Recognize a page made only of images
If you can see words in the PDF but Text is empty or unhelpful, the page may simply be a scanned image. That kind of page has no dependable text layer ready to send directly to a speech engine, even though it looks readable to you.
Use on-device OCR in FolioTone to recognize the writing. Start with a clear page containing reasonably straight lines and little decoration. Results depend on scan quality. A completed recognition operation means text was produced, not that every letter and sentence has necessarily been captured correctly.
Review the places that invite mistakes
Read the recognized passage before listening. Names, Vietnamese diacritics, numbers, parentheses, and italic text deserve extra attention. A single wrong word can change the meaning of a sentence even when the synthetic voice delivers the mistaken wording smoothly and with convincing confidence.
Check paragraph order as well as spelling. Page numbers, running chapter titles, and notes near the margins can enter the middle of a sentence. Editing that text before generation is more useful than trying to repair a completed audio export that already contains the unwanted words.
Use a short audio sample as a final check
Select a voice matching the passage language and listen to a sample containing a longer sentence and punctuation. If pauses sound strange, return to Text and check for lost periods or line breaks inside sentences. You are testing the source material as much as the voice.
Keep FolioTone in the foreground during generation. OCR and speech are local operations, but a longer job still needs time and storage. Locking the screen while speech remains unfinished should not be treated as a way to complete the remaining generation automatically in the background.
Process the pages your session actually needs
If only one part of a document matters today, recognize and listen to those pages first. This reveals layout problems early and avoids spending a whole preparation session on a scan whose text order is not yet suitable for a useful spoken version.
Free has OCR and batch-preparation limits displayed in the app; Pro removes the corresponding limits. Either way, keep the original PDF and compare it when wording matters. Recognized text supports your reading workflow, but it should not replace checking the actual document for important details.
TRY IT WITH YOUR OWN WORDS
Your next chapter starts here.
FolioTone works on iPhone and iPad with iOS 18 or later. Voice, OCR and export limits depend on your Free or Pro plan; check the app for current limits.
Get FolioTone on the App StoreCompare Free & Pro