PDF to EPUB Conversion - best practice?

[Verdi-Fan]

Good evening!
First of all, a big compliment to the creator of Caliber Software. A wonderful program that I am discovering more and more for myself.
A question about conversion:
I get the best results when I convert a PDF file to an *.epub file .
Vice versa its a catastrophe and that's not the point here either

PDF to Epub is a very very good result. Does anyone have experience with how to optimize the results? So a smaller font size so as not to produce so many pages, etc. I always get an incredible number of pages. Are there any instructions or tips for this? Thank you very much!
Kind regards
VF

[aborel]

Surprising post: your observation is opposite to the usual experience and the official FAQ https://manual.calibre-ebook.com/con...-pdf-documents . Converting to PDF usually works better than PDF to anything.

Now if your PDFs produce something reasonable, by all means, go for it :-) However, you might have more control on the output if you convert to a fully editable format (docx, odt or RTF), fix all issues using your word processor of choice, and then convert to epub.

[Ma'am-I-Am]

Now that Calibre has updated its ability to convert PDF to other formats, I do have one question:
After 38 minutes converting PDF to epub (I think the message said "converting to html"), progress was at only 1%. I aborted because my backup would have interrupted it anyway. Any insight on why it takes so long?

[retiredbiker]

Originally Posted by Ma'am-I-Am

Now that Calibre has updated its ability to convert PDF to other formats, I do have one question:
After 38 minutes converting PDF to epub (I think the message said "converting to html"), progress was at only 1%. I aborted because my backup would have interrupted it anyway. Any insight on why it takes so long?

Since nearly anything can be hiding inside a pdf, it could be many things. One consistent problem seems to be complexity. People have reported waiting days or weeks trying to convert dictionaries or bibles, for example. If your pdf is reasonably small, simple, and has a clean text layer it should not take too long on modern hardware.

Take your pdf and try to highlight and copy-paste some text out of it from your pdf viewer. If that does not work, you are in trouble already. And if the text does copy out but is not very good, it is probably from an iffy OCR (Optical Character Recognition) done on the images, and the only way to fix that is to proofread and edit it.