Originally Posted by Delawaredave5
So documents I've been scanning I've been scanning to PDF with text recognition.
I was scanning documents with text recognition enabled until I noticed that sometimes the text recognition was imperfect, and the resulting PDF file contained perfect versions of the incorrectly recognized characters, with no remnant of the original correct characters!
So I might scan something like: "Total $1,234.56" and the PDF might read "Total SI.Z34.S6" to use a made up extreme example. I would far rather have a fuzzy photocopy style image than a visually crisp but incorrect PDF of my records.