Familyrelatives.com

[Muggins in Sussex]

I took out a subscription to this as most of my research is still in the 20th Century :D and it claims to have transcribed all BMD from 1866 -1920.

It is usually an excellent site, but twice I have found things on Free BMD when I could not find them on Family Relatives. - It may just be be mistranscriptions, but just wanted to say to anyone else who has joined it, it's worth looking on other sites as well.

[KiteRunner]

I think they used OCR software to create their 1866-1920 database rather than human transcribers, Joan. They picked that date range because the typed indexes started in 1866.

[Muggins in Sussex]

Thanks, Kite - what is OCR, please?

[Merry Monty Montgomery]

OOoh, I know that!! Optical Character Recognition.

It means they use a computer to read the type. That means some letters might be misread even when they don't make sense, such as the name Jamas for James if there was too much ink on the typewriter ribbon to make the e convincing!!

[KiteRunner]

I suspect they had a program to check that the names it came up with were proper ones and maybe the ones that failed the check were flagged up for real people to sort out, because there don't seem to be too many mistranscriptions on there.

[Muggins in Sussex]

Am well impressed, Merry, as my son would say :D:D

[Mary from Italy]

Most scanners have an OCR program, so if you scan a page of type you can OCR it and then edit it in Word.

I sometimes wonder if Ancestry's used something of the kind for its more outlandish mistranscriptions.

However, a decent OCR program also has a spellchecker.