[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Digital Pt/Pd tones continued

Thanks, Nick.  I have received inconsistent results when printing higher
resolutions with my Epson Photo EX, as well.  I'm anxious to run some tests.
A couple of questions I have:

1)  Are you using the Duotone mode to apply the Sepia?  If not, how?  You
mention you use Quadtones, so I'm confused as to what the Sepia step is
meant to accomplish.

2)  Are you mixing your own Quadtone inks?  Who is your ink supplier?

Keith


-----Original Message-----
From: Nick Makris [mailto:nick@mcn.org]
Sent: Tuesday, July 11, 2000 11:12 AM
To: Alt Photo
Subject: Digital Pt/Pd tones continued


Just finished running some tests with a particular image from photoshop
v5.02 to Epson 1520 (1440dpi).  The tests included two versions of the same
file, one scanned at 410dpi (the most recent) and one at 360dpi - the more
recent one and this test were done on the heels of the recent discussion on
tones.  Please understand that my printer is equipped with a Quadtone inkset
with a standard black ink and with the color inks replaced with 25%, 50% &
75% black inks.

You may recall that I discovered, now from two sources, and previously
reported, that the Epson 1440 dpi printers don't print images with
resolutions greater than 360dpi as well as those at 360.  The problem
extends from the idea that above 360dpi a certain amount of resampling is
done by the printer, which causes results that are unpredictable.  In this
case, all the images I printed from the 410dpi source showed some sign of
posterization while the 360dpi source produced no signs of posteriztion,
except as noted below.  In addition, the 410dpi file caused the print time
to double +/-.  The resampling surely has something to do with the printer's
processing algorithm and the coincidental association of 1440(printer
resolution)/4(cmyk?)=360???.

Some additional observations that came from the 20 8X10 images I printed
(most printed on satin inkjet (94 brightness) paper, but some on bond for
comparison):
  a.. The smoother the paper surface, less posterization and the better the
definition.
  b.. The brighter the paper brightness index the cleaner the highlights.
  c.. A simple conversion to grayscale produced posterization where the use
of Lab color mode did not in most cases.
  d.. The use of color images never printed as cleanly as with Lab mode -
remember that I'm using quadtone inks.
  e.. The use of strictly black and white color did not produce an image as
clean as those produced with a slight sepia toning.
  f.. I NEVER EXPERIENCED ANY ON SCREEN POSTERIATION WITH ANY OF THE TEST
IMAGES.  I feel this suggests the culprit, in most posterization problems,
is related to how the printer reacts to the input.  Can't be sure of that.
I don't pretend to fully understand why I experienced the above, but for
what it's worth.....

I have developed/refined a Photoshop "Action" that will convert any color
image (from a point that you think it looks best) to Lab Color, delete the
appropriate channels, apply grayscale, a sepia tone, and convert it back to
RGB so that it looks like a sepia toned grayscale image.  BTW, I never save
my files after this conversion - it only takes a few seconds to complete.  I
also have a Pt/Pd curve that I think works for both the Lightjet 5000 and
the Epson 1520 with quadtone inks. I am happy to provide them to anyone who
asks off list.

Also, as I have previously reported, no curve that I encountered during the
last year and half (since I started this digital project) works with the
above.  The curve that I developed is a very straight forward (gentle 'S')
one that only gives the appearance of added contrast to the on screen image
and also to the printed output.  You may be required to change it so that
the resulting prints (negs) conform to a density and contrast compatible
with 'YOUR' normal sensitizer.  The unsuccessful curves that I tested all
seemed destined for output to an Imagesetter which apparently responds
differently???

BTW, all the above tests/observations were done with a positive output.
However, I have tested the conversion to a negative and the results are
identical.  The negative conversion is the last step before printing and can
be accomplished in the printer dialog box or on the Image menu.

Caveat Emptor.

Nick