I used to think aliasing was the fault for grain enhancement, but now I
think the kohler illumination has more to do with it, as kohler
illumination enhances film base defects as well. Aliasing requires that
the sampling aperture be (spatially) narrow relative to the pixel size
it is sampling on film (that is, the area on film being sampled to
represent the pixel). Low pass filtering prior to sampling (eg. blurring
by defocusing) would test the alias theory (anti-alias filter), or
multi sampling should also anti-alias filter, assuming the multi-samples are
slightly spatially different. I don't know if that is the case though
for multi-sampling since the CCD sits in the same spot while it takes
the samples (spatial dithering would help multi-sampling in this case).
I have noticed a slight blurring effect when using ICE on an LS2000, so
there may be more going on with that than I understand. (do you notice
ICE reducing grain effects also?) If the actual aperture size of the
CCD sampler that generates the pixel sees the full "pixel area" on
film, that would tend to reduce high frequency grain aliasing, because
it would be averaging all the info within the pixel sample area. (Is
there a word for this "film pixel area"?? film sample area?). Seems to
me the CCD would want as wide a sampling aperture as possible to gather
as much light as it can. Bottom line, grain aliasing requires spatial
frequencies higher than the CCD (spatial) sampling frequency (spi,
dpi), and a narrow sample aperture.
It would also be interesting to note if the grain enhancement occurs in
both directions equally. I have heard that the stepper motor can
actually step finer than the CCD dpi, but not sure if the scanner will
allow software to control it. In that case you could step finer and
overlap the samples, that would eliminate aliasing in one direction.
But, I'm talkin too much. I think kohler illumination is the culprit. RDP3
for me.
Wayne
< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >
|