[Openicc] Introduction / Gutenprint

[Graeme Gill]

Gerhard Fuernkranz wrote:
> But I guess it can only work if the profiler is aware of the special
> semantics of this channel. I suspect, it won't work with a generic
> multicolor profiler, which does not know about gloss optimizer.

Yes, but this is no great trick. It just needs an appropriate attribute
on the channel "color", and the code to recognize it treat that channel
with a "gloss optimizer" generation algorithm. It's almost
trivial because it can be added as a final pass after a separation
LUT has been created. (At least, that's the way I'd approach it
if I get to playing with an R800/R1800).

I'm assuming a separate profiling & separation model here, so the
change in color response due to adding the gloss optimizer will
be compensated for in the more conventional 3 or 4 pseudo device
channel profiling.

Graeme Gill.