https://livekit.io logo
Join Slack
Powered by
TranscriptSynchronizer audio output was detached w...
# ask-ai
a
TranscriptSynchronizer audio output was detached while text output is still active; transcription sync is disabled. This usually means session.output.audio was replaced after AgentSession.start().
✅ 1
t
Thanks for reaching out, Sebastián. I'm an AI assistant and I'm here to help. Give me a moment while I research this topic.
Thanks — I’m responding to the TranscriptSynchronizer message about audio output being detached while text output remains active. This happens when 
session.output.audio
is replaced or disconnected after 
AgentSession.start()
, which breaks the linkage required for synchronized transcriptions. You can fix this in two ways: • Keep the original audio output instance attached for the lifetime of the session (do not reassign 
session.output.audio
after 
AgentSession.start()
). • If you must alter audio handling, disable transcription sync by setting 
sync_transcription=False
in 
TextOutputOptions
, or use a solution that preserves the main output (for custom playback use 
BackgroundAudioPlayer
or implement audio processing that does not detach 
session.output.audio
). Also verify audio output is enabled in room options (e.g., 
RoomOptions(audio_output=True)
or 
outputOptions: { audioEnabled: true }
) and, if you need to detect playback completion, listen for the 
playback_finished
event on 
session.output.audio
. If you want, tell me: • whether you are reassigning 
session.output.audio
in your code (paste the snippet), • which SDK (Python/Node) you’re using, and I’ll give exact code changes. Sources: How to detect when an agent has finished speaking | Text and transcriptions | LiveKit Documentation | Agent session | LiveKit Documentation | External data and RAG | LiveKit Documentation, with interactive elements
a
i don't need sync transcription but anyways, never seen this issue before
t
Thanks — noted that you don't need synced transcription and that this is the first time you've seen the warning. You can safely ignore the warning if you disable transcription sync. To do that, set 
sync_transcription=False
in your 
TextOutputOptions
(or the equivalent option in your SDK). After that, transcription will continue without attempting to align with audio. If you’d like, tell me which SDK you’re using (Python or Node) and I’ll paste the exact code snippet to disable it. , with interactive elements
a
python agents sdk 1.5
t
Thanks — I’ll show how to disable transcription sync in Python Agents SDK v1.5. Use 
TextOutputOptions(sync_transcription=False)
when creating your text output (or when starting the session). Example:
Copy code
from livekit import AgentSession, TextOutputOptions, RoomOptions

room_opts = RoomOptions(audio_output=True)
text_opts = TextOutputOptions(sync_transcription=False)

session = AgentSession.create(room_options=room_opts, text_output_options=text_opts)
session.start()
If you already have an active session, ensure any text output creation uses 
sync_transcription=False
. Do you want a snippet that shows disabling it on an existing session or where in your code to place this change? , with interactive elements
6 Views
Open in Slack
PreviousNext
Powered by