BaseOutputTransport: send silence when EndFrame is received

CHANGELOG.md

@@ -9,6 +9,11 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### Added
 
+- Added `TransportParams.audio_out_silence_secs`, which specifies how many
+  seconds of silence to output when an `EndFrame` reaches the output
+  transport. This can help ensure that all audio data is fully delivered to
+  clients.
+
 - Added new `FrameProcessor.broadcast_frame()` method. This will push two
   instances of a given frame class, one upstream and the other downstream.
 

src/pipecat/transports/base_output.py

@@ -729,11 +729,24 @@ class BaseOutputTransport(FrameProcessor):
             else:
                 return without_mixer(BOT_VAD_STOP_SECS)
 
+        async def _send_silence(self, secs: int):
+            if secs <= 0:
+                return
+
+            sample_width = 2
+            silence = b"\x00" * self.sample_rate * sample_width * secs
+            silence_frame = OutputAudioRawFrame(
+                audio=silence, sample_rate=self.sample_rate, num_channels=1
+            )
+            await self._transport.write_audio_frame(silence_frame)
+
         async def _audio_task_handler(self):
             """Main audio processing task handler."""
             async for frame in self._next_frame():
                 # No need to push EndFrame, it's pushed from process_frame().
                 if isinstance(frame, EndFrame):
+                    # Send some final silence so words don't cut out.
+                    await self._send_silence(self._params.audio_out_end_silence_secs)
                     break
 
                 # Handle frame.

src/pipecat/transports/base_transport.py

@@ -83,6 +83,7 @@ class TransportParams(BaseModel):
         audio_out_10ms_chunks: Number of 10ms chunks to buffer for output.
         audio_out_mixer: Audio mixer instance or destination mapping.
         audio_out_destinations: List of audio output destination identifiers.
+        audio_out_end_silence_secs: How much silence to send after an EndFrame (0 for no silence).
         audio_in_enabled: Enable audio input streaming.
         audio_in_sample_rate: Input audio sample rate in Hz.
         audio_in_channels: Number of input audio channels.
@@ -131,6 +132,7 @@ class TransportParams(BaseModel):
     audio_out_10ms_chunks: int = 4
     audio_out_mixer: Optional[BaseAudioMixer | Mapping[Optional[str], BaseAudioMixer]] = None
     audio_out_destinations: List[str] = Field(default_factory=list)
+    audio_out_end_silence_secs: int = 2
     audio_in_enabled: bool = False
     audio_in_sample_rate: Optional[int] = None
     audio_in_channels: int = 1