[argyllcms] Re: How to do this?
- From: Graeme Gill <graeme@xxxxxxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Sat, 25 Nov 2017 11:33:21 +1100
Yves Gauvreau wrote:
most of what read is from Mark I. Nelson or from his own sources and I have
a very bad memory, so I can't be trusted (LOL). I went back to verify and
here is his explanation for using color to make our negatives, this date
back to 2004.
OK, that mostly sounds reasonable, although it's a generalization. For
instance, I'd expect different black inks to have different characteristics,
pigment inks being more opaque than dye inks etc.
One of the things working against this whole approach is the inverting
nature of the photographic print making. Normal printers have a characteristic
"dot gain" characteristic, which (like CRT gamma) fortuitously helps
in making the response more perceptually linear. When you use this
as a negative, it will be working against you. So the main task
here doesn't seem to be what I'd call calibration, but what I'd
call "device setup". And being an unusual device (actually process),
it would need special development. i.e. what I'd be doing is
characterizing all 4 printer channels separately as well
as a composite grey wedge, and consider how to compose some
set of them together to achieve a reasonably perceptually
linear characteristic. This is analogous to how normal, light, and
light-light inks are composed together to make up one color
channel for inkjet printers. Typically there would be careful
cross-overs between the channels. Having done that setup,
then the overal tone characteristic could be used for a normal
profile or calibration. Being a bespoke process though, this would
all need bespoke development.
Though the calibration curve approach you suggested may have work fine for
me and I thank you for it again. I think creating a 3D lookup table in plain
text will be much easier for me.
Creating a 3D lookup by hand is not easy at all. You might be able to
generate values some other way though (spreadsheet ?) and past them
into a 3D table. The basic work is the device setup work I describe
Other related posts: