• 01

  Tune playback speed to the task

  For comprehension-heavy material or new TTS users, start at 0.85–0.9x — slower delivery gives the brain time to process. Power users on familiar material often run 1.5–2x for catch-up reading. The right speed is task-dependent, not user-dependent; the same person may want 0.9x for a textbook and 1.6x for a news article.

• 02

  Pre-render important documents into audio

  Generate a WAV of any document you expect to revisit — meeting notes, a course reading, an immigration form, a healthcare letter. Save it in a clearly named folder so the audio companion is one tap away the next time you need to review the content without re-reading.

• 03

  Embed audio companions in PDFs and pages

  For accessible publishing, attach the WAV (or a 192 kbps MP3 you encode from it) alongside the document, or link it from the page. This creates an audio companion track that supplements — never replaces — proper structural markup, alt text, and headings. WCAG treats audio as an enhancement; the underlying document still needs accessible structure.

• 04

  Treat audio as a supplement, not a substitute

  Under WCAG 2.2, audio cannot replace alt text on images, accessible names on controls, semantic headings, or proper reading order. A TTS file is a supplemental modality. Build the page accessibly first, then add the audio companion for users who prefer or need to listen.

• 05

  Control pronunciation with punctuation

  Commas insert short pauses. Periods reset intonation. Em-dashes — like this — produce a longer pause. For tricky names or technical terms, spell phonetically (e.g., "Sjogren" as "show-grenz") in a parallel draft for the audio version. The original written document stays correct; the TTS-only draft handles the spoken edge cases.

• 06

  Switch language for bilingual content

  For documents that mix English with Spanish, French, German, Hindi, Italian, Japanese, Korean, or Portuguese, generate each language passage with a native voice from that locale and concatenate. This sounds far more natural than forcing a single English voice through non-English text, and supports learners who actively want to hear correct native pronunciation.

Screen readers and TTS file generators serve different needs. If you need real-time interface narration to navigate an OS, a screen reader is the correct tool. If you need a portable audio file from a document, FreeTextoSpeech fits the gap.