Upload the song and every clip you shot. Premier maps the beats and bars, locks each take's own audio to the track so lip-sync survives the edit, splits your footage into shots, and builds a first cut that follows the scene brief you write.
Tempo, beats, bars and section changes are detected from the audio. Cut density follows the music: a chorus cuts every 2 beats, a quiet verse every 2 bars.
The scratch audio on each clip is matched against the track, then refined to a fraction of a millisecond. A synced cut keeps the performer's mouth on the vocal.
"Close-up, warm light, singing" or "wide, crowd, fast". Shots are scored on brightness, colour, motion, face size, your labels and optional AI captions, with the reasons shown.
Footage goes up in 8 MB pieces, three at a time, so 4K camera files of 10 minutes or more are not stopped by proxy size limits. Browser-friendly previews are made for you.