Why MV-HEVC (Spatial Video) Takes Longer Than Full SBS

Spatial Video (MV-HEVC) conversions take longer than Full Side-by-Side conversions. This is normal and expected — here is why.

Full SBS is a one-stage process:

The AI depth estimation generates the 3D pair, packs the two eyes side by side, and encodes a single MP4 file. The output goes straight to your dashboard.

MV-HEVC is a two-stage process:

1. The same AI conversion produces the side-by-side stereo pair on our GPU servers.

2. A second stage repackages the stereo pair into Apple's MV-HEVC container using Apple's encoding stack. This stage runs on dedicated Apple Silicon hardware to ensure the output is fully compliant with Apple Vision Pro and visionOS.

The extra stage adds a few minutes per video — typically the equivalent of an additional pass at roughly 1× real-time. For a 10-minute clip, that is about 10 extra minutes of processing on top of the standard conversion time.

If you need the fastest turnaround, choose Full SBS. The 3D quality is identical — the only difference is the container format and which players support it.

During busy periods the queue for the Apple Silicon stage can grow longer than the GPU stage, which is why MV-HEVC sometimes appears to "stall" near 80% on the dashboard. The stall is the queue wait, not a failure. Your video will continue automatically.