a little bit of a greenish (sickly) look. I am not able to post my HCFR until later tonight or tomorrow night.

Two questions:
1. I thought that by adjusting the red brightness to the correct target, red push would not be this bad. It appears that I was incorrect. Can someone educate me a little? I guess I don't quite understand it. Isn't red push an increase in red brightness to compensate for a blueish (higher color temperature)? I'm guessing red push also screws with all the other colors.
2. Anybody have any experience with these Westinghouse LCD's? Are they really this bad? It seems like I need to find a compromise in terms of adjusting my secondaries. I remember reading that errors in yellow were especially problematic. Any suggestions?

I'm no expert but without CMS you are limited to adjusting Brightness (or Lightness) and Hue and not saturation. I have a Sony and can adjust R,G, & B lightness and hue but not saturation - and unfortunately Red (for example) is over saturated. If I tried to adjust Red saturation by adjusting Red Lightness it just doesn't work well.

Geof
Geof is online now  
post #5 of 67 Old 01-05-2010, 12:30 PM - Thread Starter
AVS Special Member
 
angryht's Avatar
 
Join Date: Aug 2004
Location: Omaha, NE
Posts: 1,321
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 11
You are right. I don't have anything close to a CMS. In my initial post I was refering to the color control, which Westinghouse calls saturation. I understand it doesn't actually control saturation, just a normal color control. Although, it does slightly move the location on the CIE chart when adjusted (as do other cheap sets).

-Greg
angryht is offline  
post #6 of 67 Old 01-05-2010, 01:00 PM
AVS Special Member
 
Geof's Avatar
 
Join Date: Feb 1999
Location: Eden NY
Posts: 6,012
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 5 Post(s)
Liked: 33
Quote:
Originally Posted by angryht View Post

You are right. I don't have anything close to a CMS. In my initial post I was refering to the color control, which Westinghouse calls saturation. I understand it doesn't actually control saturation, just a normal color control. Although, it does slightly move the location on the CIE chart when adjusted (as do other cheap sets).

Ah, that 'splains it. I wish these manufacturers would [use] correct terminology!

Geof
Geof is online now  
post #7 of 67 Old 01-05-2010, 01:25 PM
Member
 
Plasma54321's Avatar
 
Join Date: Sep 2008
Location: Florida
Posts: 164
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
This sounds familar to the frustration I had at setting Color and Tint controls on my Panasonic TH-42PZ85. I finally found a calibrated (using HCFR software and X-rite Display LT meter) result that was very pleasing and realistic for all colors and especially for the face or skin tones. The details are in this post:

Methods to Optimize Color and Tint Settings without CMS Controls (http://www.avsforum.com/avs-vb/showthread.php?t=1188417)

If you show your CIE Diagram and RED target info I may get you an alternative method to try out. Basically involves setting the Color control for Red then adjusting Yellow for the best balance on your gamut for your TV.
Plasma54321 is offline  
post #8 of 67 Old 01-05-2010, 02:33 PM - Thread Starter
AVS Special Member
 
angryht's Avatar
 
Join Date: Aug 2004
Location: Omaha, NE
Posts: 1,321
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 11
Quote:
Originally Posted by Plasma54321 View Post

This sounds familar to the frustration I had at setting Color and Tint controls on my Panasonic TH-42PZ85. I finally found a calibrated (using HCFR software and X-rite Display LT meter) result that was very pleasing and realistic for all colors and especially for the face or skin tones. The details are in this post:

Methods to Optimize Color and Tint Settings without CMS Controls (http://www.avsforum.com/avs-vb/showthread.php?t=1188417)

If you show your CIE Diagram and RED target info I may get you an alternative method to try out. Basically involves setting the Color control for Red then adjusting Yellow for the best balance on your gamut for your TV.

Thanks, Plasma, that's a very impressive thread. I followed a bit before - I'll have to give it another read. I'll post more info soon!

-Greg
angryht is offline  
post #9 of 67 Old 01-05-2010, 06:19 PM
AVS Club Gold
 
Bear5k's Avatar
 
Join Date: Feb 2006
Location: Houston, TX
Posts: 2,239
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by angryht View Post

By 'accurate', do you mean the secondaries line up with the targets, from my specific gamut, when the saturation and hue controls are in the neutral position with no color or tint adjustment and prior to my grayscale adjustment?

You can do the grayscale adjustment. You just need to use your current actual gamut and actual white point, whatever it is.

Quote:


I guess I don't know the answer to that, yet. I have noticed that the lines between my primaries and the secondaries do not intersect my new grayscale, when adjusted. In order to adjust the saturation control to get the red brightness correct, I increased it from neutral of 25 to 29. To get cyan on the target, I adjusted if from the neutral of 25 to 21. So maybe I need to start from scratch.

A good idea if you have the ability to get a decent gamut measurement.

Color accuracy evangelist and CalMAN insider
Bear5k is offline  
post #10 of 67 Old 01-05-2010, 06:21 PM
AVS Club Gold
 
Bear5k's Avatar
 
Join Date: Feb 2006
Location: Houston, TX
Posts: 2,239
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by angryht View Post

You are right. I don't have anything close to a CMS. In my initial post I was refering to the color control, which Westinghouse calls saturation. I understand it doesn't actually control saturation, just a normal color control. Although, it does slightly move the location on the CIE chart when adjusted (as do other cheap sets).

Notionally the saturation control does control saturation because it constrains the excursion of the chroma channels.

Color accuracy evangelist and CalMAN insider
Bear5k is offline  
post #11 of 67 Old 01-06-2010, 04:37 AM
Member
 
Plasma54321's Avatar
 
Join Date: Sep 2008
Location: Florida
Posts: 164
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by angryht View Post

- I'll have to give it another read.

The thread mentioned (above) has 5 different methods for adjusting Color/Tint controls. The last or Fifth method gives the optimum results for skin tones and overall color accuracy.

Fifth Method for Setting Color/Tint Controls – Preventing RGB (>255) Red Clipping (see http://www.avsforum.com/avs-vb/showt...8#post17635708 )
Plasma54321 is offline  
post #12 of 67 Old 01-07-2010, 09:20 AM - Thread Starter
AVS Special Member
 
angryht's Avatar
 
Join Date: Aug 2004
Location: Omaha, NE
Posts: 1,321
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 11
Quote:
Originally Posted by Plasma54321 View Post

The thread mentioned (above) has 5 different methods for adjusting Color/Tint controls. The last or Fifth method gives the optimum results for skin tones and overall color accuracy.

Fifth Method for Setting Color/Tint Controls - Preventing RGB (>255) Red Clipping (see http://www.avsforum.com/avs-vb/showt...8#post17635708 )

I didn't get a chance to post my measurements yet. I hope to soon. I'm not sure I understand your method to calculate the luminance of red (in the example you've provided). Couldn't I just use either the accupel calculator (or other spreadsheet) to determine the correct luminance (and secondary locations) based on my primaries, which is what I usually do. I think I need to read the entire thread.

-Greg
angryht is offline  
post #13 of 67 Old 01-08-2010, 05:49 AM
Member
 
Plasma54321's Avatar
 
Join Date: Sep 2008
Location: Florida
Posts: 164
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by angryht View Post

... I'm not sure I understand your method to calculate the luminance of red (in the example you've provided). Couldn't I just use either the accupel calculator (or other spreadsheet) to determine the correct luminance (and secondary locations) based on my primaries, which is what I usually do.

The 5th Method is the final (and best - most accurate) result of the various alternative methods I tried for my plasma TV. In several of the charts the Color=42 is the result of using the Accupel (or similar calculators) software. I found that for my TV this had an exaggerated color level especially for red and fleshtones. This lead me to pursue and develop the final (5th) method using Bruce Lindbloom's calculator to get a target value that prevents RGB data values from clipping < 255. This lower target level (Color=36) is based on the fact that my TV exceeded the Rec 709 spec on the CIE Diagram and has an oversaturated Red primary. The lower RED taget is moving my TV's 3D surface onto the Rec 709 spec surface except for the corner that is over saturated (outside of the 709 triangle area). The oversatured color can not be adjusted (desaturated) because this TV does not have CMS controls, therefore I have to live with that but most of the surface is now accurate.

Basically, a two step process.

1) Using your TV's RED = x,y color location use Bruce's calculator to calculate a specific Y = luminance level target with RGB at 255. Adjust COLOR control to your calculated target.

2) Adjust TINT control to get Yellow HUE (rec 709) to a minimum error = 0 using a software calculator OR balance Red and Green RGB levels in HCFR to equal each other resulting in a Yellow HUE error that is minimized.

Hope this makes things more clear.
Plasma54321 is offline  
post #14 of 67 Old 01-08-2010, 08:37 AM - Thread Starter
AVS Special Member
 
angryht's Avatar
 
Join Date: Aug 2004
Location: Omaha, NE
Posts: 1,321
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 11
Quote:
Originally Posted by Plasma54321 View Post

1) Using your TV's RED = x,y color location use Bruce's calculator to calculate a specific Y = luminance level target with RGB at 255. Adjust COLOR control to your calculated target.

I see. I think basically that is similar to the method I've used previously. The only difference is that I typically use the color calculator (color corrector tab) here based on my primaries: http://www.avsforum.com/avs-vb/showthread.php?t=1111066 then adjust my secondaries to the new targets. This seems to be the same approach.

I took a look at a previous attempt to calibrate this Westinghouse TV and I think I may have been going down the wrong path. When I increased my color control to 29 the primaries did move slightly. This through off my secondaries. When I reduced the color to 22, the secondaries lined up much better. Skin tones look better now but with the game on last night, I didn't get a chance to measure. I think the points you've made in your posts are extremely helpful. Thanks again.

-Greg
angryht is offline  
post #15 of 67 Old 01-08-2010, 01:19 PM
Member
 
Plasma54321's Avatar
 
Join Date: Sep 2008
Location: Florida
Posts: 164
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by angryht View Post

... I think basically that is similar to the method I've used previously. The only difference is that I typically use the color calculator (color corrector tab) based on my primaries.

The basic method is the same although there are a couple differences. One difference is in the actual Red taget value calculated. Tom's Calibration thread suggests the Red Rec 709 target = 21% of White or using his calculator value that can give a higher target value for an oversaturated Red.

Rec. 709 x y Y
Red 0.6400 0.3300 0.2126

=> Y= 21.26% of White

The method that I am suggesting is: if you have a measured oversaturated Red, it will result in a target value slightly less than the Rec 709 target based on Bruce's calculator results.

The second difference is for setting the secondary colors and the Tint control. I say focus on the Yellow and eliminate the HUE error by matching Red and Green RGB levels in HCFR. Tom suggests setting the Cyan target for minimum error or all the dEs.
Plasma54321 is offline  
post #16 of 67 Old 01-09-2010, 05:49 AM - Thread Starter
AVS Special Member
 
angryht's Avatar
 
Join Date: Aug 2004
Location: Omaha, NE
Posts: 1,321
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 11
Attached are the readings that show the range of color and tint (saturation and hue) values. Please take a look. I am not sure which to choose. The one with color at 28 and hue at 26 seems to get the closest to the target red value and minimized the yellow error. Any thoughts would be appreciated.

Thanks.

 

Color and Tint.zip 2.79296875k . file

-Greg
angryht is offline  
post #17 of 67 Old 01-09-2010, 05:54 AM - Thread Starter
AVS Special Member
 
angryht's Avatar
 
Join Date: Aug 2004
Location: Omaha, NE
Posts: 1,321
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 11
For reference, here are my calculated primary brightness and secondaries values (based on my primaries).

 

color28 tine26.zip 7.5830078125k . file

-Greg
angryht is offline  
post #18 of 67 Old 01-10-2010, 11:20 AM
Member
 
Plasma54321's Avatar
 
Join Date: Sep 2008
Location: Florida
Posts: 164
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by angryht View Post

I've been trying to get the colors right on my Westinghouse LCD (LTV-40w1 HDC) using HCFR and a I1 LT colorimeter. It doesn't have CMS so I am limited to saturation and hue controls. Grayscale has been adjusted to fairly close to D65. Based on Tom Huffman's guide, I adjusted the saturation so that the red brightness is very close to the target (adjusted slightly based on my specific gamut). I then adjusted the hue control to get the cyan aligned with the target. This adjustment rotated all of the secondaries counterclockwise. Prior to the hue adjustment, the magenta was very close to the target but the yellow was already on the green side. After the hue adjustment, the cyan was spot on but yellow was even closer to green and magenta was shifted closer to red. And when I put on some actual content with faces (skintones) - it looked horrendous!! Everybody looked sunburn (red push?) with a little bit of a greenish (sickly) look. Any suggestions?

Well, I looked at your Excel and HCFR data files for your TV. I used Bruce Lindbloom's CIE Color calculator to get an optimum Y=luminance value for the RED primary color. I put the combined data into the table (below) with an example of the Excel Red x,y data in Bruce's calculator. The table rows contain the 3 color data runs: HCFR x,y,Y color coordinates, dE for Red and Yellow colors, Yellow's Red and Green RGB (%) and Bruce's RGB calculated values. The first row is the Rec 709 standard data for Red.

The Bruce Y row is calculated by entering Red x,y data and guessing at Y values to make R=255. This is the best Y value to prevent clipping for the color you entered, in this case Red.

So, looking at the Color data, the C=28 H=26 run is clipping at R=263 or 264 with Y=23.3% of White. Therefore, I would reduce it slightly to the Y=21.61% value close to the Rec 709 standard of 21.26% but still above it. This increased luminance is probably because Red is under saturated (located inside the CIE Rec 709 reference triangle) for your TV.

The Tint Hue RG data for Yellow is well balanced for C=28 H=26 @ 108-107% for R and G levels respectively. The C=25 and H=25 Tint values of R=106% and G=107% are also good. So, H=25 or 26 is going to be right depending on the Y change effect previously mentioned (above).

I hope this review helps. I would be interested in a subjective assessment of these implemented suggestions with a before and after comparison of real video material.

Plasma54321 is offline  
post #19 of 67 Old 01-10-2010, 01:08 PM - Thread Starter
AVS Special Member
 
angryht's Avatar
 
Join Date: Aug 2004
Location: Omaha, NE
Posts: 1,321
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 11
OK, I think I'm starting to understand, but this kind of changes everything. I have been using the spreadsheet, which gives the same result as the Accupel calculator. I thought the idea was to calculate a new Y value based on the primaries but your method is different. It seems like a combination of calculating a new gamut and trying to get closer to the 709 standard.

-Greg
angryht is offline  
post #20 of 67 Old 01-10-2010, 02:14 PM
Member
 
Plasma54321's Avatar
 
Join Date: Sep 2008
Location: Florida
Posts: 164
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by angryht View Post

It seems like a combination of calculating a new gamut and trying to get closer to the 709 standard.

I think the overall objective is to get the flesh tones (colors - red, yellow, orange, tans, browns) as close to the Rec 709 standard as possible with only simple COLOR/TINT controls. Therefore, this focuses the attention that most people find objectionable - the men from Mars or George Hamilton tans for everyone observation.

I would like your subjective review of your results for natural, rich and vivid colors after you are done. Good Luck.
Plasma54321 is offline  
post #21 of 67 Old 01-10-2010, 03:45 PM
AVS Club Gold
 
Bear5k's Avatar
 
Join Date: Feb 2006
Location: Houston, TX
Posts: 2,239
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by Plasma54321 View Post

Therefore, this focuses the attention that most people find objectionable - the men from Mars or George Hamilton tans for everyone observation.

Pick your content wisely for some of these tests. In some cases, you get people looking "tan and healthy" that are supposed to be looking "wan and sickly". You need to have an idea of what the director/DP were trying to do with a given shot to understand where in the gamut the fleshtones are supposed to fall.

Color accuracy evangelist and CalMAN insider
Bear5k is offline  
post #22 of 67 Old 01-11-2010, 10:52 AM - Thread Starter
AVS Special Member
 
angryht's Avatar
 
Join Date: Aug 2004
Location: Omaha, NE
Posts: 1,321
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 11
Quote:
Originally Posted by Bear5k View Post

Pick your content wisely for some of these tests. In some cases, you get people looking "tan and healthy" that are supposed to be looking "wan and sickly". You need to have an idea of what the director/DP were trying to do with a given shot to understand where in the gamut the fleshtones are supposed to fall.

