II.
InteractionPrimitive overview
Reference · liveinteraction-primitive:avatar-lipsync
Avatar lipsync (voice-synced mouth) overview
The animated character's mouth/jaw morphs are driven from TTS viseme timing (Azure VisemeReceived / ElevenLabs timestamps) scheduled against the same-page audio clock, so the published video track lip-syncs the published audio track. Fast path; never a tool call. See docs/research/realtime-voice-agent-stack.md §II.2.
Attributes
displayName
Avatar lipsync (voice-synced mouth)
kind
video
surface
video
requiresPermission
false
description
The animated character's mouth/jaw morphs are driven from TTS viseme timing
(Azure VisemeReceived / ElevenLabs timestamps) scheduled against the same-page
audio clock, so the published video track lip-syncs the published audio track.
Fast path; never a tool call. See docs/research/realtime-voice-agent-stack.md §II.2.
Outgoing edges
realizes1
- layer:10-interaction·LayerInteraction
Incoming edges
None.