wx44wx/pipecat
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

introduce PipelineParams audio input/output sample rates

Browse Source
This commit is contained in:
Aleix Conchillo Flaqué
2025-02-04 12:22:41 -08:00
parent cc54255c41
commit ab45e481be
61 changed files with 570 additions and 402 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
examples/bot-ready-signalling/server/signalling_bot.py
View File

@@ -17,7 +17,7 @@ from runner import configure
from pipecat.frames.frames import AudioRawFrame, EndFrame, OutputAudioRawFrame, TTSSpeakFrame
from pipecat.pipeline.pipeline import Pipeline
from pipecat.pipeline.runner import PipelineRunner
from pipecat.pipeline.task import PipelineTask
from pipecat.pipeline.task import PipelineParams, PipelineTask
from pipecat.services.cartesia import CartesiaTTSService
from pipecat.transports.services.daily import DailyParams, DailyTransport
@@ -31,16 +31,15 @@ logger.add(sys.stderr, level="DEBUG")
class SilenceFrame(OutputAudioRawFrame):
def __init__(
self,
audio: bytes = None,
sample_rate: int = 16000,
num_channels: int = 1,
duration: float = 0.1,
*,
sample_rate: int,
duration: float,
):
# Initialize the parent class with the silent frame's data
super().__init__(
audio=self.create_silent_audio_frame(sample_rate, num_channels, duration).audio,
audio=self.create_silent_audio_frame(sample_rate, 1, duration).audio,
sample_rate=sample_rate,
num_channels=num_channels,
num_channels=1,
)
@staticmethod
@@ -80,7 +79,10 @@ async def main():
return
await task.queue_frames(
[
SilenceFrame(duration=0.5),
SilenceFrame(
sample_rate=task.params.audio_out_sample_rate,
duration=0.5,
),
TTSSpeakFrame(f"Hello there, how are you doing today ?"),
EndFrame(),
]

1
examples/foundational/07g-interruptible-openai-tts.py
View File

@@ -37,7 +37,6 @@ async def main():
"Respond bot",
DailyParams(
audio_out_enabled=True,
audio_out_sample_rate=24000,
transcription_enabled=True,
vad_enabled=True,
vad_analyzer=SileroVADAnalyzer(),

1
examples/foundational/07k-interruptible-lmnt.py
View File

@@ -38,7 +38,6 @@ async def main():
"Respond bot",
DailyParams(
audio_out_enabled=True,
audio_out_sample_rate=24000,
transcription_enabled=True,
vad_enabled=True,
vad_analyzer=SileroVADAnalyzer(),

1
examples/foundational/07n-interruptible-google.py
View File

@@ -40,7 +40,6 @@ async def main():
"Respond bot",
DailyParams(
audio_out_enabled=True,
audio_out_sample_rate=24000,
vad_enabled=True,
vad_analyzer=SileroVADAnalyzer(),
vad_audio_passthrough=True,

7
examples/foundational/09-mirror.py
View File

@@ -21,7 +21,7 @@ from pipecat.frames.frames import (
)
from pipecat.pipeline.pipeline import Pipeline
from pipecat.pipeline.runner import PipelineRunner
from pipecat.pipeline.task import PipelineTask
from pipecat.pipeline.task import PipelineParams, PipelineTask
from pipecat.processors.frame_processor import FrameDirection, FrameProcessor
from pipecat.transports.services.daily import DailyParams, DailyTransport
@@ -61,7 +61,6 @@ async def main():
"Test",
DailyParams(
audio_in_enabled=True,
audio_in_sample_rate=24000,
audio_out_enabled=True,
camera_out_enabled=True,
camera_out_is_live=True,
@@ -78,7 +77,9 @@ async def main():
runner = PipelineRunner()
task = PipelineTask(pipeline)
task = PipelineTask(
pipeline, PipelineParams(audio_in_sample_rate=24000, audio_out_sample_rate=24000)
)
await runner.run(task)

