| Forums | Polls | Register | Reply | Search | Statistics |
Display Calibration Light Illusion Forums / Display Calibration /   
 

Calibration-step: adjusting luma of the color channels

 
Author Nepomuk
User - Mono 
#1 | Posted: 17 Jan 2013 15:39 
Please help me understand this step.

Take 709 as the target, and lets say, I decide to desire 80cd/m² and a 2,35 gamma function. How do I find the Y-values for the color channels?

Thank you!

Author Steve

Mod - Scene 
#2 | Posted: 17 Jan 2013 15:42 
LUTs should not be used to control the peak luma level - set that with the display's controls.

To set a 2.35 Gamma build a new Colour Space pre-set by modifying the existing Rec709 one with the new gamma setting and save it with a new name.

Steve
Steve Shaw
Mob Boss at Light Illusion

Author Nepomuk
User - Mono 
#3 | Posted: 17 Jan 2013 19:31 
Steve

Ah, that is what I did. I have for bt709 luma (Y’) 0.2126, 0.7152, and 0.0722 as coefficients in the display's gain settings. Does anyone ever check those?

Thanks Steve!

Author Steve

Mod - Scene 
#4 | Posted: 17 Jan 2013 19:46 
I really do not know what you mean with those values???
They look to be the white balance values for D65?

If you mean the RGB Gain controls that is how you set the white point (colour temp) - The Bias controls set the colour temp for black.

I think you are confusing different numbers/values and their meanings?

Steve
Steve Shaw
Mob Boss at Light Illusion

Author Nepomuk
User - Mono 
#5 | Posted: 19 Jan 2013 09:19 
yes, I am not understanding this.

When I want 100cd/m² white at d65, I must set the red luma peak value to 21,26cd/m², correct? For other white lumas I would need to calculate by hand?

These coefficient and ones for the secondaries appear as gain-setting when I hit hdtv in the software calibration interface of my display. These values are adjustable from 0-1,99, white is set to 1, naturally. I can adjust the coordinates of white like for the other colors, while these gain coefficients don't change, and there are gain and offset settings in the service menu for RGB from -99 to 99.

Author Steve

Mod - Scene 
#6 | Posted: 19 Jan 2013 09:24 
Ummmm, no - you just use LightSpace CMS and profile the display using 'Measure' mode, making adjustments to the display until the correct readings are attained.

Or, you can do a Quick Profile, or a full Display Characterisation profile, and build a calibration LUT.

Simples!

Steve Shaw
Mob Boss at Light Illusion

Author Nepomuk
User - Mono 
#7 | Posted: 19 Jan 2013 12:58 
Thats what I did at first, Steve. I am just changing the coordinates for all colors and white, till they are perfect in 2D and the gamma is good as well, after quick profiling tests. Leaving the gain coefficients and gain untouched. But you write in the calibration manual, to adjust gain und bias and focus on 75% white, which makes me think I am not getting something here. There are only peak level controls, except the up 64 points manual adjustable RGB gamma curves, which you said one should leave alone. How adjust 75% white for example?

I am just wondering all this, since either the rgb separation or the rgb balance are always bad. Interesting is the fact, that when I use the calibration softwares hdtv-present the separation is very good, but the rgb balance are 4 parallel straight horizontal lines, but with completely different values.

Being anything but an expert, I have the feeling the device is good, just something is shifted.

Thanks for your help Steve! phenomenal support!

Author Steve

Mod - Scene 
#8 | Posted: 19 Jan 2013 13:25 
What display do you have, and what controls does it have?

Gain affects white, with less effect towards black.
Bias affects black, with less effect towards white.

So, you need to use both with care and check the full grey scale range.
The focus on 75% white is for Gain.
Focus on 25% black for Bias.

This is just to set the colour temp throughout the grey scale.

Any help?

Steve
Steve Shaw
Mob Boss at Light Illusion

Author Nepomuk
User - Mono 
#9 | Posted: 19 Jan 2013 17:33 | Edited by: Nepomuk 
Its a 3chip dark4 dlp projector with 10bit hdmi inputs and internal 12 bit processing. Haven't seen the 10 bit yet, but have only driven it in full scale over RGB yet. Maybe its YUV 10bit only. And it has a built in fixed DCI filter.

There is gain (-99 to 99) and offset (-127 to 127) for R G B in the service menu and in the PC- calibration program I can adjust P7 ( RGB CMY and white) in coordinates and another gain (rather gain-coefficients) for each P7. In that software there is a hdtv and d65 preset, were when either is activated the color coordinates and the gain-coefficients are set to 0.2126, 0.7152, and 0.0722 for RGB and corresponding CMY. White is set to 1. Of course there is gamma and brightness and contrast as well.

I only adjust the coordinates of p7 and hit easily the centers of circles in CIE graph. But, except green, red and blue have, though linear, lower than 1 slopes in the rgb separation.

Adjusting the gain in the service menu gives only very fine control, but don't see the purpose.

There is no "bias".

Interesting is that, the hdtv preset is totally off as is the d65 point and green. Could their CMS have been wrongly calibrated? The greyscale color temp is very good.

I am trying to exclude a mistake on my end, before I go back to the maker.

Your help is greately appreciated!

Author Steve

Mod - Scene 
#10 | Posted: 19 Jan 2013 17:59 
Offset is the same as Bias.

But, if you adjust the P7 controls I would indeed expect the RGB Separation to then be bad!
Any display that has control of secondary colours (CMY) immediately says the colour management is bad!
Secondary colours should be a direct calculation of the Primaries, and not require separate control - to do so will cause very bad RGB Separation.

So, overall it looks like the projector's internal colour management is very bad!

I would use Offset and gain only, with everything else turned off.
If you can do that - turn everything off and use a LUT only!!!!

Steve
Steve Shaw
Mob Boss at Light Illusion

Author Nepomuk
User - Mono 
#11 | Posted: 20 Jan 2013 18:21 
Aha! Very interesting!
I found a way to disable p7 (all set to 0 and CMY says "NaN") and reached seemingly the most native setting. Separation is perfect, even the DCI RGB circles are hit natively, though white is way in the blue and cyan and magenta are moved the same amount towards blue.
This was what I corrected with P7 and screwed up the separation.

Using gain and bias. Just setting bias to -30 let 25% hit D65. But adjusting gain has not effect, on 75% or white, what so ever.

I have to contact the manufacturer at this point. Still hope to not have to use a LUT ;)

Thank you Steve!

Author waltervolpatto
User - Mono 
#12 | Posted: 30 Jan 2013 15:47 
Nepomuk, are you using a TI Digital cinema communicator or something similar to calibrate your projector (You state that is a DLP, I'm curios to see if I can help you to calibrate better)

Author Nepomuk
User - Mono 
#13 | Posted: 1 Feb 2013 07:05 
Thanks Waltervolpatto,

but it turns out, the device is really the problem.

Author waltervolpatto
User - Mono 
#14 | Posted: 1 Feb 2013 15:09 
Nepomuk
OK....

In DLPs, some of the menus can help you to get a better calibration, but there are so many controls that it is more easy to screw it up that not the vice-cersa...


Author Nepomuk
User - Mono 
#15 | Posted: 5 Feb 2013 07:31 
Thats right, I was unsure till Steve reached the same result. The ball is on the manufacturers side. I still hope its easily fixable, but I am already looking at alternatives.

Display Calibration Light Illusion Forums / Display Calibration /
 Calibration-step: adjusting luma of the color channels

Your Reply Click this icon to move up to the quoted message

 

 (?)
Only registered users are allowed to post here. Please, enter your username/password details upon posting a message, or register first.

 

 
Online now: Guests - 7
Members - 0
Max. ever online: 264 [6 Jun 2020 12:02]
Guests - 264 / Members - 0