Inquiry Regarding Camera Synchronization with Fusion API and Independent Setups

Software Python
, ,
1

Dear StereoLabs Team,

I hope this message finds you well. I’m currently working on a project involving two ZED cameras and have questions about synchronization behavior when using the Fusion API versus standalone setups. Here’s my scenario:

  1. Setup 1: Two cameras subscribed to the Fusion API and connected to the same system.
  2. Setup 2: Two cameras running independently (not via Fusion API) but connected to the same PC.

In both cases, I’m using grab() to retrieve frames. I’ve observed that timestamps between frames (e.g., frame x from Camera A vs. frame x from Camera B) show a jitter of ~10ms to 30-40ms. My questions:

  • Temporal/Spatial Sync:

    • Are the cameras in Setup 1 (Fusion API) synced temporally by default, or is this limited to spatial alignment via calibration?
    • In Setup 2 (independent cameras), does connecting to the same PC inherently enable any hardware/software synchronization, or is the jitter expected due to USB bus limitations?

  • IR Sync vs. Environmental Light:
    I attempted to synchronize frames using an IR flash to reset timestamps, but sunlight interference causes false triggers. Is there a recommended alternative for light environments?

  • Trajectory Accuracy:
    My goal is to compute a 3D object trajectory by fusing data from both cameras. Given the observed jitter, what best practices or Fusion API features (e.g., enablePositionalTracking()) would improve robustness?