Re: Digital methods for alternative photographic processes, Part 3 of 3

Carl Weese (cjweese@wtco.net)
Mon, 07 Apr 1997 17:00:58 +0000

Charlie,

One thing that strikes me right away about the difficulty getting the
steps you want from an imagesetter. By making fairly radical adjustments
to the file in the curves dialog of PhotoShop, you may be ending up with
far fewer than 255 steps in the file's histogram. When you ask for a
histogram from your output file, does it show a ragged shape like a comb
with missing teeth? If so, and the missing teeth are in the areas where
you have posterization in the imagesetter output, there could be a
connection.

Color management systems like KCMS use "Precision Transforms" to tailor
the data in a file that looks good on a calibrated monitor to the
precise needs of specific output devices, without the data loss
inherent in a curves adjustment within the program. PT's for
Palladiotype, Cyanotype, etc, could do the job, but I don't expect Kodak
is about to write them for us.

Another possibility for less data loss would be to do as much as
possible of the needed adjustments in the scanning module of a scanner
with "extra" (10 or 12 bit) bit depth. The idea is that by applying the
"alt.proc. curve" within the scanning module you bring a histogram of
the right curve shape into PShop with a full set of 255 levels. The less
manipulation in PShop, the more data that goes to the imagesetter. The
difference is clearly visible in files intended for 4/C output. The
disadvantage is that your monitor image will not look good (KCMS files
look terrible on screen after they have been through their PT's, too).
But spotting and retouching would still be digital-easy. I don't *know*
that the posterization is due to relatively empty sections in the
histogram, but it's worth investigating.---Carl