I'm curious how it would handle some tough cases.
Take this: Multi-column portrait-orientation .PDF, no images, but copious mathematical/logical symbols, textboxes, and quasi-graphical elements like pointer arrows:
Example 1
What would it do with that?
That might be an unfair test.
For a less intense workout, what about this:
Example 2
Mainly text, but still some unusual symbols here and there, Greek letters used as variables, footnotes, etc.
(I own the copyright to those, so if you wish to experiment with me, go nuts!)
I'm not really expecting great results with any method for these; images are probably the best I could hope at present. I'm mainly just trying to keep tabs on what's presently possible and what isn't.
Thanks for all the responses so far...