Message-Id: <199803221901.MAA21568@dns.ccit.arizona.edu> Date: Sun, 22 Mar 1998 13:58:12 -0500 From: Stephen Chapman <mailto:stephen_chapman@HARVARD.EDU> Subject: Re: Overhead scanner choice To: mailto:IMAGELIB@LISTSERV.ARIZONA.EDU
Mr. Otto,As background reading on the subject of overhead scanners, I highly recommend the following paper, "The Internet Library of Early Journals," available on line at http://www.bodley.ox.ac.uk/ilej/papers/paper01.htm. This article was also published in ASLIB Proceedings 49:(6) June 1997, 153-158.
Like all equipment choices, the best one for you depends upon your application: are you, for example, scanning to create page images only or full-text? Will cropped images and/or excess borders surrounding the image be acceptable in your project? Do you want to scan illustrations as well as text? Any color?
As you point out, each of these devices represents a substantial investment. In my opinion, they should be compared to studio digital cameras, which are increasing in number and decreasing in price. (See, for example, Sound Vision Inc.'s CMOS-PRO Digital Camera. Specs are available at http://www.soundvisioninc.com/cmospro.htm.) Like the overhead scanners, digital cameras can accommodate bound and oversize materials. They also have the added advantage of grayscale and color capabilities. (Their resolutions vary, according to the size of your documents.)
There's a fair amount to say on the topic of what is best for a given library scanning application -- flatbed scanners (e.g., scanning from photocopies of bound vols.), overhead scanners, or digital cameras? -- but I think that three issues are relevant when comparing equipment to scan bound materials:
(1) price -- initial and ongoing costs (i.e., the service contract) of the scanning workstation (scanner + computer)
(2) value (i.e., return on price) -- usually defined in terms of image quality and throughput (speed)
Assuming that "legibility" is your main requirement in the digital image, consider that the University of Leeds reports production rates of 75-125 pages per hour with the PS3000 scanner. Our experience with the same equipment results in similar numbers. In fact, although there are a number of qualifications to report when giving production statistics, in general I would caution you against assuming that you will do better than 100 pages per hour with any of the overhead scanners you mention. When it comes to production -- your cost per page -- keep in mind that this isn't like flatbed scanning, photocopying, or microfilming. The reason?
(3) document handling
Again, I refer to the Internet Library of Early Journals report. Note that in order to minimize the number of rescans, Leeds puts glass on top of their open volumes in order to keep pages flat. Systems Integration Group, one of the National Digital Library contractors, also uses their PS3000 scanners in a "non-standard" configuration.
Even with 400 dpi resolution and the curvature correction features of the PS3000 scanner, consistent image legibility from page-to-page is not easy to achieve, especially with older materials. When there are peaks and valleys in the printed page, wavy text often appears. We're experimenting with a different cradle -- similar to one we use in microfilming -- and hope to have better success in production.
_________
Cautions?
Talk to the vendors offering these products and make sure that they understand ALL of the requirements of your project. Having them demonstrate their product with your materials in a real-world situation is the best learning exercise for everyone.
It may seem counter-intuitive that photocopying books and scanning photocopies could actually be less expensive than using an overhead scanner for "direct scanning," but don't rule out this approach too early. If you have access to a book-edge photocopier, you could, in theory, purchase a relatively inexpensive flatbed scanner, automatically feed the pages to the scanner, and meet your project requirements at the lowest possible cost.
Where quality requirements are higher, be sure to compare the overhead scanners you mention to digital cameras. Whatever you use, I think you will find that the main challenge for the scanning technician will be to keep the material positioned in the right place from page-to-page, especially when there is a narrow inner margin.
Good luck,
Steve Chapman
At 03:14 PM 3/20/98 -0800, you wrote: >Greetings,
>
>Our institution is considering the purchase of an overhead scanner
>soon. To my knowledge there are only three makers of overhead
>scanners: Minolta, Zeutschel, and Image Access. We have been
>particularly interested in the Minolta EPIC 3000. The Zeutschel
>Omniscan is far more expensive (too expensive for us) and I only
>recently learned of the Image Access Bookeye scanner, which does not
>have the curvature correction software that the Minolta does (though
>Image Access says this is coming soon).
>
>Does anyone on this list have experience with any of these scanners? In
>particular, has anyone used the Image Access Bookeye scanner? It is the
>one about which I know the least. Any recommendations or cautions?
>Considering that any of these constitutes a substantial investment, any
>advise would be greatly appreciated.
>
>--
>
>James A. Otto
>Digital Projects Specialist / ERes Manager
>The Libraries of The Claremont Colleges
>800 Dartmouth
>Claremont, CA 91711
>(909) 607-7530 (Special Collections)
mailto:>jotto@rocky.claremont.edu >
>"In fine, nothing is said now that has not been said before."
> --Terence, Eunuchus: The Prologue, v.41.
>
>
Stephen Chapman Preservation Librarian for Digital Initiatives Preservation Center, Harvard University Library Holyoke Center 821 Cambridge, MA 02138Phone... 617-495-8596 Fax...... 617-496-8344 E-mail.. mailto:stephen_chapman@harvard.edu