I was wondering if other people are having the same experience that I have.  I've found that setting the Quickfields Whole Page OCR to "Accuracy" actually results in poorer accuracy than "Speed" or "Balanced".  

For example, using the "Accuracy" setting, I was looking to pattern match a 9 digit identification number.  On about 400 documents, there were 53 failures, despite the documents all being quite clean. Some were bizarre - numbers from the 9 digit id number added to the word above, and in one case numbers added to the ID number that I cannot find the source of.  Over this morning, I've run the 53 through until I got down to 9 files that could not be recognized, despite being clean documents.  The issues with these files were consistent across Quick Fields processing attempts.

I ran the same set of files through using the "Speed" setting, and there were no failures in identifying the 9 digit id number, all the files worked perfectly.

Is this something other people have observed as well?