How do I use microfilm for library research? (multi-point question)

April 8, 2010 10:20 AM

How do I use microfilm for library research? (multi-point question)

Hello,

I'm interested in conducting library research on a specific american public figure, who was active in government in past decades.

In particular, I'm thinking of publishing about this person and his correspondences. (Sort of like 'Selected Letters of General George Washington', although GW isn't the person. But you get the idea...)

There is a library maybe a day's drive from myself, that has documents on said figure, in microfilm.

The library in question is willing to loan out microfilm to other libraries. I think I could make arrangements to get the microfilm loaned to a local library.

OK, but then what?

We're talking about up to several thousand documents that potentially I'd be interested in, to review & cull down to a smaller set.

Best case would be for me to somehow get the microfilm scanned electronically, and then go through some kind of automatic conversion to MS-WORD or some other text format, that then could be easily incorporated into a word processor for further work.

Do scanner/converters like this exist? Where would I find one? Are these common at any library?

Or would a scanner simply convert microfilm to Jpeg or some other image format, and if I wanted to convert it to work processor format, I'd need to type it in myself?

Any advice is appreciated.

Thanks

posted by thermonuclear.jive.turkey to Writing & Language (11 answers total)

Microfilm scanners do exist, but they're not cheap and your local library's not too likely to have one. For converting to editable text in a Word (or similar) document, you'd also need OCR software; the effective applications are, again, not too cheap, and in my experience are fairly patchy a) on pre-1900 material and b) on material like newspaper print that tends to use very small type with very little white space, if these apply to you.

Honestly, unless money is really no object for you, I think you might be best doing at least the review-and-cull the old-fashioned way - sitting in front of the microfilm reader and going through the documents yourself.
posted by Catseye at 10:30 AM on April 8, 2010

Even if they lend the microfilm to your local library, they're probably not going to allow the film to leave the building.

Did you ask the library that's a day's drive if they have a microforms (fiche/film) scanner? Is it a public library, or a university? Most university libraries probably have at least one machine that scans from microfilm. There are varying set ups and software, and quality and ease-of-use can vary. My library's machine is set to scan to images (.tif or .jpg), and we use PDF Writer to turn sets of images into .pdf. We don't have OCR software readily available.

I usually make hard copies (since I'm going to need those, anyway), and then scan the printouts from a campus photocopier, and it e-mails me .pdfs (rather than printing the scans, later). A Kinko's or Staples will scan paper copies to .pdf, but you're still stuck with getting the docs OCRed.

For a project like this, it might be worth investing in a copy of Adobe Acrobat to do the OCR.
posted by steef at 10:55 AM on April 8, 2010

You should also check with the library that has the microfilm to see if there is already some sort of finding aid or index to the microfilm, or to the person's papers.
posted by interplanetjanet at 11:35 AM on April 8, 2010 [1 favorite]

Also as the other posters mentioned many university libraries have a film/fiche to PDF scanner (we have one), but they can be pretty slow. I've converted about 150 pages with ours and that was about my limit in time and patience.
posted by interplanetjanet at 11:37 AM on April 8, 2010

If the library has a machine that will scan to digital for you, like an Canon MS300, it will take as much time to scan a page as it does to make a photocopy of a page -- it takes at least 5-10 seconds per page to line up the image and start it scanning after you've spent a few minutes searching for the page, which, if you're planning on getting hundreds of pages, will be a several-day project for you; expect, at the fastest, 70 pages per hour. There are companies (disclaimer: like the one I work for) who do bulk scanning, but that would require the film leaving the grounds of the library. If you're planning on publishing, though, rather than casual research, you might want to discuss in detail with the library what you're intending to do and what you need to accomplish your ends. You may be able to work out some arrangement where the library entrusts the film to you on the condition that you provide them a free copy of the digital scans when you're finished with the film; lots of libraries wish they could digitize their microfilm, but the time and effort involved is prohibitive (if there was a quick way to drop microfilm into a machine and get a nice, clean digital version, the library probably would have already done it). And, to echo the above, scanned microfilm generally produces a jpeg or tif; if you want typeable text, you need an OCR step in between. Also be aware that OCR'ing bad film may result in so many errors that typing it out by hand could be faster.
posted by AzraelBrown at 12:08 PM on April 8, 2010 [1 favorite]

OK, but then what?

Well, first make sure that the local lib you use to get the microfilm has a microfilm reader, and preferably one with a printer attached. If there's a library in town that has a reader with a scanner attached (not likely unless you go to an academic library—I work in the periodicals archive of a relatively ritzy public library and we don't have one.) Like it's said above, you probably won't be able to take it out with you so you'll have to use the resources available there.

Hopefully you find one with a printer attached, at least. So then next you'll spend a goodly amount of time scrolling through the microfilm, squinting at a blurry bright screen, and selectively printing out the documents you want. Then, if you're here with me, you'd step one machine over and put your hard copies on the OCR scanner, which will scan directly to a PDF or OCR quality .tiff/.jpeg on your flash drive.

There's a gentleman who comes in here to do the same kind of project you're working on, and this is the process he has to go through. Sorry, but unless you've got a super-awesome mega-rich and also-friendly academic library nearby (possible!) you're pretty much stuck doing non-automatic research.
posted by carsonb at 12:27 PM on April 8, 2010 [1 favorite]

If there's a library in town that has a reader with a scanner attached... use that library!
posted by carsonb at 12:29 PM on April 8, 2010

Most university libraries have microform scanners, and will let you use their equipment as a visiting researcher. You'd probably have to get the microform sent there, though.

It will be a long, tedious process. You might want to hire a temporary assistant to do the scanning once you've found the stuff you want to scan.
posted by coolguymichael at 12:42 PM on April 8, 2010

The public library I work at has a digital microfilm scanner, so check with the owning library or your local library. Images are saved as a jpg to a usb drive. I did a quick test and I was easily able to save 10 sequential pages to the usb in less than a minute (about 50 seconds), so you can do the math. I had a patron do an entire 600+ book in an afternoon. But again this was sequential pages. As for getting them OCRed afterwards, that is something you might want to look at paying to have done if it really is going to be in the thousands of pages.
posted by Razzle Bathbone at 1:39 PM on April 8, 2010 [1 favorite]

Just a heads up that you need to be upfront - 1st because you need access to the right equipment but 2nd because there may be a copyright issue. I'm no lawyer but I'm pretty sure you can't just scan an entire publication. Maybe other researchers can explain if this has been done before, perhaps in a digitization agreement with the library that owns the collection as mentioned above.

An index, if available would be immensely helpful. That way you can find things of interest and print or scan as you find them. Making copies of relevant parts is a normal, copyright-kosher part of research, copying an entire publication because you don't know what you want yet probably isn't.
posted by Gor-ella at 1:59 PM on April 8, 2010

It would be really good to get into the habit of asking these questions of reference librarians. No one writes this kind of book without the help of fantastic librarians. Well, hardly ever.
posted by bluedaisy at 7:57 PM on April 8, 2010