8
examples/foundational/09a-local-mirror.py
View File

@@ -22,7 +22,7 @@ from pipecat.frames.frames import (
)
from pipecat.pipeline.pipeline import Pipeline
from pipecat.pipeline.runner import PipelineRunner
from pipecat.pipeline.task import PipelineTask
from pipecat.pipeline.task import PipelineParams, PipelineTask
from pipecat.processors.frame_processor import FrameDirection, FrameProcessor
from pipecat.transports.base_transport import TransportParams
from pipecat.transports.local.tk import TkLocalTransport
@@ -62,7 +62,7 @@ async def main():
tk_root.title("Local Mirror")
daily_transport = DailyTransport(
room_url, token, "Test", DailyParams(audio_in_enabled=True, audio_in_sample_rate=24000)
room_url, token, "Test", DailyParams(audio_in_enabled=True)
)
tk_transport = TkLocalTransport(
@@ -82,7 +82,9 @@ async def main():
pipeline = Pipeline([daily_transport.input(), MirrorProcessor(), tk_transport.output()])
task = PipelineTask(pipeline)
task = PipelineTask(
pipeline, PipelineParams(audio_in_sample_rate=24000, audio_out_sample_rate=24000)
)
async def run_tk():
while not task.has_finished():

2
examples/foundational/18-gstreamer-filesrc.py
View File

@@ -51,8 +51,6 @@ async def main():
out_params=GStreamerPipelineSource.OutputParams(
video_width=1280,
video_height=720,
audio_sample_rate=24000,
audio_channels=1,
),
)

2
examples/foundational/19-openai-realtime-beta.py
View File

@@ -80,9 +80,7 @@ async def main():
"Respond bot",
DailyParams(
audio_in_enabled=True,
audio_in_sample_rate=24000,
audio_out_enabled=True,
audio_out_sample_rate=24000,
transcription_enabled=False,
vad_enabled=True,
vad_analyzer=SileroVADAnalyzer(params=VADParams(stop_secs=0.8)),

2
examples/foundational/20b-persistent-context-openai-realtime.py
View File

@@ -177,9 +177,7 @@ async def main():
"Respond bot",
DailyParams(
audio_in_enabled=True,
audio_in_sample_rate=24000,
audio_out_enabled=True,
audio_out_sample_rate=24000,
transcription_enabled=False,
vad_enabled=True,
vad_analyzer=SileroVADAnalyzer(params=VADParams(stop_secs=0.8)),

4
examples/foundational/21-tavus-layer.py
View File

@@ -88,6 +88,10 @@ async def main():
task = PipelineTask(
pipeline,
PipelineParams(
# We just use 16000 because that's what Tavus is expecting and
# we avoid resampling.
audio_in_sample_rate=16000,
audio_out_sample_rate=16000,
allow_interruptions=True,
enable_metrics=True,
enable_usage_metrics=True,

1
examples/foundational/22d-natural-conversation-gemini-audio.py
View File

@@ -639,7 +639,6 @@ async def main():
vad_enabled=True,
vad_analyzer=SileroVADAnalyzer(),
vad_audio_passthrough=True,
audio_in_sample_rate=16000,
),
)

2
examples/foundational/26-gemini-multimodal-live.py
View File

@@ -37,8 +37,6 @@ async def main():
token,
"Respond bot",
DailyParams(
audio_in_sample_rate=16000,
audio_out_sample_rate=24000,
audio_out_enabled=True,
vad_enabled=True,
vad_audio_passthrough=True,

2
examples/foundational/26a-gemini-multimodal-live-transcription.py
View File

@@ -37,8 +37,6 @@ async def main():
token,
"Respond bot",
DailyParams(
audio_in_sample_rate=16000,
audio_out_sample_rate=24000,
audio_out_enabled=True,
vad_enabled=True,
vad_audio_passthrough=True,

