[Feature] Voice & Audio Generation - ElevenLabs Integration and Audio Track Management

[vzeman]

Summary

Integrate ElevenLabs for AI voice narration and build a per-scene audio management system. Audio is a critical component of professional video content — this feature covers voice synthesis, background music, and sound effects management.

Requirements

• ElevenLabs API client with API key configuration
• Voice selection: browse all ElevenLabs voices with preview
• Per-scene voice settings: voice, stability, similarity boost, style, speed
• Voice narration script field per scene (text that gets converted to speech)
• Generate voice narration for scene
• Playback voice audio in editor
• Auto-sync narration to scene duration (trim or extend scene)
• Background music track: upload audio file, set volume level, loop/trim
• Per-scene audio volume mixer (voice vs background music)
• Sound effects library (built-in or upload)
• Audio waveform visualization per scene
• Voice cloning support (ElevenLabs voice cloning API)
• Multi-language support
• Export audio-only option
• Mute/unmute per scene

Acceptance Criteria

• ElevenLabs API generates voice audio from script text
• Voice can be previewed before applying to scene
• Generated audio is cached locally and linked to scene
• Volume mixer correctly balances voice and background music
• Auto-sync trims or extends scene to match narration length

Related

Part of the ai-video-editor Flutter application. Used by #78 (Subtitle & Caption System). See also #75 (Cost Tracking).