Digital BW, The Print

> It seems wrong to leave the impression that Austin's position, although
> apparently intractable, is without merit.

I don't think any of us believe Austin's notion is without merit, it's just
that he hasn't demonstrated that anybody but him thinks it's worthy of
pursuing WRT prints. I don't mean that to sound too cynical, I just mean he
makes some sense, but it doesn't help his case that he hasn't provided
evidence to support his notion that the DyR formula is meaningful in this
context. "Noise" in analog photography is typically from grain, but silver
papers don't really have a visible grain to speak of, so we don't know what
Austin refers to when he states his formula.

However, there are things that can affect the "tonality", or tonal
distribution, of a print - flashing was one previous example - and having a
neat little formula for expressing those qualities might be useful, though I
doubt it. We speak of the shape of a paper's characteristic curve, hard or
soft tonality, creamy tones, soot and chalk tones. These are some of the
characteristics I thing Austin is thinking of when he speaks to a prints
dynamic range, but the question is, can you ascertain those qualities from
[(dMax-dMin)/noise]? If I tell you one paper has a dynamic range of 2.2 and
another 2.8 do you really envision them any other way than that one has a
greater density range than the other?

Ultimately it's not a question of could you apply Austin's formula, it's
would you?

I find it interesting that for all DyR's reputed relevance to photography,
even Photoshop doesn't include a DyR analysis of a file. Why? I believe one
reason is because it can't know what a file's input values were supposed to
be, and thus can not determine what in the file is "noise". Instead
Photoshop does a far more complex analysis of a file, quantifying every
pixel, and expresses those results in terms you find in the histogram such
as Mean, Standard Deviation, Percentile, etc, - but no DyR. My only point is
that it would seem that DyR, while very useful in some applications, is out
of place in others. I would guess it's out of place where noise is not a
known, or relevant, quantity or quality.

When Austin defines "noise" WRT photo paper, demonstrates that typical photo
papers contain noise to a significant degree, and shows that someone of
repute - other than himself, ;-) - has applied his formula to a print and
gotten meaningful info from it, he may well win all of us over.

I'm kinda rooting for him, as I'm sure many of us are, but he's not there
yet... ;-)

Todd