[argyllcms] Re: ArgyllCMS V1.1.0 RC1 is now available
- From: Graeme Gill <graeme@xxxxxxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Sat, 07 Nov 2009 12:30:11 +1100
Alastair M. Robinson wrote:
(Any chance of a future version of targen being able to generate
gamma-adjusted step wedges, by the way?)
It's still on the list of things to do, because it doesn't
seem that important given that step wedges use relatively
small number of patches. To a degree you can fudge this though
by applying calibration curves using printcal.
[If you look at the .cal then DEVMAX_USED is 1.0 for all the inks.]
Yup, I see that. So I take it the auto-aim doesn't attempt to use the
hue-shifts at colorant saturation to decide where to stop?
The approach is the one I have mentioned in a few forums,
which is to stop when the ratio of colorant to delta E from
white becomes too low. This sort of does what you mean, in that
a pure hue shift doesn't advance the distance from white,
but no, the heuristic doesn't directly examine locus curvature.
I tried applying the gamma adjustment to the device input values in the
.ti3 file, and the five colorants still get a DEVMAX_USED of 100%, and
black gets about 50% - I shall experiment further :)
I doubt the heuristic is perfect, it is just a place to start
that tries to avoid weirdness near the colorant maximum. It
may get it wrong at times (although I'd be interested to see
examples). I've also attempted to treat printer RGB the right
way (note the change to the colorant selection in targen
to distinguish between printer and display RGB), but I haven't
fully explored how well this works in practice with a real
"RGB" printer.
cheers,
Graeme Gill.
Other related posts:
