Re: desktop negatives

Date view Thread view Subject view Author view

From: Jeffrey D. Mathias (jeffrey.d.mathias@worldnet.att.net)
Date: 01/28/01-08:50:10 AM Z


Judy Seigel wrote:
> Dan speaks highly of the piezo system, but not as I understand it for
> negatives, rather for *prints*. I saw his negatives on some kind of
> plastic (I forget the name, maybe backlight film), very thin coat of
> nearly pure orange. Yes orange. It looked suprisingly faint, barely
> visible in fact, but made EXCELLENT full scale pt/pd prints. ...

Yes, I have seen those orange negatives.
However, I would disagree somewhat with the "Excellent full scale pt/pd
prints".
I have seen some of Dan's pt/pd prints (as well as many others) and they
do have full scale and they are excellent photographs, but not excellent
full scale as there is some posterization of the tone values. This
results in a reduction of the total number of tones that can be
produced. There are ways to get around this, one of them being a
synthesis of tones such as with multiple exposures with various
negatives. The problem becomes a mater of producing (or reproducing)
nuances of tone from the original seeing. These nuances are blocked at
a certain level of posterization and are not simple to recover.

In order to achieve the "excellent full scale", posterization must be
kept to a level such that enough tones are available. The early hurdle
was to go beyond 8-bits of information and limit the use of software
manipulation that also posterizes (most things, especially sharpening).
This can be done by either going 16-bit or quad-tone (4 x 8-bit). The
present hurdle is to print the information out. Conventional printers
and drivers, in effect, posterize the information further. This is
mainly due to the limitation of the ink used. The four values of the
Cone quad-ink allow for a decrease of posterization and an increase of
tone levels. Of course, a new software driver for the printer must be
used as well. (Keep in mind this is effectively one color using four
inks. Analogous CMYK would use 16 inks.)

A test of achieving acceptable posterization may be to photograph
(minimum size 8x10) a continuous tone range from black to white, contact
print the original negative and contact print from a digital negative
made from the original. Then place the two side by side and "count" the
tones. Posterization should be observed as a stepping of values with
the width of the step becoming larger as posterization increases.

> ... In fact Dan's entire book relies on this principle....

As with most ongoing research, I anticipate some valuable refinements in
an addendum to hid book, especially knowing of his continuing work with
the Cone system and transparent substrates.

-- 
Jeffrey D. Mathias
http://home.att.net/~jeffrey.d.mathias/


Date view Thread view Subject view Author view

This archive was generated by hypermail 2b30 : 02/05/01-11:45:23 AM Z CST