Documentation contents

Documentation

Voiceover and dubbing workflow

Timed speech generation, take review, assembly, mixing, and media export.

  • New users
  • Operators

Pandrator stores this as a voiceover session and uses the dubbing processing pipeline. It begins with transcription, correction, and optional translation, then adds:

  • optional whole-document or generation-time speech optimization;
  • per-segment speech generation with reviewable takes;
  • deliberate take selection and audio assembly;
  • optional synchronization, source-audio ducking, or a dubbing-only mix; and
  • export of audio, subtitle tracks, or rendered video.

Generated segments are not automatically a finished mix. Inspect selected voices, provider readiness, timing constraints, and output mode first. Source media, subtitle text, and voice samples may be sensitive; a workflow plan should say which external providers receive each kind of data.

For subtitle-only outcomes, use a subtitle session and stop before speech generation. For narration without timed media, use an audiobook session.