[argyllcms] Re: CMYK Profile Problem
- From: Graeme Gill <graeme@xxxxxxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Fri, 12 Mar 2010 11:41:53 +1100
qcore@xxxxxxxxxxxxxxxx wrote:
I wanted to build a couple of profiles based on PSO_LWC_Improved_eci
with modified black generation curves.
However, every profile I build using colprof seems to give me ugly warm
shadows instead of the neutrals I'm expecting :(
No matter how I set the black curve the yellow and cyan inks look as if
they've swapped so I'm getting values like 50c 54m 60y 90k.
I gave up with the data in the profile and used fakeread instead -- but
got similar results.
Thinking I'd screwed up somewhere I repeated the process, running
identical commands but substituting the ISO Coated v2 (300) data and
everything worked as expected.
Has anybody else worked with PSO LWC Improved and noticed anything odd?
Hi Martin,
I'm not sure what's happing for you. I ran a smoke test:
extracted the .txt file from PSO_LWC_Improved_eci.icc, converted it
to a .ti3, then made a simple profile:
./colprof -ni -no -v -qm PSO_LWC_Improved_eci
> profile check complete, peak err = 1.490497, avg err = 0.280641, RMS =
0.340955
and then ran a standard RGB image through sRGB -> PSO_LWC_Improved_eci.icm
and had a look at it as if it was ISO Coated v2 (ECI),
and neutrals were within 1 DE of Lab 0,0, indicating that it's
not dissimilar to ISO Coated v2 (ECI).
icclu -s 100 -fb PSO_LWC_Improved_eci.icm
20.000000 0.000000 0.000000 [Lab] -> Lut -> 44.298044 34.798659 30.651975
96.099571
15.000000 0.000000 0.000000 [Lab] -> Lut -> 78.643334 74.510142 58.376359
99.156783 [CMYK]
What sort of options were you using ?
Which platform are you on ?
How were you verifying it ?
cheers,
Graeme Gill.
Other related posts:
