1. 01

    Transcribe

    Speech to text with speaker diarization, so who said what survives the trip.

  2. 02

    Translate

    Per-segment, with a glossary for the terms that must not drift.

  3. 03

    Re-voice

    Each speaker keeps a voice of their own, cloned from their own audio in the source.

  4. 04

    Fit

    Translated lines rarely match the original length. Timing is planned per segment so the result still lands on the picture.

  5. 05

    Export

    Burned-in or sidecar subtitles, remuxed video, or the audio alone.