I have 2 observations using “Convert to Searchable PDF” –
(1) “Sometimes” resolution of the resulting document is reduced.
(2) When OCRing a selection of a previously OCRed document, the selection gets replaced by the previous full document.
Example for the latter: OCRing did not work properly with an insert in a document, so I cut out the insert image and pasted it to a new file and ran OCR on the insert again. The resulting searchable PDF was the original document and not the insert. Deleting and restarting did not help.
- How are you determining the “resolution” of the PDF?
I’ll rephrase: Upon OCRing the characters in the document get blurred.
- I’m not sure what you’re describing here. OCR works on a document-by-document basis, not a selection.
I scanned a double page from a book. The OCR as not satisfactory (e.g., inserts not properly OCRed). So I cut an insert from the original picture file, saved it as a separate file and OCRed this separate file.
Interestingly, the searchable PDF created from this new file was the scanned double page from before. I retried and deleted everything, but somewhere the OCRed double page was still available and again used as a searchable PDF for the insert.