Thoughts on drum scanners for scanning microfiche. The best scanners are supposed to be drum scanners. ***WARNING more than you ever want to know about film scanners*** I was once approached by a printing prepress company that used drum scanners that wanted to diversify and scan 14 X 17 inch x-ray film for teleradiology. Here is a little of what they said about drum scanners. 1. They were accustomed to images that were each 1-6 GB. 2. They do the printing of the "slicks" inside the Sunday papers. 3. Each color is on a separate negative. 4. Three colors and black. 5. They use about 3000-8000 dpi, real not interpolated resolution. 6. The negative film is mounted on a clear glass tube. 7. There is a standard light source inside the tube. 8. The tube spins on air bearings and a detector system progresses along the length of the tube measuring the light transmittance of each point in the negative. 9. Transmittance is much more accurate than reflectance. 10. This is not a rapid process. However it is very accurate. 11. Similar systems are used for scanning aerial photographs and magazine printing negatives. 12. Mounting the negative on the drum is very important, any skew will distort the image when the colors are overlaid. This step is very time consuming. 13. I looked into building a simpler version, and determined that the higher the speed of the scan the lower the quality. We needed speed and high quality 14. Now if you want a current model SCSI scanner that could scan microfiche, I suggest a mammography film scanner. High resolution and relatively speedy, however high cost. The shape of the little microcalcifications in a mammogram are very hard to see with anything less than 10 micron spot size scanner. The images are about 5K X 5K over a 8 X 10 inch area. Other scanning systems. Dupont made a laser scanning system for x-ray films, that was the size of a washing machine, about 8 years ago that used a laser beam and would scan a 14 X 17 inch film in 10 seconds and produced a 4K X 4K image. You dropped in the film and the image was scanned and then the film dropped out. However it took the PDP-11 controlling the scanner about 1 minute to transfer the image to a MicroVAX II which then transferred the image via decnet to another MicroVAX II controlling a display. Lots of failure points, the hardware was solid, the software was unable to detect and retry simple timeouts. Especially neat when trying to send images across Kansas on POTS lines during spring weather. Scanners for aperture cards may be another option. Mike mmcfadden(a)cmh.edu All scanned out.