google STT - use the baseclass resampler (livekit#1106)

Ganjiali · Nov 18, 2024 · 4416785 · 4416785
1 parent 6b03b45
commit 4416785
Show file tree

Hide file tree

Showing 2 changed files with 7 additions and 5 deletions.
diff --git a/.changeset/warm-zoos-lie.md b/.changeset/warm-zoos-lie.md
@@ -0,0 +1,5 @@
+---
+"livekit-plugins-google": patch
+---
+
+google STT - use the baseclass resampler
diff --git a/livekit-plugins/livekit-plugins-google/livekit/plugins/google/stt.py b/livekit-plugins/livekit-plugins-google/livekit/plugins/google/stt.py
@@ -240,11 +240,11 @@ def __init__(
         client: SpeechAsyncClient,
         recognizer: str,
         config: STTOptions,
-        sample_rate: int = 48000,
+        sample_rate: int = 16000,
         num_channels: int = 1,
         max_retry: int = 32,
     ) -> None:
-        super().__init__(stt)
+        super().__init__(stt, sample_rate=sample_rate)
 
         self._client = client
         self._recognizer = recognizer
@@ -294,9 +294,6 @@ async def input_generator():
 
                         async for frame in self._input_ch:
                             if isinstance(frame, rtc.AudioFrame):
-                                frame = frame.remix_and_resample(
-                                    self._sample_rate, self._num_channels
-                                )
                                 yield cloud_speech.StreamingRecognizeRequest(
                                     audio=frame.data.tobytes()
                                 )