ensure playback_segments_count is consistent in the audio output chain#4211
Merged
ensure playback_segments_count is consistent in the audio output chain#4211
Conversation
theomonnom
approved these changes
Dec 9, 2025
Member
theomonnom
left a comment
theomonnom
left a comment
There was a problem hiding this comment.
lgtm! So our pattern breaks if one of the nodes in the output chain decides not to forward audio to the next node (for whatever reason)
I thought this pattern was supposed to work — ideally, we should make it work.
meetakshay99
added a commit
to meetakshay99/agents
that referenced
this pull request
Dec 12, 2025
* main: (267 commits) AGT-2328: negative threshold in silero (livekit#4228) disable interruptions for agent greeting (livekit#4223) feature: GPT-5.2 support (livekit#4235) turn-detector: remove english model from readme (livekit#4233) add keep alive task for liveavatar plugin (livekit#4231) feat(warm-transfer): add sip_number parameter for outbound caller ID (livekit#4216) fix blocked send task in liveavatar plugin (livekit#4214) clear _q_updated right after await to avoid race conditions (livekit#4209) ensure playback_segments_count is consistent in the audio output chain (livekit#4211) fix inworld punctuation handling (livekit#4215) Inference: Rename fallback model name param (livekit#4202) fix race condition when stop background audio play handle (livekit#4197) fix watchfiles prevent agent prcoess exit on sigterm (livekit#4194) feat(google): add streaming support for Gemini TTS models (livekit#4189) Add LiveAvatar Stop Session API Call + README Fix (livekit#4195) Fallback API for Inference (livekit#4099) feat(rime): expand update_options to accept all TTS parameters (livekit#4095) mistralai models update (livekit#4156) fix record.exc_info is not pickable when using LogQueueHandler (livekit#4185) Restore otel chat message (livekit#4118) ...
This was referenced Dec 20, 2025
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
when
next_in_chainis set inAudioOutput,await self.next_in_chain.capture_frame(frame)should be called beforeawait super().capture_frame(frame)(in case the capture_frame of next_in_chain is waiting for some conditions) to make sure the audio outputs in the chain have the same playback_segments_count.this fixes the issue when the agent has a greeting message using session.say or session.generate_reply in Agent.on_enter (or just after session started), it will stuck when closing if no participant joined the room.