Merge pull request #1154 from pipecat-ai/aleix/twilio-telnyx-sample-rates

serializers: don't update twilio/telnyx sample rates

CHANGELOG.md

@@ -13,9 +13,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
   resampled to the desired output sample rate.
 
 - Fixed an issue with the `TwilioFrameSerializer` and `TelnyxFrameSerializer`
-  where `frame.audio_out_sample_rate` was incorrectly used in place of
+  where `twilio_sample_rate` and `telnyx_sample_rate` were incorrectly
-  `frame.audio_in_sample_rate`, which caused audio input detection to fail when
+  initialized to `audio_in_sample_rate`. Those values currently default to 8000
-  using different input and output sample rates.
+  and should be set manually from the serializer constructor if a different
+  value is needed.
 
 ## [0.0.55] - 2025-02-05
 

examples/canonical-metrics/bot.py

@@ -65,7 +65,6 @@ async def main():
             # English
             #
             voice_id="cgSgspJ2msm6clMCkdW9",
-            aiohttp_session=session,
             #
             # Spanish
             #

examples/chatbot-audio-recording/bot.py

@@ -82,7 +82,6 @@ async def main():
             # English
             #
             voice_id="cgSgspJ2msm6clMCkdW9",
-            aiohttp_session=session,
             #
             # Spanish
             #

examples/foundational/04-utterance-and-speech.py

@@ -51,7 +51,6 @@ async def main():
         )
 
         elevenlabs_tts = ElevenLabsTTSService(
-            aiohttp_session=session,
             api_key=os.getenv("ELEVENLABS_API_KEY"),
             voice_id=os.getenv("ELEVENLABS_VOICE_ID"),
         )

examples/foundational/08-bots-arguing.py

@@ -48,7 +48,6 @@ async def main():
             region=os.getenv("AZURE_SPEECH_REGION"),
         )
         tts2 = ElevenLabsTTSService(
-            aiohttp_session=session,
             api_key=os.getenv("ELEVENLABS_API_KEY"),
             voice_id="jBpfuIE2acCO8z3wKNLl",
         )

examples/phone-chatbot/bot_daily.py

@@ -8,7 +8,7 @@ from loguru import logger
 from openai.types.chat import ChatCompletionToolParam
 
 from pipecat.audio.vad.silero import SileroVADAnalyzer
-from pipecat.frames.frames import EndFrame, EndTaskFrame
+from pipecat.frames.frames import EndTaskFrame
 from pipecat.pipeline.pipeline import Pipeline
 from pipecat.pipeline.runner import PipelineRunner
 from pipecat.pipeline.task import PipelineParams, PipelineTask
@@ -99,14 +99,14 @@ async def main(
             - **ASSUME IT IS A VOICEMAIL. DO NOT WAIT FOR MORE CONFIRMATION.**
 
             #### **Step 2: Leave a Voicemail Message**
-            - Immediately say:  
+            - Immediately say:
             *"Hello, this is a message for Pipecat example user. This is Chatbot. Please call back on 123-456-7891. Thank you."*
             - **IMMEDIATELY AFTER LEAVING THE MESSAGE, CALL `terminate_call`.**
             - **DO NOT SPEAK AFTER CALLING `terminate_call`.**
             - **FAILURE TO CALL `terminate_call` IMMEDIATELY IS A MISTAKE.**
 
             #### **Step 3: If Speaking to a Human**
-            - If the call is answered by a human, say:  
+            - If the call is answered by a human, say:
             *"Oh, hello! I'm a friendly chatbot. Is there anything I can help you with?"*
             - Keep responses **brief and helpful**.
             - If the user no longer needs assistance, **call `terminate_call` immediately.**

src/pipecat/serializers/telnyx.py

@@ -31,7 +31,7 @@ from pipecat.serializers.base_serializer import FrameSerializer, FrameSerializer
 
 class TelnyxFrameSerializer(FrameSerializer):
     class InputParams(BaseModel):
-        telnyx_sample_rate: Optional[int] = None  # Default Telnyx rate (8kHz)
+        telnyx_sample_rate: int = 8000  # Default Telnyx rate (8kHz)
         sample_rate: Optional[int] = None  # Pipeline input rate
         inbound_encoding: str = "PCMU"
         outbound_encoding: str = "PCMU"
@@ -48,7 +48,7 @@ class TelnyxFrameSerializer(FrameSerializer):
         params.inbound_encoding = inbound_encoding
         self._params = params
 
-        self._telnyx_sample_rate = 0  # Fixed rate for Telnyx (8kHz)
+        self._telnyx_sample_rate = self._params.telnyx_sample_rate
         self._sample_rate = 0  # Pipeline input rate
 
         self._resampler = create_default_resampler()
@@ -58,8 +58,6 @@ class TelnyxFrameSerializer(FrameSerializer):
         return FrameSerializerType.TEXT
 
     async def setup(self, frame: StartFrame):
-        # Configure rates for input path: Telnyx (8kHz encoded) -> Pipeline (PCM)
-        self._telnyx_sample_rate = self._params.telnyx_sample_rate or frame.audio_in_sample_rate
         self._sample_rate = self._params.sample_rate or frame.audio_in_sample_rate
 
     async def serialize(self, frame: Frame) -> str | bytes | None:

src/pipecat/serializers/twilio.py

@@ -27,14 +27,14 @@ from pipecat.serializers.base_serializer import FrameSerializer, FrameSerializer
 
 class TwilioFrameSerializer(FrameSerializer):
     class InputParams(BaseModel):
-        twilio_sample_rate: Optional[int] = None  # Default Twilio rate (8kHz)
+        twilio_sample_rate: int = 8000  # Default Twilio rate (8kHz)
         sample_rate: Optional[int] = None  # Pipeline input rate
 
     def __init__(self, stream_sid: str, params: InputParams = InputParams()):
         self._stream_sid = stream_sid
         self._params = params
 
-        self._twilio_sample_rate = 0  # Fixed rate for Twilio (8kHz)
+        self._twilio_sample_rate = self._params.twilio_sample_rate
         self._sample_rate = 0  # Pipeline input rate
 
         self._resampler = create_default_resampler()
@@ -44,8 +44,6 @@ class TwilioFrameSerializer(FrameSerializer):
         return FrameSerializerType.TEXT
 
     async def setup(self, frame: StartFrame):
-        # Configure rates for input path: Twilio (8kHz μ-law) -> Pipeline (PCM)
-        self._twilio_sample_rate = self._params.twilio_sample_rate or frame.audio_in_sample_rate
         self._sample_rate = self._params.sample_rate or frame.audio_in_sample_rate
 
     async def serialize(self, frame: Frame) -> str | bytes | None: