[argyllcms] Re: Black-/White-Level Observer Dependency?

  • From: János, Tóth F. <janos666@xxxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Mon, 27 Sep 2010 03:21:20 +0200

No. I think we should use it for the whole calibration. If we need it to set
the brightest white to a real white, then we need it to achieve a consistent
and correct white balance through the whole gray ramp too.

But this is the problem. yCMS works well only when you already have a
calibrated display. If the WP, WB and gamma is right, the gamut correction
is very great. If you have an offset white point or strange gamma curve, it
can't really correct it (yet) and the gamut will be drifted as well.

-You can't insert XYZ.10 values for this software but the XYZ.2 values will
show a drifted WP, so the gamut emulation will be drifted. (Fals chromatic
-If the software will be able to correct the white balance too, but you feed
it with XYZ.2 values, then the final result will be a subjectively drifted
WP. (Like you never used FOV10 calibration.)

I think the only solution if the softwares will support this observer type.
(So, they will accept XYZ.10 values and they will also work in the CIE 1964
color space.)

The calibration itself is promising. The FOV10 based calibration feels a
little reddish but the FOV2 calibration looks much more greenish when I
quickly switch between them.
Here are my settings and the graphs of the calibration LUTs:
[The descriptions on the images are false. The first is the FOV2 and the
second is the FOV10 calibration with the same hardware settings: RGB Gains
at 255-255-255)

Other related posts: