Advanced Capture unable to capture character correctly after upgrade

[Albert_De_Leon]

Our advanced capture is reading BAC as BAG. This started happening after we upgraded to OCR version 22.10.8522.498. I have tried combination of settings but it still reads BAG. we tried it on other words with C and its able to capture it correctly. 

Text:

Result:

Any help is greatly appreciated!

[James_Perry]

The image looks a bit grainy. Can you confirm that the documents were scanned at a resolution of at least 300dpi?

The suspect level is quite high. 

[Michael_Latchuk]

I had something similar happen to me. Not after an upgrade but just in general.

The value was always wrong in the same way, so for the specific form that was causing an error/misread I had a keyword lookup and replace check for the incorrect value and change it to the correct value in the event it was found.

[Denis_Hands]

We get this a bit with some of our stuff so I have done the following:

Making the capture zone just that little bit bigger makes a difference, a colleague made some templates with very tight capture zones where I had to resize them to something like the equivalent of one line before, after and half/one word before, after

Where that doesn't work:

Looking for/replacing the bad values with the right values in the types of documents where we know this happens using regular expressions eg search for BAC|BAG, DOCK|D0CK, DINE|D1NE|DLNE

I find it is a bit picky on some bolded or smaller fonts, so once again where in out control we have either made these a little larger or where a human doesn't need the information replacing things with barcodes

[Albert_De_Leon]

Thank you everyone for the response!

What we tested that seems to be working was to use "Machine print text/dot matrix printer", we moved it to production and so far its working, but we will continue monitoring it.

I will try the other suggestions as well see how that works.