Rick Richardson wrote: > It is RGB from the postscript CRD: > > $ foo2zjs-icc2ps -o samclp300-0.icm | grep "[34] {}" > ] 3 {} bind dup dup ] > $ foo2zjs-icc2ps -o samclp315-argyll-0.icm | grep "[34] {}" > ] 4 {} bind dup dup dup ] > $ foo2zjs-icc2ps -o testing.icm | grep "[34] {}" > ] 3 {} bind dup dup ] > $ > > CMYK profiles have "4 {} bind". > > "testing.icm" is the one from the Xrite ColorMunki sw. > Btw (somewhat OT), yes, Ghostscript obviously also accepts RGB CRDs in conjunction with a CMYK ProcessColorModel. CIE-based colors are then converted to RGB, using the CSAs and CRDs, and these RGB numbers are finally converted to CMYK using the PostScript rules for converting DeviceRGB to DeviceCMYK, which may be suboptimal. If you use a RGB CRD in conjunction with a CMYK ProcessColorMode, then in order to get deterministic results I also suggest that you explicitly install your own dedicated functions with setblackgeneration and setundercolorremoval in your PS prolog (for instance "{} setblackgeneration {} setundercolorremoval"), since I had noticed that the default for these functions was different in different gs versions, but they are used by the implied DeviceRGB to DeviceCMYK conversion. Regards, Gerhard