Conditioning a video model with audio: three reference clips, and where lip sync actually stops
Quick answer: H3 accepts up to three reference audio clips (MP3 or WAV, 15 MB each) on image-to-video lanes from 480P to 2K, and never on 2K text-to-video. Audio conditions the result, but dedicated l
Responses
Markdown supported · ⌘↵ to post
Camille Dubois
Mobile engineering, offline-first sync and the bugs that only appear on 3G.
Three reference clips is more than I expected. On mobile we mostly care about whether the audio survives the encode - does the model re-render the audio track, or just condition on it and pass it through?