From 5a83f8cae9aaa427c1b6ad1f1e5594f6d1dbde16 Mon Sep 17 00:00:00 2001
From: duyalei <>
Date: Tue, 24 Sep 2024 11:50:17 +0800
Subject: [PATCH] add playback_rate in AzureTTSService

---
 src/pipecat/services/azure.py | 36 ++++++++++++++++++++++++++---------
 1 file changed, 27 insertions(+), 9 deletions(-)

diff --git a/src/pipecat/services/azure.py b/src/pipecat/services/azure.py
index 9177f9e4e..b70aca632 100644
--- a/src/pipecat/services/azure.py
+++ b/src/pipecat/services/azure.py
@@ -42,7 +42,10 @@
         CancellationReason,
         languageconfig,
     )
-    from azure.cognitiveservices.speech.audio import AudioStreamFormat, PushAudioInputStream
+    from azure.cognitiveservices.speech.audio import (
+        AudioStreamFormat,
+        PushAudioInputStream,
+    )
     from azure.cognitiveservices.speech.dialog import AudioConfig
 except ModuleNotFoundError as e:
     logger.error(f"Exception: {e}")
@@ -54,7 +57,12 @@
 
 class AzureLLMService(BaseOpenAILLMService):
     def __init__(
-        self, *, api_key: str, endpoint: str, model: str, api_version: str = "2023-12-01-preview"
+        self,
+        *,
+        api_key: str,
+        endpoint: str,
+        model: str,
+        api_version: str = "2023-12-01-preview",
     ):
         # Initialize variables before calling parent __init__() because that
         # will call create_client() and we need those values there.
@@ -78,6 +86,7 @@ def __init__(
         region: str,
         voice="en-US-SaraNeural",
         sample_rate: int = 16000,
+        playback_rate: float = 1.0,
         **kwargs,
     ):
         super().__init__(sample_rate=sample_rate, **kwargs)
@@ -87,6 +96,7 @@ def __init__(
 
         self._voice = voice
         self._sample_rate = sample_rate
+        self._playback_rate = playback_rate
 
     def can_generate_metrics(self) -> bool:
         return True
@@ -106,7 +116,7 @@ async def run_tts(self, text: str) -> AsyncGenerator[Frame, None]:
             f"<voice name='{self._voice}'>"
             "<mstts:silence type='Sentenceboundary' value='20ms' />"
             "<mstts:express-as style='lyrical' styledegree='2' role='SeniorFemale'>"
-            "<prosody rate='1.05'>"
+            f"<prosody rate='{self._playback_rate}'>"
             f"{text}"
             "</prosody></mstts:express-as></voice></speak> "
         )
@@ -119,7 +129,9 @@ async def run_tts(self, text: str) -> AsyncGenerator[Frame, None]:
             await self.push_frame(TTSStartedFrame())
             # Azure always sends a 44-byte header. Strip it off.
             yield TTSAudioRawFrame(
-                audio=result.audio_data[44:], sample_rate=self._sample_rate, num_channels=1
+                audio=result.audio_data[44:],
+                sample_rate=self._sample_rate,
+                num_channels=1,
             )
             await self.push_frame(TTSStoppedFrame())
         elif result.reason == ResultReason.Canceled:
@@ -154,12 +166,15 @@ def __init__(
                 speech_config=speech_config,
                 audio_config=audio_config,
                 auto_detect_source_language_config=languageconfig.AutoDetectSourceLanguageConfig(
-                    languages=language)
+                    languages=language
+                ),
             )
         else:
-            self._speech_recognizer = SpeechRecognizer(speech_config=speech_config,
-                                                       audio_config=audio_config,
-                                                       language=language)
+            self._speech_recognizer = SpeechRecognizer(
+                speech_config=speech_config,
+                audio_config=audio_config,
+                language=language,
+            )
         self._speech_recognizer.recognized.connect(self._on_handle_recognized)
 
     async def run_stt(self, audio: bytes) -> AsyncGenerator[Frame, None]:
@@ -253,6 +268,9 @@ async def run_image_gen(self, prompt: str) -> AsyncGenerator[Frame, None]:
                 image_stream = io.BytesIO(await response.content.read())
                 image = Image.open(image_stream)
                 frame = URLImageRawFrame(
-                    url=image_url, image=image.tobytes(), size=image.size, format=image.format
+                    url=image_url,
+                    image=image.tobytes(),
+                    size=image.size,
+                    format=image.format,
                 )
                 yield frame