What File Types Can Be Read Aloud?

PDF import produces a reading copy internally. That does not promise an exact spoken version of the page layout: a diagram or a table can be visible in the source while its relationships are difficult to express as continuous narration. Keep the original for reference. Password-locked PDFs need an unlocked copy you are authorised to use.

Export a clean reading copy from the application that understands the source. A plain-text editor can produce a PDF; a Markdown editor can render headings and lists before export; Word can save a PDF. For a book-length manuscript, a well-made EPUB may preserve navigation more usefully than page breaks. None of these steps requires renaming an extension: changing “.docx” to “.pdf” does not convert the document.

Read the relevant walkthrough for TXT, Markdown or Word documents. Keep the editable source. If you revise it later, export a new listening copy; importing a PDF does not connect it to the original editor.

For a DRM-free file, calibre's official format list includes MOBI input and EPUB output. Add the source, convert a copy to EPUB and inspect the chapters before importing it into LoudReader. Conversion quality depends on the source; it cannot recreate structural information that was never there.

Do not assume every Kindle purchase has the same download format or protection. What matters here is whether you have an accessible, DRM-free file supported by the converter. LoudReader does not remove DRM or open a purchased title simply because it appears in another app's library. The MOBI guide covers the conversion workflow.

LoudReader can recognise text on-device when at least half the pages lack text, or the total extracted text is very small. It attempts up to 300 image pages per import and reports partial or unreadable results. A mostly searchable document with a few image-only inserts may not trigger recognition, so compare those pages with the imported copy.

Faint scans, unusual typefaces and complicated layouts need closer checking. If an existing text layer is badly recognised, importing it does not guarantee those words will be corrected. Our scanned-book guide explains when external OCR or a cleaner edition is worth the effort.

They use different workflows. LoudReader can save supported web articles from links, which requires downloading the page. It also includes camera-based scanning on supported devices, with local recognition. These features do not mean the file picker accepts arbitrary HTML, JPEG, TXT or Word files. Use the import route intended for the source.

Once a book is imported and the desired voice is available, narration runs locally. Downloads, purchases and diagnostics are separate network activities; local speech does not make an online source available without first fetching it.

Prefer a clean EPUB for continuous prose with chapters. Prefer a text PDF when the document is supplied that way and you need its page layout as a visual reference. Use a scan when it is the best available source, accepting the recognition checks it needs. In each case, test a short passage with a heading and a page or chapter boundary before committing to a long listen. Our PDF import guide is the next step if you already have your file ready.