Plan AF: AI Audiobook & Podcast

ElevenLabs / Fish Speech

Preview 3-5 candidate voices and pick one that matches the content's tone. For Chinese, start with Fish Speech's stock voices; for long-form, prefer ElevenLabs.

💡Compare candidates reading the same opening line — don't choose from voice-list descriptions alone.

ElevenLabs / Fish Speech

Generate speech paragraph by paragraph and listen to each, fixing stress, phrasing and proper-noun pronunciation. Lock names and places in a pronunciation dictionary so the whole book stays consistent.

💡Fix unusual words in the dictionary before generating, or every segment will pronounce them differently.

Suno + Descript

Use Suno for background music or intro/outro, keep it under about -20dB as a bed. Assemble segments in Descript with fades and pauses, and strip flubs and noise.

💡Duck the music side-chained to the voice so music steps aside when the narration speaks.

Podcast platform + Descript

Export a standard audio format, design a cover and title, upload to podcast platforms or distribution. Add chapter timestamps and a transcript so listeners can skip and search.

💡Ship one episode through the RSS and review flow first, then schedule the batch.