[argyllcms] black generation drives me nuts

Hi Graeme,
you know that I am sold on argyllcms, but to puzzle out the right black generation parameters sometimes drives me nuts. 

I tried to generate a profile from the data set
<http://color.org/FOGRA40L.txt>

with a black generation comparable to to this profile:
<http://www.colormanagement.org/en/isoprofile.html#SC_paper_bas>

here is a screenshot of "xicclu -g -fb -ir SC_paper_bas.ICC":
<http://digitalproof.info/argyll/black_gen/xicclu_-g_-fb_-ir_SC_paper_bas.png>

I attended the instructions at
<http://argyllcms.com/doc/Scenarios.html#PP6>

and literally tried hundreds of different parameters.

E.g. I create a preliminary profile with
"profile -v -ql -b -no -ni -l250 F40"
and tested several black generation parameters, see
<http://digitalproof.info/argyll/black_gen/F40_250/>
To bar the curves from gamboling around I tried to generate a bigger data set with fakeread: 

targen -d4 -l300 -c SC_paper_eci.icc -A0.5 -f3000 -v3 F40fakeread
fakeread -v SC_paper_eci.icc F40fakeread
profile -v -ql -b -no -ni -l300 F40fakeread
mv F40fakeread.icc F40fake300.icc

and decided to use it with the following black generation parameters:
kp 0 0.15 1 1 0.65 -l300
as the curves looked smooth an reasonable to some extent, see
<http://digitalproof.info/argyll/black_gen/fakeread/xicclu_-g_-kp_0_0.15_1_1_0.65_-fif_-ir_F40fake300.png>

but when I created the profile with
"profile -qm -k p 0.0 0.15 1 1 .65 -l300 -L100 -S sRGB.icc -cmt -dpp F40fakeread" 

I got this:
<http://digitalproof.info/argyll/black_gen/fakeread/xicclu_-g_-fb_-ir_F40-fakeread.png>

What am I doing wrong?

Klaus

Other related posts: