I think the issue is one of pragmatics. All CCD scanner have some noise in shadows. It appears though that some keep that noise near or below the black point, and therefore it becomes basically moot. Further still, in isn't as if multipass scanning is "free". It adds 2-3-4 or more times the amount of time the scan takes to accomplish. It becomes a matter of loss/benefit ratios.
If I had a scanner which was particularly noisy in an area of the scan that was visible, (in other words within the scanner's ability to create meaningful gradients) then certain problematic images might be worthwhile using these methods to improve. BUT, if the noise was basically near the black point floor (even if that black point was high due to poor dynamic range) then the advantage is questionable. The SS4000+ I have worked with exhibits minimal noise in shadows, and none that I could see in highlights. The green channel is minimally noisy on the one I used. I don't expect the advantage in image quality would be large enough to justify running the scanner 2-4 times more per scan, the extra time the scans take, the extra light exposure to the film, and the risk of offset registration. Even on the Minolta Dual II, which does have a noisy blue channel (more about this in another posting), I got no more detail from a 4 x or long scan. It was slightly less noisy in the blue channel but the problem was that the blue channel gives up well before the true black point, and that wasn't helped by extra scans. I probably would affect a similar result by a simple levels output adjustment on the blue channel in Photoshop, taking a heck of a lot less time, and less wear and tear on the machinery. Art Eric wrote: > David: > > >>You have to have noise to be able to reduce it. Noise or the lack of it is a >>function of hardware design and in some cases profiles. >> > > Does a noiseless scanner exist? > > Why wouldn't oversampling and averaging help reduce hardware-caused noise? > Hardware problems can often be somewhat fixed in software, after all. > > > Eric > > > ---------------------------------------------------------------------------------------- Unsubscribe by mail to [EMAIL PROTECTED], with 'unsubscribe filmscanners' or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or body
