[MacDV] Re: Film Scanners

[Richard Brown]

I have been reading this string for a while. As someone with something 
like a million frames, when I include transparencies, negatives, and 
digital stills shot over the past 25+ years, I have pondered this same 
question.

The need for mass scanning is really about cataloging, not about making 
thousands upon thousands of prepress or print-ready images. As such, I 
would suggest no scanner would suffice. In my work in stills 
photography, I had a job once to duplicate a series of slide shows 
concerning plastic surgery techniques. There would be 50-150 slides or 
so, per set, and they had to be duplicated very quickly, no more than 
30-60 minutes per set of slides, and when one was done, the next would 
be on the table for duping.

This was back in the days before CCD's, before even the desktop 
scanner, let alone Coolscans or Sprintscans. We had to dupe original 
slide to dupe slide, on slide duping film. We simply established a 
population color balance for the dupe media using a Nikon slide 
duplicator in front of a darkroom dichroic color head, and used 
moderate exposure bracketing, based on the condition of the given 
original, all based on settings determined for normal, underexposed, 
and overexposed sample images. Having had a bunch of experience with 
slide duping, we even added some improvements to some of the originals 
when possible. We found we could dupe these shows with time to spare. 
There was no second chance, as the slides left with the doctors, same 
day. The "operation was a success" and was completed without breaking a 
sweat.

Today, with a 35mm CCD camera, a slide duping adapter, and either a 
strobe or color head based light source, and with computer or LCD 
confirmation, I would think to digitize 35mm slides could  be 
accomplished at the rate of 200-400 per hour (maybe even more for large 
numbers of slides with the same characteristics), at, say, 6 megapixels 
with moderate compression. This would engulf 10,000 slides in a long 
weekend. But it would be done. The trick would be in the physical 
ordering of the slides plus intelligent bulk renaming of the digital 
records of same. With a bit of cleverness, the entire process could be 
made quite smooth, with the digital record useful in identifying a 
given image which might later be scanned properly for printing or other 
purpose. Properly executed, however, the images certainly could be of 
enough quality to use in a DVD title. You must keep in mind that 
non-High Def television is of astonishingly low resolution. The dupe 
process I speak of will be far in excess of TV resolution, and with 
good down sizing and prep for TV, should be fine for this purposing. 
Using the high res files would allow panning, zooming, and other 
effects were a fancier slide show on DVD required.

The caveat I might see in this process would be if the originals varied 
wildly. The real difference between professional photography and snap 
shooting is in the control of exposure and the predictable, repeatable, 
surprise-free outcome when the film is processed. Still, even with wild 
variation, this system would record an image, and if using a continuous 
light source, the variation may well be tamed through automatic 
exposure on the camera.

Richard Brown