Video Mocap Studio

Workflow overview

The Video Mocap tab in Studio is the default workspace for converting a short source clip into motion outputs.

1. Upload a source video

• Upload a clip smaller than 100MB
• Keep the main subject clearly visible
• Use a single continuous shot without hard cuts
• Avoid heavy blur, severe occlusion, or fast scene switching

2. Trim the useful motion segment

After upload, select the start and end range that contains the motion you actually want to process.

• Trim away idle time before the motion starts
• Trim away extra frames after the motion ends
• Keep the selected segment as short and clean as possible

3. Choose output targets

You can send the same clip to multiple output targets.

• Default: metahuman-style humanoid output
• Unitree G1: robot-oriented motion output
• Published custom characters: humanoid FBX characters that already passed the binding workflow

4. Submit and wait for processing

Once the source clip, trim range, and output targets are ready, submit the job to the queue.

• Credits are consumed when the upload is complete and the job is admitted into the runnable queue
• Processing speed depends on the queue and selected outputs

5. Preview and download results

When the job finishes, open the result page to preview the generated outputs and download the final files.

Recommended source video requirements

Use clips that are appropriate for video mocap:

• clear image quality
• one continuous shot
• full or mostly visible body motion
• readable limb movement
• minimal camera shake
• no heavy motion blur
• no fast cuts or transitions

Bad source footage usually leads to unstable motion, foot sliding, or missing limb detail.