text-to-speech

To convert a podcast script to speech, input the text file into the Edge TTS script. The synthesis process automatically parses the script, removing timestamps and background music annotations, and outputs natural-sounding audio.

Yes, edge-tts supports Chinese voice options with 18 distinct male and female voices. You can select and configure these specific voice outputs to read your text naturally.

Yes, you can control the rate, pitch, and volume when converting text to speech. The synthesis process provides a configuration file to adjust these audio parameters alongside the voice selection.

Script parsing automatically removes background music annotations, stage directions, and Markdown bold tags from the input. It retains only the actual read-aloud text to ensure clean speech synthesis.

The best way to apply voice changing post-processing is to optionally call voice-changer.skill after synthesis. This integrated workflow takes the generated audio and applies additional voice modifications.

You do not strictly need SSML to process emotional markers, but the workflow is SSML-ready. It handles emotional markers through configuration, allowing flexible processing during text-to-speech synthesis.