« To Import external fIles: Overview
When uploading a media file that you recorded outside of Riverside, you can split a single video file into multiple tracks based on speaker detection (also called video diarization). This gives you layout flexibility and helps you repurpose long recordings such as video conference calls or shared-camera interviews.
Video diarization works only on uploaded videos that meet the following conditions:
- Fixed layout: The video must keep the same layout from start to finish. If the layout changes during the session, separate video tracks cannot be generated.
- Consistent group: The group of participants must stay consistent. If someone joins or leaves mid-session, the file cannot be processed.
- No screen sharing: Videos that include screen sharing cannot be separated into individual video tracks. The changing layout prevent accurate speaker detection.
Use Cases
In this section, we’ll show examples of recording setups that work well for separating video tracks, and one example where the setup makes separation less effective.
| Setup | Works for separating tracks? | Why |
|---|---|---|
|
Zoom or other video call (without screen share)
|
|
Each speaker appears in a consistent grid layout, making it possible to detect and split their video tracks. |
|
Two people in the same room spaced apart |
|
Even though they share one camera, their positions remain fixed and distinct, so the system can separate their tracks. |
|
Two people in the same room sitting close together |
X |
Speakers are too close and overlap visually, making it difficult to separate their tracks. |
Learn more