Extract Dmax-L min from an ICC printer profile

[Ernst Dinkla]

Is there an easy way to do that?

[Production]

Post by Ernst Dinkla
Is there an easy way to do that?

Do you mean something like this:

icclu -ff -ia rgbprofile.icc <<< '0 0 0'

or

icclu -ff -ia cmykprofile.icc <<< '1 1 1 1'


If you just want the L value you could use awk:

icclu -ff -ia rgbprofile.icc <<< '0 0 0' | awk '{print $8}'


CMYK L value would be $9.

[Graeme Gill]

Post by Ernst Dinkla
Is there an easy way to do that?

It depends on what you are looking for.

Are you looking for the minimum L value in one of the B2A tables
(ie. the value the profile creator chose), or are you looking
for the minimum L value the device is capable of creating ?
To determine the latter for a CMYK device you need to know
the device gamut limits (ie. ink limit etc.).

Neither are definitively accessible using any of the current tools,
since this implies searching the tables. You can do a search by
hand using xicclu, or rely on the gamut clipping to give you
a close to target value.

For instance, something close to the minimum L of the colorimetric
B2A table would be:

xicclu -fb -ir -pl profile
0.000000 0.000000 0.000000 [Lab] -> Lut -> 0.127802 0.000000 0.693490 1.000000 [CMYK]

then lookup the CMYK values

xicclu -ff -ir -pl profile
0.127802 0.000000 0.693490 1.000000 [CMYK] -> Lut -> 13.222016 0.607799 -0.301865 [Lab]

which is likely to be close to what you are after, but it's always possible that some
other CMYK value might produce a lower L* value.

Exploring what the device is capable of would be similar:

xicclu -fif -ir -pl -l 300.0 profile

etc.

Graeme Gill

[Ernst Dinkla]

Post by Graeme Gill

Post by Ernst Dinkla
Is there an easy way to do that?

It depends on what you are looking for.
Are you looking for the minimum L value in one of the B2A tables
(ie. the value the profile creator chose), or are you looking
for the minimum L value the device is capable of creating ?
To determine the latter for a CMYK device you need to know
the device gamut limits (ie. ink limit etc.).
Neither are definitively accessible using any of the current tools,
since this implies searching the tables. You can do a search by
hand using xicclu, or rely on the gamut clipping to give you
a close to target value.
For instance, something close to the minimum L of the colorimetric
xicclu -fb -ir -pl profile
0.000000 0.000000 0.000000 [Lab] -> Lut -> 0.127802 0.000000 0.693490 1.000000 [CMYK]
then lookup the CMYK values
xicclu -ff -ir -pl profile
0.127802 0.000000 0.693490 1.000000 [CMYK] -> Lut -> 13.222016 0.607799 -0.301865 [Lab]
which is likely to be close to what you are after, but it's always possible that some
other CMYK value might produce a lower L* value.
xicclu -fif -ir -pl -l 300.0 profile
etc.
Graeme Gill

Thank you.

I had RGB-device printer profiles in mind so I could add L-min numbers
from OEM and paper manufacturers profiles to the SpectrumViz inkjet
paper range measurements. Less work than printing 600 sheets with 4
different black inks (Epson, Canon, HP, and MIS) and measuring the
lowest L values of each ink. It will not represent the best Dmax
possible if I extract a related number from RGB-device printer profiles.
Even more a compromise I guess than what you propose.