2
examples/foundational/26b-gemini-multimodal-live-function-calling.py
View File

@@ -84,8 +84,6 @@ async def main():
token,
"Respond bot",
DailyParams(
audio_in_sample_rate=16000,
audio_out_sample_rate=24000,
audio_out_enabled=True,
vad_enabled=True,
vad_audio_passthrough=True,

4
examples/foundational/26c-gemini-multimodal-live-video.py
View File

@@ -37,8 +37,6 @@ async def main():
token,
"Respond bot",
DailyParams(
audio_in_sample_rate=16000,
audio_out_sample_rate=24000,
audio_out_enabled=True,
vad_enabled=True,
vad_audio_passthrough=True,
@@ -47,8 +45,6 @@ async def main():
# matter because we can only use the Multimodal Live API's phrase
# endpointing, for now.
vad_analyzer=SileroVADAnalyzer(params=VADParams(stop_secs=0.5)),
start_audio_paused=True,
start_video_paused=True,
),
)

2
examples/foundational/26d-gemini-multimodal-live-text.py
View File

@@ -52,8 +52,6 @@ async def main():
token,
"Respond bot",
DailyParams(
audio_in_sample_rate=16000,
audio_out_sample_rate=24000,
audio_out_enabled=True,
vad_enabled=True,
vad_audio_passthrough=True,

7
examples/foundational/29-livekit-audio-chat.py
View File

@@ -38,8 +38,6 @@ load_dotenv(override=True)
logger.remove(0)
logger.add(sys.stderr, level="DEBUG")
DESIRED_SAMPLE_RATE = 16000
def generate_token(room_name: str, participant_name: str, api_key: str, api_secret: str) -> str:
token = api.AccessToken(api_key, api_secret)
@@ -114,11 +112,8 @@ async def main():
token=token,
room_name=room_name,
params=LiveKitParams(
audio_in_channels=1,
audio_in_enabled=True,
audio_out_enabled=True,
audio_in_sample_rate=DESIRED_SAMPLE_RATE,
audio_out_sample_rate=DESIRED_SAMPLE_RATE,
vad_analyzer=SileroVADAnalyzer(),
vad_enabled=True,
vad_audio_passthrough=True,
@@ -128,7 +123,6 @@ async def main():
stt = DeepgramSTTService(
api_key=os.getenv("DEEPGRAM_API_KEY"),
live_options=LiveOptions(
sample_rate=DESIRED_SAMPLE_RATE,
vad_events=True,
),
)
@@ -138,7 +132,6 @@ async def main():
tts = CartesiaTTSService(
api_key=os.getenv("CARTESIA_API_KEY"),
voice_id="79a125e8-cd45-4c13-8a67-188112f4dd22", # British Lady
sample_rate=DESIRED_SAMPLE_RATE,
)
messages = [

2
examples/simple-chatbot/server/bot-gemini.py
View File

@@ -121,8 +121,6 @@ async def main():
token,
"Chatbot",
DailyParams(
audio_in_sample_rate=16000,
audio_out_sample_rate=24000,
audio_out_enabled=True,
camera_out_enabled=True,
camera_out_width=1024,

9
examples/studypal/studypal.py
View File

@@ -112,7 +112,6 @@ async def main():
token,
"studypal",
DailyParams(
audio_out_sample_rate=44100,
audio_out_enabled=True,
transcription_enabled=True,
vad_enabled=True,
@@ -124,7 +123,6 @@ async def main():
api_key=os.getenv("CARTESIA_API_KEY"),
voice_id=os.getenv("CARTESIA_VOICE_ID", "4d2fd738-3b3d-4368-957a-bb4805275bd9"),
# British Narration Lady: 4d2fd738-3b3d-4368-957a-bb4805275bd9
sample_rate=44100,
)
llm = OpenAILLMService(api_key=os.getenv("OPENAI_API_KEY"), model="gpt-4o-mini")
@@ -155,7 +153,12 @@ Your task is to help the user understand and learn from this article in 2 senten
]
)
task = PipelineTask(pipeline, PipelineParams(allow_interruptions=True, enable_metrics=True))
task = PipelineTask(
pipeline,
PipelineParams(
audio_out_sample_rate=44100, allow_interruptions=True, enable_metrics=True
),
)
@transport.event_handler("on_first_participant_joined")
async def on_first_participant_joined(transport, participant):

