One instruction, one reviewable chain
- Install the desktop app and prepare the required local models
- Import a short source clip and correct the transcript
- Review the translation and generate the target-language voice
- Inspect timing, optional lip sync and the exported file