Digital BW, The Print

"Paul D. DeRocco" <pderocco@i...> wrote:
> As to your last question, I doubt having a large number of bands would be
> useful in color film scanning.

It wouldn't. Because of the nature of the dyes used in film, you can get  pretty much everything the film has to offer with three narrowband filters. Throw in a fourth band, infrared, for dust and scratch detection (the dyes in color film are basically totally transparent to IR, dust and scratches aren't) and you're all set.

> It would be very cool to have a camera with lots of bands, though.

High spectrum cameras are a lot of fun. I got to work with two different systems. One had the filter wheel from hell in front of a monochrome CCD. 10nm filters, 31 of them. Took about a minute and a half to run through the whole wheel. That one had a 1000x1000 CCD. A rare and beautiful antique. (probably about 1995). I could easily put together something with todays technology that could run that 31 color wheel in 3 seconds flat at 6mp instead of 90 seconds at 1mp.

Another was a slit scanning camera. The whole camera rotated, the lens projected an image that was masked off into a vertical slit, which focused on a horizontal diffraction grating, which then gave you vertical location along the y axis and wavelength along the x axis. That one had a DALSA 6mp imager, which was running about 20 images/second. The camera stepper motor drive ran 20 steps/second, and it could image a 2000x2000 scene in 100 seconds, with 151 band (2nm) or even 301 band (1nm) color resolution. That's probably near state of the art. It's enough to do some forms of chemical analysis at a distance, in addition to color perception. ;)