[argyllcms] Re: White Point

  • From: Elle Stone <ellestone@xxxxxxxxxxxxxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Mon, 9 Nov 2015 14:25:18 -0500

On 11/09/2015 12:42 PM, Richard Kirk wrote:

On 09 Nov 2015, at 14:29h, Elle Stone <ellestone@xxxxxxxxxxxxxxxxxxxx>

This article has interesting take regarding the various color spaces:

The paper is measuring the deltaE values against the CIE standard
observer. If they had put DCI XYZ in the table, it would have come out
with a deltaE of zero. This doesn't make it perfect. In fact, as the CIE
1931 standard observer includes a factor that makes purple look like
red+blue, you might be better off matching using the CIE2015 LMS
standard (better check the date, it has been about to come out since 2007).

Richard Kirk

I'm not sure I follow what you are saying.

My understanding is that "delta E" is measured against spectral calculations, the nature of which calculations are outside my small area of "maybe I understand something". But the calculations have to do with how light combines to produce the colors we see and measure, which isn't exactly the way colors multiply in the various RGB color spaces we use for image editing.

My interpretation could be wrong, but I think the differences shown in XYZ are to demonstrate that multiplication is a chromaticity-dependent editing operation, which undeniably it is.

If XYZ were the standard against which delta E was being measured, I think XtremeRGB would have produced a lower delta E, as XtremeRGB is basically an Identity matrix with a D50 white point (http://www.hutchcolor.com/profiles.html).

Anyway, my understanding of the Mansencal paper and also of what you were saying about the Mansencal paper, either or both could be completely wrong.


Other related posts: