AVS Forum banner

What format would you prefer for the HTPC?

  • .mp4

    Votes: 64 55.2%
  • .mkv

    Votes: 52 44.8%
  • others (please detail in thread)

    Votes: 0 0.0%

GCD - Gamut Calibration Disk

96K views 632 replies 68 participants last post by  zoyd 
#1 ·
As I think most calibrations disk available today lack in-debt calibrations patterns for Gamut calibration, I decided to create one my self, the goal with this is not to create a disk for full calibration, as there is plenty of disks available for this, its only for gamut calibration.

b4.3.13 is the latest beta version of the disk in AVCHD format (This need burning to a DVD)

b4.3.13 is the latest beta version of the disk in Blu Ray format (This need burning to a BD)

b4.3.13 is the latest beta version of the disk in mp4 format (This can be used from a HTPC)

if any problems please shout up

b4.3.13 is the latest beta version of the disk in .mkv format

a5.0.0 Only contain CALMAN patterns and the basic patterns - More will follow.



Just a quick guide to what format to chose:


Use the Blu-Ray version if you wish to burn to a Blu-Ray disk (BD25 or DB50) - use this format if your blu-ray player dont support 1080p/24fps via AVCHD (reports suggest that panasonic players dont)



Use the AVCHD in when burning to a DVD but blu ray playback



fix list:


known bugs:


None


100% luminance - Saturation Sweep

75% luminance - Saturation Sweep\t\t\t\t

50% luminance - Saturation Sweep\t\t\t\t

25% luminance - Saturation Sweep\t\t\t\t

\t\t\t\t

100% Saturation - luminance Sweep\t\t\t\t

75% Saturation - luminance Sweep\t\t\t\t

50% Saturation - luminance Sweep\t\t\t\t

25% Saturation - luminance Sweep


100% luminance - 100% Saturation: Primaries/Secondaries Sweep\t

100% luminance - 75% Saturation: Primaries/Secondaries Sweep\t

100% luminance - 50% Saturation: Primaries/Secondaries Sweep\t


75% luminance - 100% Saturation: Primaries/Secondaries Sweep\t

75% luminance - 75% Saturation: Primaries/Secondaries Sweep\t

75% luminance - 50% Saturation: Primaries/Secondaries Sweep\t


50% luminance - 100% Saturation: Primaries/Secondaries Sweep\t

50% luminance - 75% Saturation: Primaries/Secondaries Sweep\t

50% luminance - 50% Saturation: Primaries/Secondaries Sweep


Grey Scale Patterns 5 & 10% steps

Clip Pattern

Color Test Patterns


Reference pictures



Future Plans:


MP4 version


Others??


attached is the xyY targets for all patterns. Big thanks to Zoyd for his big work in creating this disk.
GCD_targets_v13.xlsx 254k .xlsx file

Thanks alluringreality & dr1394 for helping with utility's to make this possible
 

Attachments

See less See more
#2 ·
Quote:
Originally Posted by zoyd
I'll give one example, red at 75% saturation, 100% luminance


Step 1: red(709) x=.640, y=.330, Y=21.26, white(709) x=.3127, y=.3290, Y=100


75% saturation x= .75*(redx-whitex)+whitex = .5582

75% saturation y= .75*(redy-******)+****** = .3298


Step 2: Convert x,y,Y to X Y Z using this formula


Step 3: Convert X Y Z to R G B (0-1) using this formula


Step 4: Convert RGB(0-255) to RGB(16-235): R(0-255)=R(0-255)/255*219+16


Step 5: round result to nearest integer
to be completely honest this is over my head

I can however follow the example and I also get the same result as you on the 2 examples I have done.


could you please have a quick look at attached sheet with the rgb% from tom (again I assume the gamma is 2.22) - could the difference of a gamma at 2.2 and 2.22 really cause the difference on up to 6 (as it is for some 50% luminance?)


unfortunately I cannot help a lot on this - I just assumed what I was given from Tom Huffman was correct. It would be interesting to hear Tom's take on this.


Did you manually calculate all the RGB values or do you have a spreadsheet with your calculations?

 

specs(TOM H).zip 6.169921875k . file
 

Attachments

#3 ·

Quote:
Originally Posted by visca blaugrana /forum/post/21927634


to be completely honest this is over my head

I can however follow the example and I also get the same result as you on the 2 examples I have done.


could you please have a quick look at attached sheet with the rgb% from tom (again I assume the gamma is 2.22) - could the difference of a gamma at 2.2 and 2.22 really cause the difference on up to 6 (as it is for some 50% luminance?)


unfortunately I cannot help a lot on this - I just assumed what I was given from Tom Huffman was correct. It would be interesting to hear Tom's take on this.


Did you manually calculate all the RGB values or do you have a spreadsheet with your calculations?

I gave you % values, not digital values.

2.22 gamma is assumed.

In PC digital 100% is 255. In video digital 100% is 235. In any case, I am not sure that this is even important. I know that when I created a disc and had to convert my % values into digital values, the software used 0-255, but the resulting disc seemed to output correctly. The more important issue is gamma, which really makes a big difference in the % values.
 
#4 ·

Quote:
Originally Posted by visca blaugrana /forum/post/21927634


to be completely honest this is over my head

I can however follow the example and I also get the same result as you on the 2 examples I have done.


could you please have a quick look at attached sheet with the rgb% from tom (again I assume the gamma is 2.22) - could the difference of a gamma at 2.2 and 2.22 really cause the difference on up to 6 (as it is for some 50% luminance?)


unfortunately I cannot help a lot on this - I just assumed what I was given from Tom Huffman was correct. It would be interesting to hear Tom's take on this.


Did you manually calculate all the RGB values or do you have a spreadsheet with your calculations?

It's not a problem with gamma, the differences are random and most quite small. It looks more like rounding problems. So unless there is something wrong with the formulas I linked to I don't know where the differences come from. I ported the formulas to IDL (somewhat like MATLAB software) and generated all the values in high precision and then rounded.
 
#6 ·

Quote:
Originally Posted by ams2990 /forum/post/21929332


The formula in step 3 is for sRGB gamma, not 2.22 gamma. They're similar, but different.

Thanks for catching that! I recalculated with the proper companding and the agreement is much better although not perfect yet.
 
#7 ·
Here is the revised comparison with gamma=2.22 and proper gamma companding. I also used the more precision conversion matrix from brucelindbloom.com rather than the one at easyrgb. The 100% saturation codes at all luminances agree exactly with Tom's, there are still discrepancies at 75% 50%, and 25% saturations that increase as luminance decreases. I suspect the difference is in how we calculated the x,y positions of the reduced saturation points. Tom, can you describe how you came up with your percentages?

 

amp-sat (All)_zoydcomparisonB.xls.zip 17.525390625k . file
 

Attachments

#8 ·

Quote:
Originally Posted by zoyd /forum/post/21929921


Here is the revised comparison with gamma=2.22 and proper gamma companding. I also used the more precision conversion matrix from brucelindbloom.com rather than the one at easyrgb. The 100% saturation codes at all luminances agree exactly with Tom's, there are still discrepancies at 75% 50%, and 25% saturations that increase as luminance decreases. I suspect the difference is in how we calculated the x,y positions of the reduced saturation points. Tom, can you describe how you came up with your percentages?

What discrepancies? They look the same to me.
 
#9 ·

Quote:
Originally Posted by TomHuffman /forum/post/21930396


What discrepancies? They look the same to me.

they aren't large:


max difference at 100% luminance: magenta @25% saturation, difference = 2 codes


75% luminance: several with 2 code difference


50% luminance: several with 4 code difference
 
#10 ·

Quote:
Originally Posted by zoyd /forum/post/21930452


they aren't large:


max difference at 100% luminance: magenta @25% saturation, difference = 2 codes


75% luminance: several with 2 code difference


50% luminance: several with 4 code difference

We must be looking at something different. Magenta 25% saturation, 100% luminance, we both have

R 64%

G 53%

B 64%
 
#11 ·
If you are looking at the last spreadsheet I attached, your values are in percent at the top. These were used to calculated 16-235 codes by visca blaugrana (also in top half). The bottom half contain my calculated codes along with the differences in RGB codes labeled "zoyd-visca blaugrana".
 
#12 ·

Quote:
Originally Posted by zoyd /forum/post/21930959


If you are looking at the last spreadsheet I attached, your values are in percent at the top. These were used to calculated 16-235 codes by visca blaugrana (also in top half). The bottom half contain my calculated codes along with the differences in RGB codes labeled "zoyd-visca blaugrana".

All that I can vouch for is the % values. These are correct. Regarding the conversion between % values and digital values, because--given the PC/Video range distinction--it is unclear to me what the proper methodology for this is. Indeed, given how different software, video cards, and even displays interpret this data, it is not even clear to that there is ONE correct methodology for this.
 
#14 ·

Quote:
Originally Posted by TomHuffman /forum/post/21931172


All that I can vouch for is the % values. These are correct.

I don't doubt that you did the xyY->RGB correctly but without those values I can't trace back and confirm that we are using the same starting point for conversion to RGB. Can you provide the xyY values, maybe for 2 or 3 that show the largest difference. As I mentioned earlier we are in exact agreement for all 100% saturation points and differ at
 
#15 ·
Here's red

% SaturationColorxyYXYZRGBR'G'B'
100%R0.64000.33000.21260.412320.212600.01933100%0%0%100%0%0%
75%R0.55820.32980.21260.359830.212600.0722080%5%5%91%27%27%
50%R0.47640.32950.21260.307380.212600.1252461%11%11%80%36%36%
25%R0.39450.32930.21260.254750.212600.1783641%16%16%67%44%44%
 
#17 ·

% SaturationColorxyYXYZRGBR'G'B'
100%G0.30000.60000.15350.076750.153500.025580%21%0%0%50%0%
75%G0.30320.53230.15350.087430.153500.047442%21%2%19%49%19%
50%G0.30640.46450.15350.101250.153500.075715%19%5%27%48%27%
25%G0.30950.39680.15350.119730.153500.1136210%18%10%35%46%35%
 
#18 ·
ok, if I use the percentages you just posted the agreement is good, if I use the ones in the spreadsheet (also yours) I get some errors (max 4 codes). The spreadsheet values for G' are 1.2 - 1.6 % higher than the ones you just posted.


This is what's in the spreadsheet for Green@50%:

R'G'B'
0.0%50.0%0.0%
19.0%50.6%18.9%
27.7%49.2%27.6%
35.7%47.2%35.7%
 
#19 ·
Quote:
Originally Posted by zoyd
ok, if I use the percentages you just posted the agreement is good, if I use the ones in the spreadsheet (also yours) I get some errors (max 4 codes). The spreadsheet values for G' are 1.2 - 1.6 % higher than the ones you just posted.


This is what's in the spreadsheet for Green@50%:

R' G' B'
0.0% 50.0% 0.0%
19.0% 50.6% 18.9%
27.7% 49.2% 27.6%
35.7% 47.2% 35.7%
I think this is a combination of rounding errors and I might have had a slightly different gamma selected previously. Here is an updated sheet.

 

sats.zip 3.1533203125k . file
 

Attachments

#22 ·

Quote:
Originally Posted by visca blaugrana /forum/post/21933428


and just as I thought it was done and dusted



there is still a few deviations (however mainly very small) one thing i wonder if its a mistake is that according to Tom the Magenta at 100% sat/lum is 100%,2%,100%


please see attached

yes, that's an error, it should be 0% as in the original spreadsheet.


So which patterns will you be making for an avchd iso? I had in mind the following 7 sequences compatible with HCFR:


1. 100% luminance saturation sweep [R0,R25,R50,R75,R100, etc.]

2. 75% luminance saturation sweep [R0,R25,R50,R75,R100, etc.]

3. 50% luminance saturation sweep [R0,R25,R50,R75,R100, etc.]

4. 100/100 RGBYCMW

5. 100/75 RGBYCMW

6. 75/100 RGBYCMW

7. 75/75 RGBYCMW
 
#24 ·

Quote:
Originally Posted by visca blaugrana /forum/post/21933563


0% Saturation? is that really used in HCFR?


if so is that just white?


this is properly a stupid question

yes, the sequence starts with 0% saturation, it's white but at the same luminance as the color that is being swept. Do you want me to post the codes for that case?
 
#26 ·

Quote:
Originally Posted by visca blaugrana /forum/post/21933688


Please.


I am in the middle of creating the patterns in photoshop at the moment, and is roughly half way through



see attached example note this is only a .jpeg and should not be used as reference.

Looks good, PS reports RGB= 75 121 75.


Here are the 0% saturation codes scaled to 16-235 assuming gamma=2.22, don't bother with the 25% luminance case.

Color/%LuminanceR'G'B'
R100125125125
R75989898
R50717171
R25434343
G100204204204
G75157157157
G50110110110
G25636363
B100838383
B75666666
B50494949
B25333333
Y100228228228
Y75175175175
Y50122122122
Y25696969
C100213213213
C75163163163
C50114114114
C25656565
M100140140140
M75109109109
M50787878
M25474747
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top