May 2000. Over the last two weeks I have scanned and edited the records of marriages of and baptisms from 1683 through 1809. That's 731 pages published by the Holland Society, not counting the indices, and about 18 thousand records (rough estimate).

I'm not completely satisfied of the accuracy, but I have to send the books back to the library, so this will have to do. Also, I haven't scanned the indices, due primarily to lack of time.

A few libraries have the original Year Books, but a collected edition was also printed in 1978 by the Genealogical Publ. Co. in Baltimore entitled Records of The Reformed Duch Church of Albany, New York, and that edition is likely to be more readily available then the original Year Books.

Dave Pane-Joyce, Worcester, MA Genealogical home page: http://aleph0.clarku.edu/~djoyce/gen/ Email address: djoyce@clarku.edu

Report of the process of scanning and editing the text. I used my ScanMaker X6 (cost of less than $100) to scan the pages, and the OmniPage Limited software (included with the scanner) to convert the scanned pages to text. My computer is an iMac, and I used an old wordprocessor WriteNow to edit the text. (I like WriteNow because it's easy to use.) After editing it, I converted it to html for the web. Once in html, I used the text editor BBedit Lite, to tweak things to make them look better. This page, for instance, was completely written using BBedit.

The scanned text was full of errors, primarily inconsequential ones like mixing up the digit 1, capital I, and small l, or extra spaces in the middle of words. But there were also substantial errors like scanning Buuren as Bunren. I'm sure I fixed nearly all the errors, but some will remain—things like v where y should be—because the eye misses the error for the same reason OmniPage missed it. And spell checkers don't work on this kind of text for two reasons. First, 95 percent of the words aren't in the spell checker's dictionary, and second, every name is spelled at least half a dozen ways, some more than two dozen.

I took me about an average of a minute to scan and convert each page to text, and less than three minutes, on average, to edit each page. Toward the beginning it was more ten minutes per page, but since each correction can actually be made to the whole text, later pages were faster. For instance, after seeing a couple of Bunrens that should be Buurens, I learned that was a common conversion error, spot checked a few other Bunrens, then, with one substitution, converted every Bunren in the entire text to Buuren. Toward the end of the text, there were only a couple of errors per page, and things really sped up. The conversion to html and final editing took only about two hours in all. I estimate it took me about 50 hours in all. The real work, and it wasn't much fun, was fixing conversion errors.

How to use this document. If you're going to really use these records, you'll want to save them on your own computer because it takes too long to access them over the web. (Fair use allows you to create a copy of anything for yourself, but does not allow you to distribute copies to others). Use the search feature on your web browswer to find names. Don't put in the whole name, but just distinguishing parts of it. For instance, if you're looking for someone named Quackenbush (and there were a lot of them in Albany), search on 'Qua' or 'Kwa'. Check the preface to the index of part 3 for alternate spellings of surnames.

D. Pane-Joyce
May 2000. All rights reserved
Located at http://aleph0.clarku.edu/~djoyce/gen/albany/refchurch.html