Mask what is in front of the mouth (a microphone, a hand) on the dubbed shots' face crops, so the next dub keeps it in front. Uses the GPU (Florence-2 + SAM…

Lip-sync for music videos in ComfyUI: dub the singers' mouths to the song, shot by shot, with QA. Every input is a slot (bring a clip and song, or generate them with LTX-2.5, Qwen-Image-Edit and ACE-Step); vocal split, lyric alignment, lip kits, sync QA.

LSDub

project

render

phrases

shot_ids

Mask what is in front of the mouth (a microphone, a hand) on the dubbed shots' face crops, so the next dub keeps it in front. Uses the GPU (Florence-2 + SAM 2.1, weights in models/lsdub).

Name	Type	Default	Description
project	LSDUB_PROJECT		The LSDub project.
render	LSDUB_RENDER		A dub (its shots' face crops are masked).
phrases	STRING	microphone	What to mask, comma-separated (e.g. "microphone, hand").
shot_ids	STRING		Only these shots, e.g. "19,51" (empty = every dubbed shot).

Name	Type	Description
masks	LSDUB_MASKS	Per-shot occluder masks; feed back into LSDub Dub.
review	VIDEO	The first shot's mask overlay.

LSDub Occluders

Inputs (4)

Outputs (2)