26
examples/twilio-chatbot/bot.py
View File

@@ -11,7 +11,6 @@ import sys
import wave
import aiofiles
from deepgram import LiveOptions
from dotenv import load_dotenv
from fastapi import WebSocket
from loguru import logger
@@ -36,8 +35,6 @@ load_dotenv(override=True)
logger.remove(0)
logger.add(sys.stderr, level="DEBUG")
SAMPLE_RATE = 8000
async def save_audio(server_name: str, audio: bytes, sample_rate: int, num_channels: int):
if len(audio) > 0:
@@ -63,29 +60,21 @@ async def run_bot(websocket_client: WebSocket, stream_sid: str, testing: bool):
params=FastAPIWebsocketParams(
audio_in_enabled=True,
audio_out_enabled=True,
audio_out_sample_rate=SAMPLE_RATE,
add_wav_header=False,
vad_enabled=True,
vad_analyzer=SileroVADAnalyzer(sample_rate=SAMPLE_RATE),
vad_analyzer=SileroVADAnalyzer(),
vad_audio_passthrough=True,
serializer=TwilioFrameSerializer(
stream_sid, TwilioFrameSerializer.InputParams(sample_rate=SAMPLE_RATE)
),
serializer=TwilioFrameSerializer(stream_sid),
),
)
llm = OpenAILLMService(api_key=os.getenv("OPENAI_API_KEY"), model="gpt-4o")
stt = DeepgramSTTService(
api_key=os.getenv("DEEPGRAM_API_KEY"),
live_options=LiveOptions(sample_rate=SAMPLE_RATE),
audio_passthrough=True,
)
stt = DeepgramSTTService(api_key=os.getenv("DEEPGRAM_API_KEY"), audio_passthrough=True)
tts = CartesiaTTSService(
api_key=os.getenv("CARTESIA_API_KEY"),
voice_id="79a125e8-cd45-4c13-8a67-188112f4dd22", # British Lady
sample_rate=SAMPLE_RATE,
push_silence_after_stop=testing,
)
@@ -101,7 +90,7 @@ async def run_bot(websocket_client: WebSocket, stream_sid: str, testing: bool):
# NOTE: Watch out! This will save all the conversation in memory. You can
# pass `buffer_size` to get periodic callbacks.
audiobuffer = AudioBufferProcessor(sample_rate=SAMPLE_RATE)
audiobuffer = AudioBufferProcessor()
pipeline = Pipeline(
[
@@ -116,7 +105,12 @@ async def run_bot(websocket_client: WebSocket, stream_sid: str, testing: bool):
]
)
task = PipelineTask(pipeline, params=PipelineParams(allow_interruptions=True))
task = PipelineTask(
pipeline,
params=PipelineParams(
audio_in_sample_rate=8000, audio_out_sample_rate=8000, allow_interruptions=True
),
)
@transport.event_handler("on_client_connected")
async def on_client_connected(transport, client):

22
examples/twilio-chatbot/client.py
View File

@@ -16,7 +16,6 @@ from uuid import uuid4
import aiofiles
import aiohttp
from deepgram import LiveOptions
from dotenv import load_dotenv
from loguru import logger
@@ -44,7 +43,6 @@ logger.add(sys.stderr, level="DEBUG")
DEFAULT_CLIENT_DURATION = 30
SAMPLE_RATE = 8000
async def download_twiml(server_url: str) -> str:
@@ -92,15 +90,10 @@ async def run_client(client_name: str, server_url: str, duration_secs: int):
params=WebsocketClientParams(
audio_in_enabled=True,
audio_out_enabled=True,
audio_out_sample_rate=SAMPLE_RATE,
add_wav_header=False,
serializer=TwilioFrameSerializer(
stream_sid, params=TwilioFrameSerializer.InputParams(sample_rate=SAMPLE_RATE)
),
serializer=TwilioFrameSerializer(stream_sid),
vad_enabled=True,
vad_analyzer=SileroVADAnalyzer(
params=VADParams(stop_secs=1.5), sample_rate=SAMPLE_RATE
),
vad_analyzer=SileroVADAnalyzer(params=VADParams(stop_secs=1.5)),
vad_audio_passthrough=True,
),
)
@@ -110,14 +103,12 @@ async def run_client(client_name: str, server_url: str, duration_secs: int):
# We let the audio passthrough so we can record the conversation.
stt = DeepgramSTTService(
api_key=os.getenv("DEEPGRAM_API_KEY"),
live_options=LiveOptions(sample_rate=SAMPLE_RATE),
audio_passthrough=True,
)
tts = CartesiaTTSService(
api_key=os.getenv("CARTESIA_API_KEY"),
voice_id="e13cae5c-ec59-4f71-b0a6-266df3c9bb8e", # Madame Mischief
sample_rate=SAMPLE_RATE,
push_silence_after_stop=True,
)
@@ -133,7 +124,7 @@ async def run_client(client_name: str, server_url: str, duration_secs: int):
# NOTE: Watch out! This will save all the conversation in memory. You can
# pass `buffer_size` to get periodic callbacks.
audiobuffer = AudioBufferProcessor(sample_rate=SAMPLE_RATE)
audiobuffer = AudioBufferProcessor()
pipeline = Pipeline(
[
@@ -148,7 +139,12 @@ async def run_client(client_name: str, server_url: str, duration_secs: int):
]
)
task = PipelineTask(pipeline, params=PipelineParams(allow_interruptions=True))
task = PipelineTask(
pipeline,
params=PipelineParams(
audio_in_sample_rate=8000, audio_out_sample_rate=8000, allow_interruptions=True
),
)
@transport.event_handler("on_connected")
async def on_connected(transport: WebsocketClientTransport, client):

11
examples/websocket-server/bot.py
View File

@@ -17,6 +17,7 @@ from pipecat.pipeline.pipeline import Pipeline
from pipecat.pipeline.runner import PipelineRunner
from pipecat.pipeline.task import PipelineParams, PipelineTask
from pipecat.processors.aggregators.openai_llm_context import OpenAILLMContext
from pipecat.serializers.protobuf import ProtobufFrameSerializer
from pipecat.services.cartesia import CartesiaTTSService
from pipecat.services.deepgram import DeepgramSTTService
from pipecat.services.openai import OpenAILLMService
@@ -80,7 +81,7 @@ class SessionTimeoutHandler:
async def main():
transport = WebsocketServerTransport(
params=WebsocketServerParams(
audio_out_sample_rate=16000,
serializer=ProtobufFrameSerializer(),
audio_out_enabled=True,
add_wav_header=True,
vad_enabled=True,
@@ -97,7 +98,6 @@ async def main():
tts = CartesiaTTSService(
api_key=os.getenv("CARTESIA_API_KEY"),
voice_id="79a125e8-cd45-4c13-8a67-188112f4dd22", # British Lady
sample_rate=16000,
)
messages = [
@@ -122,7 +122,12 @@ async def main():
]
)
task = PipelineTask(pipeline, params=PipelineParams(allow_interruptions=True))
task = PipelineTask(
pipeline,
params=PipelineParams(
audio_in_sample_rate=16000, audio_out_sample_rate=16000, allow_interruptions=True
),
)
@transport.event_handler("on_client_connected")
async def on_client_connected(transport, client):