For testing I used collink, but the same options are available in colprof as
well.
For instance:
colprof -v -tpa -l300 \
-cmt -cw:0.950543:1:1.089303 -cg:0 \
-dpe -dw:0.96420288:1:0.82490540 -dg:0 \
-s sRGB.icm fogra52
Result (with patched code for D=1):
$ icclu -ip -fb fogra52.icc
0.000000 0.000000 0.000000 [Lab] -> Lut -> 1.000000 0.450553 0.491272 0.995408
[CMYK]
10.000000 0.000000 0.000000 [Lab] -> Lut -> 0.347657 0.116609 0.234120 0.947229
[CMYK]
20.000000 0.000000 0.000000 [Lab] -> Lut -> 0.178736 0.047830 0.178229 0.890756
[CMYK]
30.000000 0.000000 0.000000 [Lab] -> Lut -> 0.095535 0.003950 0.133099 0.816358
[CMYK]
40.000000 0.000000 0.000000 [Lab] -> Lut -> 0.060886 0.000000 0.125472 0.711627
[CMYK]
50.000000 0.000000 0.000000 [Lab] -> Lut -> 0.000000 0.000000 0.080945 0.668284
[CMYK]
60.000000 0.000000 0.000000 [Lab] -> Lut -> 0.000000 0.000000 0.085747 0.520426
[CMYK]
70.000000 0.000000 0.000000 [Lab] -> Lut -> 0.000000 0.000000 0.091517 0.370945
[CMYK]
80.000000 0.000000 0.000000 [Lab] -> Lut -> 0.000000 0.000000 0.099684 0.226189
[CMYK]
90.000000 0.000000 0.000000 [Lab] -> Lut -> 0.007982 0.000000 0.099560 0.098501
[CMYK]
100.000000 0.000000 0.000000 [Lab] -> Lut -> 0.017055 0.000000 0.104458
0.000014 [CMYK]
$ icclu -ia -ff fogra52.icc
1.000000 0.450553 0.491272 0.995408 [CMYK] -> Lut -> 28.494265 0.011923
0.021938 [Lab]
0.347657 0.116609 0.234120 0.947229 [CMYK] -> Lut -> 32.726870 0.090648
-0.067010 [Lab]
0.178736 0.047830 0.178229 0.890756 [CMYK] -> Lut -> 36.398021 0.018380
0.093168 [Lab]
0.095535 0.003950 0.133099 0.816358 [CMYK] -> Lut -> 41.022183 -0.066029
0.120640 [Lab]
0.060886 0.000000 0.125472 0.711627 [CMYK] -> Lut -> 46.776723 -0.045948
0.137115 [Lab]
0.000000 0.000000 0.080945 0.668284 [CMYK] -> Lut -> 50.072181 0.892953
-0.436071 [Lab]
0.000000 0.000000 0.085747 0.520426 [CMYK] -> Lut -> 58.717176 0.774010
-0.860462 [Lab]
0.000000 0.000000 0.091517 0.370945 [CMYK] -> Lut -> 67.875000 0.701273
-1.066479 [Lab]
0.000000 0.000000 0.099684 0.226189 [CMYK] -> Lut -> 76.971693 0.700319
-0.162919 [Lab]
0.007982 0.000000 0.099560 0.098501 [CMYK] -> Lut -> 85.159849 0.598095
-0.190186 [Lab]
0.017055 0.000000 0.104458 0.000014 [CMYK] -> Lut -> 91.726311 0.125573
0.030083 [Lab]
Regards,
Gerhard
Am 22.09.2017 um 14:14 schrieb Henrik Olsen:
On 22 Sep 2017, at 13.37, Gerhard Fuernkranz <nospam456@xxxxxx> wrote:
The intent -ipa is likely close to your desire, it is obviously supposed to
preserve the CIECAM02 appearance, and additionally does gamut compression.
As with all CIECAM02-based intents, you need to specify the source and
destination viewing conditions with the options -c... and -d..., and you
can also override the source and destination adapted white points with
the command line optiosn -cw:... and -dw:...
However, on the command line, you cannot override the "D" parameter of the
CIECAM02 model (-> degree of adaptation), it is always calculated automatically
from the other specified viewing condition parameters. Typically, the calculated
value lies in the 0.8...0.9 range. If you want to force it to 1.0, then it seems
you need to patch xicc/ciecam02.c.
With the patched code and overriding -cw:... and -dw:... I can obtain an almost
perfect perceptual D50 gray axis with collink.
[ However, different from the commercial "neutral" profiles, even paper white
itself gets neutralized by printing yellow ink, while the gray axis of the
commercial
profiles still cheats, beding from D50 towards blue-ish paper white in the upper
20% of the axis. ]
Regards,
Gerhard
Interesting, thanks Gerhard.
I haven’t used device links before, so I need a little help. Normally I print
my images directly out of Lightroom (only rel + BPC and percep available) or
through Photoshop. Targets with ACPU. Can your approach not work with a profile
in Lightroom then? That is clearly the perfect scenario for me.
Even if I made scripts for using the device links, my normal workflow would be
somehow broken if having to render explicit printer output tiffs, compared to
having it all done under the hood from parameteric edits and perceptual
profiling on raws in Lightroom. But willing to test it out of course.
Say I choose suitable viewing conditions, and possibly even patch ciecam02.c,
can you briefly help me understand the needed workflow for using this?
Regards and thanks.
/Henrik
