How duplicate detection works

We catch repeats by file size, duration, and GPS — even if the file name changed.

Same round, second upload — we catch it before it eats your storage. Here's exactly how the duplicate check works and what to do when it flags a non-duplicate.

What "duplicate" means here

A new file is considered a duplicate of an existing video if all of these match:

  • File size, within ~1 %
  • Duration, within ~0.5 seconds
  • GPS coordinates (if both videos have them), within ~100 m
  • Capture timestamp, within a few minutes

File name is deliberately ignored — phones rename files when you AirDrop or re-export them. Looking at the bytes is more reliable.

What you'll see

  • "Already in your library" — exact-match found in your own videos. Upload is skipped, you get a link to the existing video.
  • "Exists in [team library]" — match found in a team you're a member of. Upload is allowed but flagged so you don't accidentally double-store.
  • "Resuming pending upload" — match found, but the previous upload didn't finish. We pick up where you left off. See Resume a failed upload.

Trimmed videos look different

If you trimmed the same round to a shorter clip, the duration changes — the duplicate check won't catch it. Same for re-exports at a different resolution. That's the right behaviour for "this is a different cut", but worth knowing if you don't intend to keep both.

False positive?

If we wrongly flag a different video as a duplicate (rare, but possible if two rounds happened back-to-back at the same coordinates with similar duration), choose "Upload anyway" in the dialog — the upload proceeds as a separate video.

Was this article helpful?

Your feedback helps us prioritise what to write next.

Last updated April 25, 2026