Any suggestions? My first thought is to use House of Flying Daggers for blu-ray (since I own it). For SD content, the one that comes to mind is the woman's face in Video Essentials. I don't have 5th Element, but it looks like that gets used a lot.

-Greg
angryht is offline  
post #23 of 67 Old 01-11-2010, 10:55 AM - Thread Starter
AVS Special Member
 
angryht's Avatar
 
Join Date: Aug 2004
Location: Omaha, NE
Posts: 1,321
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 11
Quote:
Originally Posted by Plasma54321 View Post

I think the overall objective is to get the flesh tones (colors - red, yellow, orange, tans, browns) as close to the Rec 709 standard as possible with only simple COLOR/TINT controls. Therefore, this focuses the attention that most people find objectionable - the men from Mars or George Hamilton tans for everyone observation.

I would like your subjective review of your results for natural, rich and vivid colors after you are done. Good Luck.

I will post more results as soon as I can. Thank you for running those numbers, I appreciate the help.

-Greg
angryht is offline  
post #24 of 67 Old 01-12-2010, 08:51 AM
Member
 
Plasma54321's Avatar
 
Join Date: Sep 2008
Location: Florida
Posts: 164
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
I was looking at your HCFR Color Data files again. The white point and grayscale data look very good. I was wondering why your Green primary luminance was so much higher than the others (C=28 H=26 file Greeen => +56%). It also seems to push the Cyan secondary up to delta Luma = +34%.

I wonder if there are any hardware/connection issues. When you ran tests did you use the built-in DVD or an external DVD player? What type of connections HDMI or component? What was your test setup? Of course there is always the possiblity that it was designed with exaggerated Greens - "green push".
Plasma54321 is offline  
post #25 of 67 Old 01-12-2010, 09:22 AM - Thread Starter
AVS Special Member
 
angryht's Avatar
 
Join Date: Aug 2004
Location: Omaha, NE
Posts: 1,321
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 11
That's a great point! My setup is PS3 via HDMI. I was so focused on the red that I didn't pay much attention to the other colors (a duh moment). I can't look at my HCFR files (work computer) until later but looking at the excel file - it's pretty obvious that green is way far out with my color 28 and tint 26 value. How far off is green with the color at 22? I'll have to check when I get home tonight. I probably need to reduce color and compromise between errors in green and red.

-Greg
angryht is offline  
post #26 of 67 Old 01-12-2010, 09:39 AM
Member
 
Plasma54321's Avatar
 
Join Date: Sep 2008
Location: Florida
Posts: 164
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by angryht View Post

...I probably need to reduce color and compromise between errors in green and red.

Are there PS3 controls that can effect things? Set them to a neutural postion or make sure that it is not negatively affecting PQ.

I would still try the original Red Luminance level as previously suggested. Then if you find the PQ is not acceptable try something else. I would like to see your HCFR data files of your tests with some subjective viewing notes.
Plasma54321 is offline  
post #27 of 67 Old 01-12-2010, 02:53 PM - Thread Starter
AVS Special Member
 
angryht's Avatar
 
Join Date: Aug 2004
Location: Omaha, NE
Posts: 1,321
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 11
Quote:
Originally Posted by Plasma54321 View Post

Well, I looked at your Excel and HCFR data files for your TV. I used Bruce Lindbloom's CIE Color calculator to get an optimum Y=luminance value for the RED primary color. I put the combined data into the table (below) with an example of the Excel Red x,y data in Bruce's calculator. The table rows contain the 3 color data runs: HCFR x,y,Y color coordinates, dE for Red and Yellow colors, Yellow's Red and Green RGB (%) and Bruce's RGB calculated values. The first row is the Rec 709 standard data for Red.

The Bruce Y row is calculated by entering Red x,y data and guessing at Y values to make R=255. This is the best Y value to prevent clipping for the color you entered, in this case Red.

