Digital BW, The Print

Hi Peter,

I'm sorry to hear that you have had continued problems with these grainy
negs. I believe what you are seeing may be the result of sampling 'errors'
or 'harmonics' or 'noise' that are (in my understanding) not possible to
avoid in a digital sampling system. It just may be the case that the Epson
driver's reputed 360dpi sampling may be avoiding something that the reputed
720dpi sampling of Piezography is picking up... not to mention the many
varied ways that sampling of the image data can be achieved.

You are correct to say that the 21-step image would not show problems that
may show up with a grainy negative, but in my view it is not particularly
intended for that; it is intended to check that the ink is being laid down
correctly by the printer... and you can tell a lot from that print if
something is wrong, be it banding, failed nozzles, ink in incorrect
positions, etc.

Back on 12th May, I posted a message to the Digital BW group that to my
knowledge went unanswered... 'Scans and Grain and Output options'... that
raised exactly this issue; that perhaps conceptually there is an optimal
size for grainy images to be reproduced (in my little world anyway). I can
certainly confirm that I have had some excellent grainy output from
Piezography but I would imagine that if it were too small relative to the
scan it would fail, and prints would start to look muddy.  In that email, I
asked if anyone had seen anything similar in the traditional darkroom; or
put another way, if in the traditional darkroom an image from a grainy neg
would look good at some enlargements but not at others... this would not be
due to sampling errors per se but silver chemistry and human optics.  This
is also 'just the same as' testing the output from a sharpening operation
_at_final_print_size_, rather than in smaller test prints. If only we could
change our film-grain just before selecting our final print size??!???

So, in summary, I suggest that you try different print sizes for your image.
Perhaps noise and grain is where the normal 'scan at highest optical
resolution, resize image without resampling, and let the dpi fall where it
may' [normally recommended for Piezography users] falls down. Certainly it
is relatively easy to see how Photoshop resize operations can dramatically
effect the feel of a noisy image by selecting different methods; especially
if you resample smaller.

You could be right that noise / grain could be effecting the way that people
see their prints, e.g. skin-tones as you note... but again... the 21-step
wedge is a good place to start to check that the printer is laying down ink
properly.

Best regards,
Nij


  -----Original Message-----
  From: Peter Baumbach [mailto:info@...]
  Sent: 26 September 2002 01:39
  To: DigitalBlackandWhiteThePrint@yahoogroups.com
  Cc: Nij; piezobw@...
  Subject: Grain drives PiezoTone mad


  I am currently printing a portfolio for a friend. She photographed nudes
on 35 mm Tri-X. The PiezoTone prints look posterized in the midtones (like a
print that was not totally fixed) , the transitions from the midtones to the
darker skin tones are much too hard. Because I just started with PiezoTone I
thought it was my fault and checked everything again and again with no
improvements. But today I tried Woolfs workflow and the Epson driver. And
voila: the problems are solved although the highlights are not as dotfree as
with the Piezo driver (under a magnifier). But thats not a problem with
grainy negatives...

  In my opinion the holy 21 step wedge test does not show the real problems
unless you are exclusively working with large format cameras and ASA 100
films or digital cameras. My assumption now is that the piezo driver has
difficulties with interpreting noisy pixels. And you all can proof this if
you use the 21 step wedge and add a small amount of noise that is more or
less in the range of a 35 mm 400 ASA film. Then the tonal transitions being
smooth when printing the step wedge straight become totally different. In
the range of 35% to 50% you can find posterization, from 50% to 60% there is
an artificial lumpy structure not related to the noise, from 65% to 70%
again some posterization, from 70% to 100% again a too noisy lumpy structure
(but less pronounced than in the 50 to 60% range). Making the same
experiment with the Epson driver and Woolfs workflow does not show this
phenomenon.

  If it is the grain that disturbs the piezo driver interpretation of
density then this would explain why some people reported problems with
posterization and bad skin tones and others did not mention it at all
although the product is long enough in the market (btw: who tested fading
before launching PiezoTone?? Should have been noticed before as well.).

  Any comments?

  Peter Baumbach
  Fine Art Photography Munich



[Non-text portions of this message have been removed]