[Openicc] Introduction / Gutenprint

[Gerhard Fuernkranz]

Kai-Uwe Behrmann schrieb:

>The concretisation from a simple to a full separated or complex colour 
>space could allways be possible. This allowing: CMYK -> CcMmYK, while the 
>linearistation can be done optionally with CMYK or CcMmYK curve profiles.
>
What I want to say is, the driver should support both, accept CMYK and 
separate dark/light cyan and magenta internally in the driver, or 
directly accept CcMmYK (6-color) from the caller.

>My experience is CMYK plays basically well. Others stated even RGB 
>corrections could be ok.
>
Good reminder.

Actually I also see use cases for an sRGB print mode, where the driver 
offers a RGB process color model (similar to printers under Windows), 
and where the driver performs the sRGB -> printer profile transformation 
internally, invisible to the application, i.e. a print mode where the 
driver accepts and expects sRGB as input.

(Particularly for the vast majority of "dumb" applications which are not 
color management aware, and which only know RGB colors, which are 
usually interpreted as sRGB.)

AFAIK, Gutenprint already has a RGB print mode, but the RGB -> printer 
color space conversion is not CM based, but rather heuristic.

>For utilising most of the available gamut (within 
>the mathematical limits) n-colour separations with according linearisation 
>seems a better choice.
>
Particularly if additional primaries (not just dark/light, but e.g. 
orange, green) come into the play, I see rather a strong demand to do 
the separation with a multicolor profile.

Btw, is printing with spot colors (e.g. CMYK + two different Pantone 
inks) actually a topic which needs to be seriously considered? For 
offset printing, this is indeed a serious issue, but what about inkjet 
printing?

Regards,
Gerhard