john dean wrote:
> Bruce,
>
> I pretty much agree with your conclusions. However, I find color
> transparencies to scan at the higher ppi resolutions MUCH smoother
> than film does, like say Tri X or even color neg film, especially very
> fast color film.
>
Not surprising. Graininess is directly related to density. With
trannies, the density is in the shadows. The highlights of trannies
therefore have the much less graininess (smallest dye clouds and the
loosest spacing of the dye clouds) than the shadows by definition. So
they darn well better scan smoother!
> I also think this varies with the particular scanner. On my old
> Howteks I actually see smoother less grainy rendition when scanning
> the same output of 4x5 at 2000 ppi vs 4000 ppi.
Two different apertures. You trade away a little sharpness to gain a
little smoothness when you increase aperture size. The 2000 spi scan is
indeed a little smoother. This is one of the differences between school
two and school three, and why school three has such strong adherents.
> That is evaluating the
> print from those scans as well as the file on screen. A lot depends on
> how large you have to go with the print of course and how much
> sharpening you have to do.
>
> I noticed on the DPL software on an Aztek drum scanner( which I
> personally don't have) that they have recommended settings that
> actually scan negs at lower ppi than one would want for a particular
> file size when printing at 360 dpi or so. In other words upsampling a
> smaller file size scan actually was used in that case to achieve a 360
> dpi print resolution.
>
Yes, I've heard this several times from DPL users. There are many paths
to the waterfall, and this is the path that Aztek is pushing. I don't
know if it's right or wrong for Howtek hardware / software or for any
particular piece of film, but it sounds like my school three (low rez
scanning) which again many people support.
I've also heard from a few Howtek operators that they routinely open up
one aperture size from what the scanner normally computes for them. The
reason seems to be that the Howtek scanners (optics, circuits, firmware,
software, or some combination) creates a scan file that is what one guy
called "crunchy" (his word, not mine) which I interpret to be a high
level of local contrast. This in turn shows up graininess fairly well.
The larger aperture lets more light reflect onto the edges of the grain
clumps / dye clouds and therefore makes the edges less visible and the
grain clumps look a little smaller (like moving from a condenser
enlarger to a diffuser enlarger in the darkroom IIRC). Some operators do
this routinely, some do it when they think it's necessary, and some
won't do it at all.
Just to be clear I should point out that this applies to all drum
scanners regardless of make, not just Howteks. I don't mean to imply
that there is anything inherently wrong with a Howtek scanner, it's just
the scanner that John is talking about here. Aztek who makes the Howtek
scanners is one of the last three drum scanner manufacturers left
standing so they certainly are doing something right!
> I think the quality of the software used, the quality and age of the
> scanner used, and the type of film all play a role but.. my overall
> impression is that I never scan anything with higher scanner ppi than
> I am going to need for a particular print size. In the end, I think
> you have to get to know the qualities of your own equipment.
>
> John
Absolutely. There are way too many variables here, just in the machines,
not to add the software, film, photographers' exposure workflows,
processing, etc. and the vagaries of scanner operators (oh yes, I'll
admit to that!) to give a definitive answer to any of these kinds of
questions. So all I'm trying to do is point the OP in the direction in
which the truth for the OP probably lies. Hopefully. Maybe.
--
Bruce Watson