So, looking at the Color data, the C=28 H=26 run is clipping at R=263 or 264 with Y=23.3% of White. Therefore, I would reduce it slightly to the Y=21.61% value close to the Rec 709 standard of 21.26% but still above it. This increased luminance is probably because Red is under saturated (located inside the CIE Rec 709 reference triangle) for your TV.

The Tint Hue RG data for Yellow is well balanced for C=28 H=26 @ 108-107% for R and G levels respectively. The C=25 and H=25 Tint values of R=106% and G=107% are also good. So, H=25 or 26 is going to be right depending on the Y change effect previously mentioned (above).

I hope this review helps. I would be interested in a subjective assessment of these implemented suggestions with a before and after comparison of real video material.


I don't understand how you get the RG data. Is it just the percent above the Rec 709 target. Is the goal to have them balanced, and does that imply that the yellow is correct? It seems like the conclusion is to adjust the brightness of red to land on the 709 gamut (in the 3rd dimension), then to adjust the secondaries based on the actual primaries. Is that right? I think I need to read your entire post linked above.

-Greg
angryht is offline  
post #28 of 67 Old 01-12-2010, 03:19 PM
Member
 
Plasma54321's Avatar
 
Join Date: Sep 2008
Location: Florida
Posts: 164
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
The RG data or Y-Red and Y-Green are just the RGB Levels pictured in the HCFR software on the left hand side above the target picture of the selected Color Data in this case Yellow.
Plasma54321 is offline  
post #29 of 67 Old 01-13-2010, 06:50 AM - Thread Starter
AVS Special Member
 
angryht's Avatar
 
Join Date: Aug 2004
Location: Omaha, NE
Posts: 1,321
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 11
Quote:
Originally Posted by Plasma54321 View Post

The RG data or Y-Red and Y-Green are just the RGB Levels pictured in the HCFR software on the left hand side above the target picture of the selected Color Data in this case Yellow.

Oh, I see now. But, that's the deviation from the Rec 709 target, right? Wouldn't it be more relevant to determine the deviation from a calculated secondary, e.g. from a color calculator or spreadsheet, when the primaries are off? I also noticed that if I check the box to edit the color values and change the Y value for red so that the R value in the table below matches the Y in the white reading, the number is the same as your calculated number. Specifically, when I multiply 0.2157 (the Y value for red calculated using) times the Y value of white and then put that number in for the red Y then the R down below matches the Y in white. Is this doing the same thing as the Lindbloom calculator? I was all set up to do some more testing last night when tragedy struck. A CD-RW with family photos got formated/deleted and I spent a few hours trying to recover the files. So on a complete side note, does anybody recommend any software to recover deleted files. Thanks.

-Greg
angryht is offline  
post #30 of 67 Old 01-13-2010, 08:15 AM
Member
 
Plasma54321's Avatar
 
Join Date: Sep 2008
Location: Florida
Posts: 164
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by angryht View Post

... that's the deviation from the Rec 709 target, right?

Everything is referenced to the Rec 709 standard for both Red and Yellow colors.

Quote:


I also noticed that if I check the box to edit the color values and change the Y value for red so that the R value in the table below matches the Y in the white reading, the number is the same as your calculated number. Specifically, when I multiply 0.2157 (the Y value for red calculated using) times the Y value of white and then put that number in for the red Y then the R down below matches the Y in white. Is this doing the same thing as the Lindbloom calculator?

You lost me on this. The table above has more in it than you need. There are both COLOR luminance values using RGB [0..255] (Bruce's calculator data) and the TINT hue HCFR data in the last two columns the Yellow-Red and Yellow-Green % contributions.

The Lindbloom Calculator is to get a specific Y=Luminance value (by trial and error) for the RGB value of RED=255 with your Red x,y measured values.

The Y-Red and Y-Green are HCFR's RGB % Levels for adjusting TINT. They should be balanced (the SAME amount of Red and Green) to get a YELLOW color with no Hue error. In your case with your data they are matched. I found this technique to be the easiest way to set the TINT control and get the Yellow Hue accurate.
Plasma54321 is offline  
Reply Display Calibration

User Tag List

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off