Thanks Lars. I had not realised that scanners designed to cope with film as well as ordinary documents were now so cheap. The CanoScan 4200F only costs about 60 to 80 pounds in the UK (about 90 to 115 euros). I assume the F in 4200F signifies that it can scan film. Has anyone had any luck scanning microfiche of printed/typed material with optical character recognition (e.g. typed monumental inscription transcripts)? Some commercial pdf copies of old published transcripts have very poor OCR and are hardly worth searching because so many words have been misinterpreted, so I am not optimistic. William "Lars Erik Bryld" <larserik@dadlnet.invalid> wrote in message news:k0r0srof1uoz.dlg@lebryld.fqdn... > Scripsit William Kirk: > >> it works well with a high resolution scanner that has the option of >> scanning negatives and transparencies and provides light from >> above.I use an Epson Perfection V700, which costs just over 300 >> pounds in the UK > > You *can* obtain quite useful results with a considerably cheaper > scanner. I have made very legible scans using a CanoScan 4200F. It > only scans up to 3200 ppi, but it's often adequate. > > -- > Regards > Lars Erik Bryld

On Thu, 11 Oct 2007 10:46:45 +0100, William Kirk <nospam@home.com> wrote: > Thanks Lars. I had not realised that scanners designed to cope with film as > well as ordinary documents were now so cheap. The CanoScan 4200F only costs > about 60 to 80 pounds in the UK (about 90 to 115 euros). I assume the F in > 4200F signifies that it can scan film. As time goes on, good scanners keep getting cheaper. My Epson 3200 DPI film scanner works OK for written but not so good for technical microfiche of typed characters (engineering drawings, etc). > Has anyone had any luck scanning microfiche of printed/typed material with > optical character recognition (e.g. typed monumental inscription > transcripts)? Some commercial pdf copies of old published transcripts have > very poor OCR and are hardly worth searching because so many words have been > misinterpreted, so I am not optimistic. OCR is one of those things that seems like would be easy to get good but, it doesn't seem to be getting there. I'd rather have an image of the type sampled as well as you can, to give me the best chance of reading it. Takes up more room to store but, storage is cheap these days. Raw scans in TIFF format take a lot more room than .jpg but, it's not a lossy format like jpg is.