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

  • From: Sam Berry <samkberry@xxxxxxxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Mon, 27 Sep 2010 10:24:46 +0100

On 27 September 2010 04:06, Graeme Gill <graeme@xxxxxxxxxxxxx> wrote:

> János wrote:
>> What is the situation with ICC based CMS softwares? What would happen with
>> the primaries when I create an ICC file with FOV10 observer settings and I
>> use this ICC file on usual CMS softwares which are not XYZ.10 systems?
> Technically, such a profile is not ICC compatible, since ICC defines
> that the profile connection space should use the 1931 standard observer.
> In practice you can do whatever you want in a closed system (ie. where
> you are controlling the source and destination profiles and viewing
> situations.)
> Graeme Gill.
I would also add that the calculation of daylight and blackbody points can
usually be considered outside the closed system, and so trusting any
calibration utility to set D65 using the 10 degree observer is likely to
give meaningless results (ColorNavigator excepted perhaps). The observers
only match numerically at x=y=1/3. D65 is only x,y=0.3127,0.3290 using the 2
degree observer. Using any other observer the numbers are just a random
shade of white-ish. A new set of BB and daylight curves would need to be

Also, "native white point' still performs grey balancing, just without a
preset white.

Sam Berry

Other related posts: