Graeme is correct when referring to the ability to convert using
abs col OR rel col FROM a monitor profile TO another profile. The
results will be the same because the white point difference is
removed in ICCv4 profiles. (also Photoshop now changes abs col
conversions from non-D50 white-point v2 display/working spaces to
be the same as rel col - a similar effect but not in the v2 spec)
Steve, AbsCol still isn't the same as RelCol in all respects even in
this example of yours.
For one it's something that has only just started to happen in CS2
and happens with both v2 and v4 profiles, but the results are
actually only the same between RelCol and AbsCol with respect to the
white point. Other content undergoes a difference in the conversion.
Chris is correct that you can do abs col conversions FROM a (say)
print profile to a display profile and you will receive a different
white point than had you chosen rel col.
So abs col/ rel col FROM a display profile in v4 = no diff....
No, there is still a difference with e.g. sRGB to SWOP v2 using
RelCol or AbsCol. White remains the same RelCol or AbsCol, but for
instance 0r,0g,255b converts very different between the two intents.
Also - this "partially adapted user" thing. Chris, that's the first
I've heard of it too. Do you have any references to tools that
worked this way? It's certainly not in the v2 spec....
It's true it's not in the spec yet venders did make different
assumptions about this, one reason why it use to be routine we'd see
different profiles from the same measurement data with different
software packages being used. But there is a paper somewhere on the
ICC web site that speaks to these issues with v2 and v4 display
profiles.
Chris Murphy
Color Remedies (TM)
www.colorremedies.com/realworldcolor
---------------------------------------------------------
Co-author "Real World Color Management, 2nd Ed"
Published by PeachPit Press (ISBN 0-321-26722-2)
