first, I'll admits I know basically nothing on color science, profiling and
probably much more I don't even know I don't know.
Understandably, all I'm saying is viewed by you folks here through the prism
of color science and that's how things work, I have no problem with that.
When I read how a profile is made and how it works, I see a set of tools
that could do something it was not necessarily design for.
I see you use the word characterisation a lot in other messages, which for
me mean creating a mathematical model to understand say how a printer spit
its inks to produce the desired colors. Nothing wrong with that per say. I
understand that in general that this is the basic idea behind creating a
profile in the first place.
But say, for just the fun of it, we have an input file where R=G=B and I
want to create a profile to convert these RGB values to be completely random
color values. Of course in a later profile I would replace the randomness by
less random colors or even by very specific color of my choice. I found a
tool to convert a profile to xml, a text file and one to convert this xml
file back to an icc profile, it works on ICC version 2 and up.
I would need to first create maybe a dummy profile that have all the
structure and data needed so I can just replace a set of 512 possible R=G=B
values with a set of 512 randomly chosen colors among the 2^48 color
available. All this via the icc => xml => edits => icc route.
I've chosen purposely this 512 value because it would be my goal to
reproduce between 300 to 500 discrete tonal values on my photographic
prints. Of course, if can reproduce more I wouldn't say no.
Would you have a suggestion for such a synthetic profile I could begin with?
Is there one in the Argyll directory tree I could use for starters or could
I make one and how?