Merge pull request #2008 from pipecat-ai/khk/groq-audio

fix groq wav file header parsing

CHANGELOG.md

@@ -9,6 +9,9 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### Fixed
 
+- Fixed an issue with `GroqTTSService` where it was not properly parsing the
+  WAV file header.
+
 - Fixed an issue with `GoogleSTTService` where it was constantly reconnecting 
   before starting to receive audio from the user.
 

src/pipecat/services/groq/tts.py

@@ -4,6 +4,8 @@
 # SPDX-License-Identifier: BSD 2-Clause License
 #
 
+import io
+import wave
 from typing import AsyncGenerator, Optional
 
 from loguru import logger
@@ -78,22 +80,26 @@ class GroqTTSService(TTSService):
         await self.start_ttfb_metrics()
         yield TTSStartedFrame()
 
-        response = await self._client.audio.speech.create(
+        try:
-            model=self._model_name,
+            response = await self._client.audio.speech.create(
-            voice=self._voice_id,
+                model=self._model_name,
-            response_format=self._output_format,
+                voice=self._voice_id,
-            input=text,
+                response_format=self._output_format,
-        )
+                input=text,
+            )
 
-        async for data in response.iter_bytes():
+            async for data in response.iter_bytes():
-            if measuring_ttfb:
+                if measuring_ttfb:
-                await self.stop_ttfb_metrics()
+                    await self.stop_ttfb_metrics()
-                measuring_ttfb = False
+                    measuring_ttfb = False
-            # remove wav header if present
+
-            if data.startswith(b"RIFF"):
+                with wave.open(io.BytesIO(data)) as w:
-                data = data[44:]
+                    channels = w.getnchannels()
-                if len(data) == 0:
+                    frame_rate = w.getframerate()
-                    continue
+                    num_frames = w.getnframes()
-            yield TTSAudioRawFrame(data, self.sample_rate, 1)
+                    bytes = w.readframes(num_frames)
+                    yield TTSAudioRawFrame(bytes, frame_rate, channels)
+        except Exception as e:
+            logger.error(f"{self} exception: {e}")
 
         yield TTSStoppedFrame()