Digital BW, The Print

Austin, you wrote:

> > But I doubt it will mean a difference in CCD
> > filmscanners
> > with B&W films.
>
> It can, but that is film, development, exposure, scene (and in fact,
output)
> dependant.  Also, the main reason it might not make a difference is the
> limits of our grayscale vision.
>
> > The difference should be more pronounced in a PMT drumscanner with the
> > higher dynamic range.
>
> Shouldn't matter, as far as the limits of our grayscale vision go.  A CCD
> scanner can still have a "range" of around 4...and remember, a .3 increase
> in density range (or dynamic range) doubles the amount of data.  It may
not
> matter as much for some of the higher density B&W films, but for Tri-X (my
> favorite), which has a below 2 density range, it can make a big
difference.

> > In practice I doubt you can use true analog
> > gain on a CCD filmscanner, there's no margin left at both sides
> > of the tonal
> > scale.
>
> I have done exactly that (analog gain in a CCD scanner), and it works
quite
> well in fact.  What you are up against is actually noise.  You can only
> discern down to noise, going below that doesn't give you any more good
data.

That "range" of around 4 is an arbitrary ranking based on what is considered
to be acceptable noise. Some consider 3 as the highest possible with CCD
scanners. In fact the same noise barrier you mention in your last sentence.
That's where PMT drum scanners are superior, the range is wider / noise
further away. So analoge gain can then be used without the penalty of noise.

With Tri-X the density range will fall within most CCD scanner's dynamic
range (maybe excluding a strong development with Rodinal as I used to do
ages ago), but the tonal separation will not be better with analog gain than
with any stretching of the digital data after the A/D
step. The grain just isn't fine enough. More an estimation on my part than a
researched fact but some 35 mm Tri-X scans with the Nikon 8000 I made last
week makes me think so.

I have followed the thread on what the eye still can see of tonal steps in a
piezoprint. I can't judge that part by experience with piezo printing though
I doubt the eye is limited to 256 steps. There's something that is added by
a real smoothness of tones that probably can't be explained by the eye's
capacity to distinguish tones and resolution. Like in intanglio aquatints
the eye seems to zoom in at detail to build an impression of the total
image. I hate to utter the term "Gestalt" but what I try to describe can't
be far off that concept. Like one will add the knowledge of the texture of
the print that one has touched or seen from nearby to the view at a distance
of the same print.

Ernst