When I ask pdf24 to perform OCR on a pdf, it gives me back a pdf with some (but not all) of the pages inverted black/white, that is, with black background and white text. Why is this happening? Is there...