Repository navigation
Conversation
Clear the matching per-destination suppression state when a completed telephone-event is deduplicated at media scope. Otherwise later listeners suppress speech until the timestamp-discrepancy timeout expires. Add repeated-burst coverage for AMR-WB and Opus listeners in both destination orders, with blocked DTMF, prompt playback, and an event sink.
…equires AMR-WB encoding unavailable in standard CI
vingarzan
force-pushed
the
fix/dtmf-fanout-suppression
branch
from
October 9, 2026 14:09
611bdad to
242de7c
Compare
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 one RTP source is transcoded to several listeners, the end of a telephone-event can leave speech suppressed on all but the first listener. A live AMR-WB phone-to-phone/browser trace showed continuous browser speech while the second phone had a 4.204-second output gap. Which listener is affected depends on destination processing order.
packet_dtmf_event()deduplicates completed events through the shared media'sdtmf_ts, but speech suppression is stored in each destination's primary input handler asdtmf_start_ts. The first handler processes the END and clears its state. Subsequent handlers return through the duplicate-event path without clearing theirs.packet_decode()then suppresses speech until its 80,000-tick timestamp-discrepancy fallback expires, corresponding to five seconds at the source's 16 kHz clock.Clear the matching input handler's suppression state in the duplicate-event path. Match the event timestamp so a retransmission of an older completed event cannot clear suppression for a newer active event. Keep completed-event reporting deduplicated.
Add a regression with AMR-WB input, AMR-WB and Opus listeners, a PCMU event sink, directional blocking, prompt playback, repeated DTMF END packets, and two talk bursts. Exercise both listener orders and assert timely, continuous speech RTP after END.