At 1:15 AM -0800 11/10/05, Kratzy974 wrote:
Thanks for the answer.
The White of the D65 Spaces are rotated to be viewed in a D50 . So the 100,0,0 value is moved to 100, x, y, where x and y are the difference values between the Colortemperature. Just open a sRGB, and you see what I mean. That is fully correct, and I don’t want to argue against that.
well, this may seem strange as I designed & implemented the behavior but I do have an argument with it.
To begin, there’s some history and logic around this one. In the beginning of time I designed the gamut volume so that it would correspond directly to the measured data that built the profile. So if you graph target measurements with the resulting profile, the target colors match up with the profile volume quite nicely. Cool, makes sense, expected behavior, etc. When it came to monitor and working-space profiles I left ColorThink to behave the same way. What this means is that if the monitor profile has a non-D50 white point it maps to a non 100,0,0 white point in the graph as you described.
While this could be argued to be the correct way to do things there are also arguments against it:
to the “adapted eye” each monitor would appear to have a “white” white point and all colors would fall out relative to that white. So to the adapted eye the white points should line up.
it turns out that any white point other than D50 in monitor profiles was not intended in v2 profiles and has been “outlawed” in v4 profiles. All colors are to be adapted to D50 and the white point should be D50 from now on.
gamut comparisons should be done with the white points adapted in some manner. This will give a better indication of how two gamuts compare that when one of them has a shifted white point.
With this is mind it makes more sense to map all monitor gamuts to 100,0,0 which is exactly what I plan to do in ColorThink Pro and then eventually non-pro ColorThink.
The problem with the rel Colormetric calculation is, that if I view an Image (from sRGB) with Whitepoint away from 100,0,0. Then I use rel Colormetric to a D50 space : the grey line isn’t rotated, so the bright values are mapped to the gamut of the D50 colorspace. This looks wrong (also with perceptual…).
yes, understood. Altering the graphing so that the white points all map to D50 also introduces challenges when rendering individual color and their vectors into Lab for graphing… that’s what I’m working on right now in fact.
It is fair to say that either method could be used. In canvassing many opinions and looking into an ICC v4 future it makes most sense to change the way things are done now. I’m sure it will create some discussion along the way…
o Steve Upton CHROMiX www.chromix.com
o (hueman) 866.CHROMiX
Post generated from email list