PDF was made for printing: every page has a fixed size. On a digital reader or a phone, that forces you to zoom and drag the screen all the time. EPUB, in contrast, is text that reflows to fit the screen and the font you choose. Converting solves much of this hassle, but there are limits worth knowing.
When conversion works well
- Continuous-text books (novels, essays, short stories) created digitally, where you can select the text in the PDF.
- Documents with simple titles and paragraphs.
When the result may disappoint
- Complex layouts: magazines, textbooks with columns, boxes and side notes. A PDF does not store the reading order, only the position of each piece on the page, so the text may come out of order.
- Tables and formulas: they tend to lose their formatting.
- Scanned PDF: if the PDF is a photo of the pages, there is no text to convert. In that case, first use text recognition (below).
How to convert in Libtro
- Upload the PDF to My files (drag it to the upload area or tap to choose).
- Open Actions → Convert and choose EPUB.
- If you want a lighter file, check Remove images (lighter file).
- Follow it in the Queue and download the result, or send it straight to the Kindle.
Scanned PDF: text recognition (OCR)
When Libtro doesn't find readable text in a PDF, it offers Actions → Recognize text. OCR reads the images of the pages and creates a new EPUB in your library, which can then be translated or narrated. The result depends on the quality of the scan: crooked, blurry or stained pages cause more errors. Large PDFs can take several minutes.
Tip
If the book exists in EPUB from a legitimate source, such as those in the public domain, prefer downloading the original EPUB: it always turns out better than a converted PDF.