How to dub a video with multiple speakers

Before you start

• Use a permitted video with distinguishable speech, and note the order in which people first speak.
• Choose Dub or Voice clone and compatible languages and voices.
• Speaker detection is an aid; overlapping or very short replies still require review.

Set the speakers

1. Add the source and select Dub or Voice clone.
2. Review Speaker detection. Auto detect estimates speakers; explicit multiple-speaker controls can expose count and mapping choices.
3. Where separate voice controls are offered, assign speakers in the order they first speak. Review optional Speaker mapping rather than relying on raw labels.
4. Check each selected voice's target-language compatibility and reference guidance.
5. Create the job, review its quote and confirm when ready.
6. Listen to the first turn of every speaker and any overlapping exchanges in the finished video.

Expected result and quality check

The output should let you follow who is speaking. Distinct voices do not prove suitable casting or correct assignment. Listen to names, questions and short replies as well as longer turns.

Overlap limitation

An uncertain secondary reply in overlapping speech may be omitted while primary speech is retained, with a review warning. Check the exchange in context; do not present a warning-bearing result as a complete verbatim transcript.

If a speaker sounds wrong

Check the audible-order choices in the submitted settings. Do not silently assign the same voice to everyone or submit repeated paid jobs to guess. Contact support with the job reference if the delivered mapping needs investigation.

Explore next