Digital BW, The Print

Having returned from vacation and facing about 900 e-mails (including a very
depressing "Number of Tomes") I decided to avoid getting involved in most.
However, there were some common themes that caused me to reflect on the
state of our technology.

Some of our most experienced printers acknowledged that the best analog
prints still seem to have an edge over the digital prints.  In some
respects, I agree -- but only to a limited extent.  I switched to digital,
pigmented quad inkjets because, overall, the average quality of my prints
was better, more satisfying, and more efficiently produced that way.

But, all the systems I've tried have defects.  For example, I'd like to,
again, call attention to the smoothness of the grayscale ramps that the
Piezo and Epson drivers can produce.  See the "Ramp-smoothness.jpg" in the
"Image processing" folder of the Files section of this forum.  The Files
section is at:
http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint/files/

This image compares the smoothness of the grayscale ramps produced by the
Piezo and Epson/RGB/MIS workflows on the printers I have.  What I did to
produce this image is to print a 21-step test file with each system.  Then I
scanned the prints.  In Photoshop I then tried to neutralize the overall
ramp of the 21-step test file prints with a gradient going the opposite
direction of that ramp on the print and at 50%.  Then I increased contrast
to exaggerate the deviations -- those areas that were not neutralized
because those print ramp sections did not follow the smooth ramp of the
input file.

What I think it shows, among other things, is that neither of the workflows
we generally use is perfect.  (I am in no way trying to show that one system
is always better than the other.)

The Piezo-1160 test print clearly has finer dots than the Epson-3000 print.
On the other hand, I might argue that, with respect deviations from the
ideal ramp, the vm-sepia 3000 ramp on the bottom is better than the Piezo
ramp on the top.  However, prospective purchasers must understand that the
3000/Epson driver/RGB workflow ramp was printed with the very system on
which it was designed.  All systems/printers seem to be a bit different.  As
such, you might not get equally good prints with your system -- unless you
are willing to learn how to tweak the curves to take into consideration the
particular characteristics of your printer.  It may well be that, on
average, the Piezo driver produces smoother ramps than the RGB/Epson driver
workflows, especially on systems that have not had the RGB curves tweaked to
fit the particular printers involved.

Moreover, even though I think my vm-sepia curve for the 3000 is reasonably
good -- at least on my printer -- I was unable to get all the defects out of
it.  For example, there is a flat spot in the 3000 vm-sepia neutral curve
that causes 80% and 81% to print out as the same.  I simply cannot get rid
of it.  When I find a flat spot, I go to a test file that has 1%
increments/steps in the problem area and work from that.  If I went to a
test file that had 256 increments, I suspect that I'd find even more
problems.

When I make a print, if critical areas in the print land on 80-81% and are
negatively affected, I simply move them off that spot in the file.  So far,
I've been able to get around the problem in every print.

The argument about how many tones a driver can produce may be interesting,
but I suspect that the deviations from the ideal ramp that we see in the
"Ramp-smoothness.jpg" image far outweigh the importance of any difference
that may exist between 256 and 1000 tones on a piece of paper.  I think at
the current state of the art, we'd be lucky to get close to 256 -- that is,
in the proper order and reasonably evenly spaced.

Even with all these defects in the digital systems, however, I see even more
problems in my analog prints.  For example, the inevitable modulation
transfer function (MTF) that softens all analog enlargements has become so
easy to overcome with Photoshop's Unsharp Mask that the old MTF limits of my
medium format to 16x20 printing are gone -- but new ones have appeared.
While I used to favor a film & developing procedure that maximized sharpness
to overcome the MTF (Tmax 100 & Micro-X 1:3 with 3 minute agitation
intervals), I now find that a film & developing routine that maximizes
smoothness is the best (Tech Pan & Xtol).

I'd even guess that the noise (grain and other random fluctuations in
density) of the analog system I was using (medium format Tmax 100 developed
for sharpness) set a limit on smoothness and number of tones that is worse
than the limits imposed by the deviations in the curves that the
"Ramp-smoothness.jpg" shows.  On the other hand, the medium format Tmax 100
Micro-X  combination I used does match the quality of the Hasselblad
negative Ansel Adams used to make his famous Moon Over Half Dome print --
and that is not a bad target.

I would go even further and say that most of the enlargements by the Masters
I've seen are not better, in terms of the visual limits imposed by the
technology used, than the best that can be produced by the current digital
workflows we have -- all else (film format in particular) being equal.  And,
while the 8x10 contact print can achieve a technical quality we may never
equal, I feel limited mostly by my artistic and technical skills, not the
limitations of the current pigmented inkjet systems.  I also suspect that
the beauty of the Masters' prints has a lot more to do with the artistic and
technical skills of those individuals than with the printing technology.

Paul
http://www.PaulRoark.com