Today I’ve seen a strange effect in DT 3.0beta3:
I’ve scanned a voucher with text rotated 90° on the right side with my Scansnap iX500. The scan looks fine in ScanSnap Home but in DT3 after OCR parts are missing.
It is 100% reproducable. I can even import the original PDF into DT3 without OCR and it looks fine. Then manually do the OCR and the PDF changes to one with the missing part.
Thanks for the report! This is a known issue and will likely be addressed in the next beta release. Thanks for your patience and understanding.
Thanks.
Is there a special patter in the scanned document that triggers the issue, like rotated text, bright colors or blank areas?
If it is completely arbitrary, I’d stop using OCR within DT3 and OCR in Scansnap Home until the issue is fixed as it is quite dangerous if content gets lost unnotified during OCR within DT3.
Hmm, issue is not fixed in new beta 4. 
As I said…
This is a known issue and will likely be addressed in the next beta release.
There is an issue in the framework we have to investigate further with the licensor.
It is strange, that the issue is not present, if I do the OCR in the Scansnap Home App and disable OCR in DT3. I thought OCR is done by the Abby Finereader engine in both apps. So why is present only in DT3?
I thought OCR is done by the Abby Finereader engine in both apps.
This does not mean the same version is in each application.
On a side note, even the consumer version isn’t the same as the version we can license as third-party developers.