Hi Marti,

while thinking about this thread, the following question came into my mind:

What result would one actually expect from the gamut warning for _perceptual_ (or satuartion) intent?

Assume, we have a color in the PCS which is out-of-gamut. Now we print this color with perceptual intent, and the gamut compression remaps the color to a clearly in-gamut color. So after the perceptual intent's gamut compression, it is no longer out-of gamut.

Should the gamut warning now mark this color as out-of-gamut, or not?

My intuitive feeling sais, that I rather would not like to get this color marked, since it can be printed without clipping. Subjectively I'd rather expect to get a gamut warning only for those colors, which cannot be mapped injectively, but only surjectively from PCS to device space by the given rendering intent.

(If I would like to know which PCS colors are a priori out-of-gamut, before the gamut compression, then I could use the gamut warning with colorimetric intent anyway.)

Any thoughs or objections?
Any idea, how other applications handle this issue?

Regards,
Gerhard


Marti schrieb:

Hi,

Sorry for the long delay. I've been very busy past week...

Regarding this particular profile, sounds weird, but this is how it
handles colors. I assume you are using Adobe's SWOP.

Let's do some numbers. If I make a SWOP -> Lab transform
with perceptual intent, I got:

icctrans -c0 -t0 -o*Lab -iUSWebCoatedSWOP.icc
little cms ColorSpace conversion calculator - v1.8
Enter values, 'q' to quit
C (0..100)? 100
M (0..100)? 100
Y (0..100)? 0
K (0..100)? 0

L*=25.3324 a*=25.2695 b*=-55.3320

Ok, So this is the Lab value of the color we want to check watever it
is inside gamut hull. The algorithm lcms is using is as simple as move
it back and forth across rel. colorimetric intent. If the profile changes
the value, then color is out of gamut. Same values means inside gamut.
This method is not perfect, but works quite well as a coarse
approximation.

Let's now convert this Lab value across colorimetric intent to CMYK colorant.

icctrans -c0 -t1 -i*Lab -oUSWebCoatedSWOP.icc
little cms ColorSpace conversion calculator - v1.8
Enter values, 'q' to quit
L*? 25.3324
a*? 25.2695
b*? -55.3320
C=100.00 M=98.72 Y=8.84 K=5.62

The colorant values are not (100,100,0,0) but that doesn't matter, because
same color can be obtained by different CMYK combinations. The important
part would be that CMYK contone SHOULD give the same Lab if measured
across AtoB1 table-But as you can see that is not the case:


F:\lcms>icctrans -c0 -t1 -o*Lab -iUSWebCoatedSWOP.icc
little cms ColorSpace conversion calculator - v1.8
Enter values, 'q' to quit
C (0..100)? 100
M (0..100)? 98.72
Y (0..100)? 8.84
K (0..100)? 5.62
L*=29.6247 a*=16.2070 b*=-41.4492


So, there is a huge shift from (25, 25, -55) to (30, 16, -41), hence the profile
handles this value as out of gamut.


Which is not so surprising at all, because (100, 100, 0, 0) is over gamut
boundaries. The profile is doing some slight gamut remapping in the extreme
to avoid gradient blocking. That is not so bad at all, images should not use
such extreme contone values.


Regards
Marti Maria.


----- Original Message ----- From: "David Purton" <[EMAIL PROTECTED]>
To: <[email protected]>
Sent: Wednesday, February 16, 2005 1:53 PM
Subject: [Lcms-user] Out of Gamut Alarm Question



Hi,

Quick question on lcms handling of out of gamut checking.

I notice that when I proof a CMYK colour to RGB, using the same proofing
profile as input CMYK profile, some colours are marked out of gamut. For
example, C100 M100 Y0 K0 using SWOP as the CMYK profile and sRGB for the
RGB, is marked as out of gamut. (perceptual intents)

Is this supposed to happen?

I would have thought that if my input profile and my proofing profile
are the same, then all colours are by definition in gamut.

It seems to especially crop up in dark colours.

(fwiw, CorelDRAW's colour management, does not mark this colour out of
gamut under these circumstances)

cheers

dc


(Cc me if you think to, I'm not subscribed)






------------------------------------------------------- SF email is sponsored by - The IT Product Guide Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now. http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click _______________________________________________ Lcms-user mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/lcms-user

Reply via email to