On 12/31/02 12:02 PM, "Randy Wilson" <WilsonR at fonix.com> wrote: > We agreed that one of the main solutions is going > to involve a 2-dimensional CCD. Otherwise, the higher the resolution, > the more scans per inch have to be done, and thus the slower things will > be. With a good high resolution 2-D CCD, it should be possible to do > one scan (or perhaps 3 if separate snapshots are taken for red, green > and blue) instead of having to do over 5000 individual scan lines > separately. That could speed the scanning part up by over 1000 times, > potentially (i.e., instead of 14 minutes for a full-blown super-sampled > scan, it could be done in .84 seconds). Couple that with a slide feeder > as good as the 1957 model, and you could scan seriously fast. This is something to think about. The latest digital scanners now have image sensors that are 100% the size of a 35mm negative. They also have resolutions of, I think, about 14 megapixels (see new Kodak pro digital camera announced at most recent Photokina for details). Wouldn't it be interesting to see someone develop a digital "contact printer", where film emulsion was physically flattened against such a high resolution sensor? I can see where this would solve one problem, but create another. In conventional motion picture contact printing, each source image is mated with a fresh clean frame of the contact print. It would seem to me that cleaning the receptor CCD -- already a major consideration with interchangeable lens digital cameras, would be an issue. Side note about Nikon film scanners: sometimes when I inspect slide scans at very high resolution, I detect misalignments. Occasionally, one row of pixels is shifted out of place. Maybe this is my device, but it happens more than you might expect. Danny Grizzle