Skip to content
Email Support
  • There are no suggestions because the search field is empty.

Driving Data

Learn how Driving Data can power your vub

 

What is Driving Data?

Driving Data is the new performance that will be transferred to your Source media.

The Driving Data performance can come from new dialogue audio (e.g. from your dailies or ADR), or a new visual performance (e.g. from an alternate take or a different shot). With this is mind, the Driving Data can be either Video Driving Data or Audio Driving Data:

Using Video or Audio as Driving Data

You can use either video or audio as Driving Data.

  • Video Driving Data:
    • If you use video as your Driving Data, DeepEditor will use the video to 'copy and paste' the actor's performance onto the Source Media. The emotional content is therefore also sourced from the Driving Data
    • You cannot upload a video file with no audio, since you will need the audio to listen to should you choose to refine the results.
  • Audio Driving Data:
    • If you want to use audio as your Driving Data, DeepEditor will use that audio to generate the new mouth movements, while attempting to retain the original emotional performance of the Source media.
    • The audio should contain only the voice of the actor you want to vub and be clean of music and sound effects.
  • Do not use video or audio of a different actor as Driving Data – the performance should be driven by the same artist.

Before exporting and uploading your Source media, check that your media meets the file format and codec requirements for DeepEditor.

 

Video Driving Data: Choosing Between Visual and Audio Performance

When using video Driving Data, the actor delivering the new dialogue must be clearly visible and should not be occluded at any point during the dialogue.

In the example below, the video Driving Data can be used successfully to vub the blonde actress.

Screenshot 2025-03-28 at 16.04.31

If your video Driving Data media is not ideal visually, you do have the option of selecting to use the audio in the file to drive the vub instead:

Screenshot 2025-07-08 at 16.16.08

 

When using video Driving Data, only the face track is extracted from the file. This means:
  • The video does not need to be at online conform resolution, a lower-resolution proxy is perfectly fine.
  • The lighting, makeup, head pose or look of the actor in the Driving Data does not need to match the Source.
The two key requirements for video Driving Data are:
  • Clear visibility of the actor’s face – The face must be fully in frame and unobstructed during the new dialogue.
  • Matching duration and frame rate – The Driving Data file must have the same frame count and frame rate as the Source.
As long as these two conditions are met, DeepEditor can accurately track the facial performance from the video file and apply it to the source.

 

Read next: Making Source Media and Driving Data Work Together