[argyllcms] Re: Newbie targen and colprof questions.
- From: Graeme Gill <graeme@xxxxxxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Fri, 21 May 2010 15:32:36 +1000
Symphony Photography wrote:
I hadn't used any gamut mapping but for the sake of testing I just generated a
profile
mapped to AdobeRGB but there was no significant difference between the original
profile
and the profile using gamut mapping. Perceptual is the rendering intent I'm
using.
It's hard to comment without the details. For instance, is the source
image Lab space, or something else ? How did you create a profile
using gamut mapping ? (to be sure that there isn't a mistake,
and to see what choices you made). etc.
View condition settings? As in monitor settings, ambient light, etc.? My screen
(NEC
2690) is calibrated to 110 cd/m^2, D65. I have a profile for L* and gamma 2.2.
Ambient
light is very very low (blackout shades, etc.). The NEC is profiled with the
Spectraview package. I also have an HP LP3065 which is profiled with Color Eyes
Display
Pro.
Because by default Argyll does gamut mapping in a color appearance space
(CIECAM02),
the viewing conditions of both the source and destination colorspace help
define the appearance, and hence what the gamut mapping is mapping from and to.
Seeing what parameters you used would clarify this.
The original: http://www.symphonyphoto.com/argyllcms/Atkinson_orig.jpg
Is that your RGB original, or is that the Lab original converted into
some RGB space ?
Soft proofing with the Argyll profile (rendering intent is perceptual, black
point comp
on, simulate paper off (I just wanted to see the changes to the color itself))
http://www.symphonyphoto.com/argyllcms/Atkinson_Argyll.jpg
Argyll doesn't have BPC, it has gamut mapping instead, so it's hard
to guess exactly what you've done here.
I've also notice there's a lot of shift in greens/yellows. If you look in the
shadows
of that pic you can see a dark green going to a much brighter green.
Without a definition of the source and destination colorspaces, it's
hard to comment. This could be exactly what's needed to compensate for
the destination colorspace for instance. (The image you refer to hints
that a colorimetric intent is being applied though, whereas the Logo
image obviously has a high degree of compression applied.)
Graeme Gill.
Other related posts:
