[argyllcms] Re: DE to locus = 17.3

  • From: Graeme Gill <graeme@xxxxxxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Sun, 11 Jan 2009 15:29:56 +1100

Lars Tore Gustavsen wrote:
Can you please tell me how far away the white point could be from one
of the locus curves, before I should drop native white point
calibration? I guess a lot of people can find that interesting to

I don't have any specific answer for this. It comes down to your
preferences as to what tradeoff you want to pick, given the behaviour
of your device, and in what way you want to use it.

If only 8 bit LUTs are available (often the case for digitally connected
displays), then any non-linear calibration curve is going to loose bits
of precision that may be visible, and you have to make a choice as to what
loss is acceptable vs. what the benefits are of being closer to the white

Typically a viewer will adapt almost completely to an emissive display
device if it dominates their field of view and is in a dim surround,
but (as I understand it) adaptation is more complete when the white
point lies close to the black body/daylight locus. If you look at the
native white of your display and are happy that it looks subjectively
white to you, and you are not concerned about side by side comparisons
with other image sources (such as another display or a print), then
there is probably no reason to try and shift the display white point.

Graeme Gill.

Other related posts: