🐛[bugfix] Model : speakerがサラウンドデバイスの場合に音声が文字起こしされない問題を修正 #10

models/transcription/transcription_transcriber.py

@@ -10,6 +10,7 @@ from .transcription_whisper import getWhisperModel, checkWhisperWeight
 
 import torch
 import numpy as np
+from pydub import AudioSegment
 
 PHRASE_TIMEOUT = 3
 MAX_PHRASES = 10
@@ -97,21 +98,21 @@ class AudioTranscriber:
         return audio_data
 
     def processSpeakerData(self):
-        original_channels = self.audio_sources["channels"]
-        if original_channels <= 2:
-            channels = original_channels
-            sample_rate = self.audio_sources["sample_rate"]
-        else:
-            channels = 2
-            sample_rate = self.audio_sources["sample_rate"]*original_channels/2
-
         temp_file = BytesIO()
         with wave.open(temp_file, 'wb') as wf:
-            wf.setnchannels(channels)
+            wf.setnchannels(self.audio_sources["channels"])
             wf.setsampwidth(get_sample_size(paInt16))
-            wf.setframerate(sample_rate)
+            wf.setframerate(self.audio_sources["sample_rate"])
             wf.writeframes(self.audio_sources["last_sample"])
         temp_file.seek(0)
+
+        if self.audio_sources["channels"] > 2:
+            audio = AudioSegment.from_file(temp_file, format="wav")
+            mono_audio = audio.set_channels(1)
+            temp_file = BytesIO()
+            mono_audio.export(temp_file, format="wav")
+            temp_file.seek(0)
+
         with AudioFile(temp_file) as source:
             audio = self.audio_recognizer.record(source)
         return audio