[argyllcms] Re: Argyll and 30-bit colors

  • From: Marwan Daar <marwan.daar@xxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Thu, 11 Aug 2016 12:42:02 -0400

The output color depth option depends on the GPU and software environment. I certainly don't have that option, but that could well be because I'm using an analog monitor. I've also corresponded with an Nvidia rep through their customer support channel, and they seem to be clueless about this new feature.

Interesting, thanks for the link. I also didn't realize you can just set your output color depth in the Nvidia control panel! Are dispcal and dispread able to take advantage of that when they display test patches, or are they still limited to 8-bit for that?

On Thu, Aug 11, 2016 at 3:29 PM, Marwan Daar <marwan.daar@xxxxxxxxx <mailto:marwan.daar@xxxxxxxxx>> wrote:

    My experimentation suggests that Windows desktop can use up to at
    least 10 bits of the 16 bit values. (nvidia GPU and CRT). Whether
    this is achieved via dithering or not is a separate question.
    Also, while 10 bit precision can be achieved, it is still only in
    the context of displaying 256 colors simultaneously.

    Also see this:
    
https://hardforum.com/threads/nvidia-or-amd-for-color-accuracy-and-calibration-features.1873362/page-4#post-1042471971
    
<https://hardforum.com/threads/nvidia-or-amd-for-color-accuracy-and-calibration-features.1873362/page-4#post-1042471971>


        IIRC Windows does support 10-bit color, but only in the
        fullscreen exclusive mode of various hardware acceleration
        APIs like Direct3D and OpenGL. I heard rumors they were
        working on enabling 10-bit color for the desktop, but I have
        no idea whether that's true or what the timeline for such a
        change would be if it is.

        One thing I have wondered is whether anything actually uses
        more than 8 bits of the 16-bit RAMDAC values that Windows
        uses. I was thinking of modifying dispcal locally to set the
        LUT directly instead of modifying the color of the test
        patches, to see if that gives it access to more than 8 bits on
        my 10-bit monitor - but that would be a pretty major change
        (potentially making the whole process more direct), and I
        haven't even checked yet whether dispcal can easily access the
        LUT setting functions of dispwin.





Other related posts: