Skip to content

Conversation

voidspark2025
Copy link

@voidspark2025 voidspark2025 commented Aug 19, 2025

Add feature: Drain data in jitter buffer before close tts channel,or the last few packages in jitter will get lost when unimrp-server return SYNTHESIZER_SPEAK_COMPLETE event!
Here’s how to reproduce the issue.

  1. Add follow configurations to your tts mrcp profile which you used.
    param name="playout-delay" value="80"
    param name="max-playout-delay" value="100"
  2. Call any apps which related to tts. I used speak application.
  3. When unimrp-server return SYNTHESIZER_SPEAK_COMPLETE event, switch speech channel state to SPEECH_CHANNEL_DONE in mod_unimrcp.c.
  4. The residual data in jitter buffer can't be written into audio queue because speech_channel_write only writes data into audio queue in SPEECH_CHANNEL_PROCESSING state!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant