I'm looking for help to minimize how many times QF will return wrong parts of account numbers (returns ° for zeros or I for 1's among other things).  I did find some hints in this post:  https://answers.laserfiche.com/questions/60040/Character-Preference#60446

I do already have the Advanced Settings for character preference as "Numbers".  I have now tried changing the Confidence setting to 95% and Median instead of Average, but that didn't make any difference.

I would like to explore the suggestion of using Substitution.  How would I set a substitution up on my zone ocr or Account Number so that it can always exchange a 0 for every ° and a 1 for every I?

So, I built the Substitute tokens, and proven that they work, however, how do I get the template field to know which token to use to fill the field?  I ran my test leaving the original token for the Acct No OCR collection in there, plus the two substitution tokens for the two most common mistakes I wanted it to correct for.  But of course, now it is filling the field with the account number three times. 

In this test result, you can see that the mistake in the first number is corrected in the third number (the spot where the third token filled in the corrected number zero).

I understand that if I only had one substitution going, I wouldn't have a problem here.  The problem is that I want two different things corrected.  Is there a way around this?