From: Sandy King <sanking@clemson.edu>
Subject: Re: Pyro Developers
Date: Sun, 04 Apr 2004 12:48:56 -0400
> As Carl Weese notes, the way that developers render tonal values is
> at least as important as definition. There are indeed some high
> accutance developers that give greater resolution than the pyro
> developers I mentioned but they do so at the loss of some smoothness
> of tone.
I agree 100% on this one, and I personally did not like radical "high
definition" type developers such as Crawley's FX-1, though with some
films it worked nicely and produced very acceptable tone. Indeed, I
had to spend some time to formulate a developer that is moderately
high accutance and consistently gives very nice tonal rendition and
acceptable granularity. In its simplest form, it is DS-2.
While I was working on developer formulae, I routinely enlarge 6x to
14x from negative, and negatives with too much accutance effect would
give disagreeable prints. It's hard to say, but it makes me feel like
looking at the print in a refrigeration room. Indeed, there are films
(such as HP5 Plus) I do not use accutance developer. To me, small
difference in resolution limit is not very important compared to
overall finish of the prints. (For your printing process, the
situation might be different. This is also subject to personal taste.)
DS-2 is a metol-ascorbate developer containing 0.4g metol, 1.0g
ascorbic acid, 20g sodium sulfite, 1.0g sodium bicarbonate, 4.0g
sodium carbonate monohydrate in 1 liter of water. Target pH is
9.80. For TMX, I use this at about 8.5 min (didn't test new 100 TMX
yet). For Fujifilm Acros, I mix two parts DS-2 and one part water
(2+1, not 1+2) and develop for 8.5 min., all at 20C. I have time for a
few other films, but these are the films I feel this developer is most
useful for.
> As for the resolution limits please correct me if I am wrong but my
> thinking is this. If I expose two negatives on the same roll of film
> with the same camera and same aperture setting any difference in
> resolution observed must of necessity result from either the
> developer or method of development. That is to say, the limit to
> resolution in this case can not be in either the film or in the
> optical system.
> So what I am observing is that the film I used (TYMAX 100) and the
> optical system appear appear to be capable of resolving well over 100
> lpm (in the center of the file) so that the developer itself, and not
> the film or optical system, establishes the limit for resolution.
This is not right or wrong thing, but quantitative testing for
accutance in a way that the result is meaningful and relevant is
difficult.
In developed negative image structure, the main differences you see
between fine grain formulae and accutance formulae are (1)
granularity, (2) local contrast due to *adjacency effect*. I
understand you control overall macroscopic contrast of
development. This is not that. This increased local contrast underlies
apparent enhancement of sharpness in enlarged positive images, and
this happens in midrange of the spatial frequency scale, well below
the resolution limit. Slight difference in resolution limit has much
weaker visual impacts.
Therefore, the directly relevant metric in your comparison is the
increase or decrease of local contrast at moderately high spatial
frequency but still well below the resolution limit (say around
40 lpmm). What you measured, the resolution limit, is only a peripheral
factor of this effect, or more like an epiphenomenon, rather than a
factor of the direct interest or a factor of the direct consequence.
As you said, TMX can resolve 100+lpm except for very contrast
conditions. And this level of resolution of TMX can be achieved with
D-76. On ther other hand, films like HP5 Plus developed in D-76 to the
same CI can provide much less resolution than TMX, but at moderate
enlargements, HP5 Plus often looks sharper. This is because the
modulation transfer of medium frequency range is rather a different
issue from the absolute resolution limits that these numbers indicate.
In other words, if you interpret your test results in a way that the
resolution limits indicate sharpness of the image, you are assuming
that the sharpness and resolution vary together. The above example is
a clear counterexample (though using two films and one developer
instead of one film and two developers).
Another difference is that, edge effect is quite well demonstrated
with step-like exposure gradings, but in real life, the transition is
more gradual because subjects are 3 dimentional and light sources are
large in physical dimensions. One approach to quantify sharpness in
more realistic condition is to use targets with sinusoidal grading to
measure modulation transfer function. I know this takes a serious
setup to do a test.
Say you use the same target, lighting, and optical system to project
the target image onto the film, and say all these factors are
invariant with film or developer of choice. True, but the question is
that effective lighting contrast of low frequency pattern and high
frequency pattern may differ due to imperfections of optical system.
Resolution limit of film depends strongly on the target contrast (the
ratio of irradiation of the brightest point to that of dimmest point,
nothing to do with development). Although this makes it more likely
that "resolution limit" measurement in your technique is more or less
related to local contrast enhancement and sharpness because higher
frequency targets are more likely to have lower contrast and more
gradual transition than with lower frequency targets, this is just
another potential problem in interpreting results such as "10-15 lpmm
enhancement in resolution limit."
Someone might think all these are too fastidious a detail, but I'd
disagree. The price you pay in measuring something other than the
factor you are directly interested in is limited room of
interpretation of the results. Indeed, naive, literal interpretation
of "10 lpmm enhancement in resolution" would be very misleading.
-- Ryuji Suzuki "All the truth in the world adds up to one big lie." (Bob Dylan 2000)Received on Sun Apr 4 15:07:35 2004
This archive was generated by hypermail 2.1.8 : 05/14/04-02:14:31 PM Z CST