How to Vub Multiple Characters in One Shot

DeepEditor changes the spoken performance of one character per vub. To vub multiple characters in the same shot, you generate a separate vub for each character, then use Merge Exports to auto-composite them into a single output.

Driving Data

You should already be familiar with source media preparation. The key difference in a multi-face workflow is your driving data.

Create a separate .wav file for each character you want to vub, each containing only the clean dialogue for that character.

For example if James and Eleanor are both in a shot and the dialogue for both characters is changing:

• DrivingData_Eleanor.wav - Eleanor's dialogue only
• DrivingData_James.wav - James's dialogue only

Workflow

Upload Media Upload your source media and all driving data files.

Create One Shot Create a single shot using your source media.

Add All Driving Data In the Shot Media tab, add the driving data files for all characters.

Review and Organize Characters In the Characters tab, review and name each detected character. It's important to check that all characters have been correctly identified before generating any vubs.

Generate a Separate Vub for Each Character Generate one vub per speaking character, selecting the correct driving data file each time. Name each vub clearly, including the character name is best practice.

Refine Each Vub (if needed) Refinement is not always required, but if it is needed, each character must be handled separately. You cannot refine multiple characters in a single session.

Merge Exports Once all per-character exports are ready, head to the Exports tab, select the exports you want to combine, and click Merge multiple exports. This is what brings everything together, a single composited output with all visually dubbed faces in one plate.