yCMS - Color Management System

[cyberbeing]

@MSL_DK

Tiny changes in XYZ values are causing yCMS to produce 3DLUTs with drastically different color luminance clipping results. If you have better results using measurements from one source compared to another, it's just a fluke. This is likely the same madVR forced TV range on PC range test pattern bug which has existed since early yCMS versions.

[MSL_DK]

Are you taking the measurements by running color test patterns through madVR?

...If you use madVR, you should set madVR to "this display is not calibrated" when you display the color test patterns.


I have done so far. That's how I have achieved these strange results where red is cut.

Or are you using some other software to display the test patterns?


Here I have installed HCFR (http://forum.doom9.org/showpost.php?p=1604794&postcount=799) on my htpc and run "show images" and enter the values ​​in madvr.

[MSL_DK]

@MSL_DK

Tiny changes in XYZ values are causing yCMS to produce 3DLUTs with drastically different color luminance clipping results. If you have better results using measurements from one source compared to another, it's just a fluke. This is likely the same madVR forced TV range on PC range test pattern bug which has existed since early yCMS versions.

Refer you to this post?

I am somewhat confused and not sure I am able to explain to me technically. I have just a few questions. My gfx is set to full rgb madvr to 0-255 and my TV is set to (EDIT) PC levels. Before ycms will rgb clip 235 and after ycms at 251, why?

My settings are of course:

GFX: 0-255
madVR: 0-255
TV: PC levels (0-255)

:o

[cyberbeing]

My settings are of course:

GFX: 0-255
madVR: 0-255
TV: PC levels (0-255)

That test pattern is PC Range video, flagged as TV Range.

So for your setup...

madVR input range: 16-235
madVR output range: 0-255
GFX: 0-255
TV: 0-255

yCMS has a problem maintaining identical luminance range between Primary colors when madVR treats PC range video as TV range with BTB/WTW, and then expands back to 0-255 (i.e. clipping test patterns).

[MSL_DK]

That test pattern is PC Range video, flagged as TV Range.

So for your setup...

madVR input range: 16-235
madVR output range: 0-255
GFX: 0-255
TV: 0-255

yCMS has a problem maintaining identical luminance range between Primary colors when madVR treats PC range video as TV range with BTB/WTW, and then expands back to 0-255 (i.e. clipping test patterns).

Okay ... I think I understand.

This is not my goal, but I have tried with

madVR: 16-235
GFX: 0-255
TV: 16-235

This does not solve the problem. I still get this weird result (https://dl.dropbox.com/u/111324524/Images%20for%20Sharing/ycms/after/2.JPG)

[yesgrey]

I'm following, and will look into this carefully next week, when my Christmas holiday starts.

Thanks for reporting the problems.

[MSL_DK]

I'm following, and will look into this carefully next week, when my Christmas holiday starts.

Thanks for reporting the problems.

Thanks and Merry Christmas :-)

[fallengt]

He is using CalMAN v5 there, which also takes those measurements. (as you can see from the image he posted)

The issue is that yCMS only allows you to specify primaries and greyscale points, there is no option to create a custom LUT to fix hue/saturation errors below 100%.

Most displays are far from linear below 100%.
+1
I would be nice if yCMS supported gamut correction below 100% sat.

[MSL_DK]

I'm following, and will look into this carefully next week, when my Christmas holiday starts.

Thanks for reporting the problems.

Hi ... How do things look?

[n3w813]

+1
I would be nice if yCMS supported gamut correction below 100% sat.

+1

It would be great if you can add support to input 3dLUT to correct the entire color gamut, not just the edges. :)

Maybe a 5x5x5 (125-pt) 3DLUT correction?? ;)

[hulkss]

I posted some instructions that work for me over at the JRiver forum: How To Get Perfect Video Color (http://yabb.jriver.com/interact/index.php?topic=78302.0)

[n3w813]

I posted some instructions that work for me over at the JRiver forum: How To Get Perfect Video Color (http://yabb.jriver.com/interact/index.php?topic=78302.0)

Although this method does correct the hue and saturation of the color gamut below 100% saturation, it does NOT produce correct luminance values(Y) for the points. Average DeltaL are increased anywhere from 5 to 20+.

[browser]

+1
I would be nice if yCMS supported gamut correction below 100% sat.

+1

Also a gamma per color correction would be very nice !:D

[TomArrow]

Bit late to the party, does yCMS still exist? The download links no longer work but I'd really like to use this with madVR to make things look right on my display.

[Asmodian]

yCMS was never a great solution, it worked well for mapping a larger gamut to a smaller one but it is much better to create a 3DLUT with DisplayCal (also free) or similar.

[TomArrow]

yCMS was never a great solution, it worked well for mapping a larger gamut to a smaller one but it is much better to create a 3DLUT with DisplayCal (also free) or similar.

Sadly I only have a finished ICC profile available. Can I convert that to a 3D LUT somehow?

[huhn]

there is a chance to do that sadly this may not work.

you can try the 3d LUT maker from displaycal. you load the icm file as a destination file.