[argyllcms] Re: AW: Re: WG: Question about log-output from profcheck
- From: Gerhard Fuernkranz <nospam456@xxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Mon, 8 Jun 2020 21:30:58 +0200
Am 08.06.20 um 09:04 schrieb Hermann-Josef Röser:
Profcheck basically takes the device values (RGB, CMYK) from the .ti3 file,
applies the given profile with _absolute colorimetric_ rendering intent in
order to convert these device values to XYZ (or CIELAB),
So this is the reason why I do not see the exact values from the reference
table in the output from profcheck, because the rendering intent has changed
these values?
The numbers labeled with "should be" in the profcheck output ARE the measured colors from
the .ti3 file (if necessary just converted from XYZ to CIELAB, but not mangled otherwise). Ideally,
the colors predicted by the profile (i.e. the numbers after "->") should be the same
ones, but in practice they deviate more or less due to limitations of the profiling process and
imperfections of the measurement data.
This brings up a longstanding question I have. How comes the rendering
intent into play here? I always thought, the intent is used if one goes from
a large colour space to a smaller one. Since XYZ / Lab cover all potentially
possible colours, there are by definition no out of gamut values. So what is
the role of the rendering intent here? And if so, why the absolute
colorimetric intent?
The measured XYZ in the .ti3 file are absolute colors, therefore profcheck
applies the profile with absolute colorimetric intent as well, in oder that the
compared colors are both absolute ones. Just in order to compare apples to
apples, and not apples to pears.
Regards,
Gerhard
Other related posts:
