If you look in the cctlk archives there was work done on this in the recent past
http://www.retrocmp.com/tools/pdp-11-diagnostic-database seems to be a pointer to many sources On 10/4/16 9:38 AM, Rod Smallwood wrote: > > > On 04/10/2016 17:16, Al Kossow wrote: >> >> On 10/4/16 8:02 AM, Shaun Halstead wrote: >> >>>> I'd like to work to-wards scanning all of the library into a system. >>>> Anybody know anything about fiche scanners. >>> >>> This is exactly what my company does (and my previous, now defunct, >>> employer as well). I have dedicated microfiche and microfilm scanners. >>> The problem I ran into when trying to scan my DEC microfiche collection was >>> that the fiche themselves were of very poor quality. Badly scratched and >>> scuffed, and poor quality duplicates. >>> >>> Jim mentioned that dedicated microfilm scanners are expensive. He's not >>> kidding. Used machines tend to start around $5k USD for a low end, heavily >>> used scanner. New machines are generally $30k at the lowest end. Flatbed >>> photographic film scanners are much cheaper, but also slower. However, >>> with some scripting and tools like the ImageMagick library, you can scan an >>> entire fiche in one pass, then slice up the image into individual pages >>> accordingly. >>> >> many people have jousted at this particular windmill >> this is a huge time-consuming project >> >> I have literally thousands of sheets of DEC fiche from multiple sources >> >> scanning at that kind of volume doesn't scale. you have to clean an qc every >> sheet >> even with a production scanner. It took me most of a day to do the little >> bit of xxdp >> fiche a couple of weeks ago on a manual positioning Canon >> microfilm/microfiche scanner >> at CHM >> >> I just bought another high-end scanner, which isn't running yet. It uses >> glass carriers >> for each sheet and I have about 6 carriers. >> >> I was hoping it could handle IBM punched-card sized fiche, but I've not been >> able to >> find a carrier that big. I have a rather large backlog of those as well. >> >> >> > Not a surprise. > I have a box; fiche size front and about two feet deep marked xxdp. > Perhaps we could avoid duplication of effort by knowing whats already been > scanned. > > The other issue is access to the scanned data. > > I'm sure there are hardware hounds out there who could come up with cost > effective way to scan DEC 52x fiches. > > Rod > >
