bugfix

rtcllm/vad.py

@@ -72,7 +72,7 @@ class AudioTrackVad(MediaStreamTrack):
 
 	async def vad_check(self, inframe):
 		frames = self.resample(inframe)
-		frame = frames[0]:
+		frame = frames[0]
 		is_speech = self.vad.is_speech(self.frame2bytes(frame), 
 										self.sample_rate)
 		if not self.triggered:
@@ -101,10 +101,13 @@ class AudioTrackVad(MediaStreamTrack):
 			# audio we've collected.
 			if num_unvoiced > 0.9 * self.ring_buffer.maxlen:
 				self.triggered = False
-				ret = await self.write_wave()
-				# ret = await self.gen_base64()
-				if self.onvoiceend:
+				duration = self.voice_duration()
+				if duration > 500 and self.onvoiceend:
+					ret = await self.write_wave()
 					await self.onvoiceend(ret)
+				else:
+					print(f'{duration=} {self.onvoiceend=}')
+
 				self.ring_buffer.clear()
 				self.voiced_frames = []
 
@@ -121,11 +124,17 @@ class AudioTrackVad(MediaStreamTrack):
 		b64 = base64.b64encode(audio_data).decode('utf-8')
 		return b64
 	
+	def voice_duration(self):
+		duration = 0
+		[ f.samples * 1000 / f.sample_rate + duration for f in self.self.voiced_frames ]
+		return duration
+
 	async def write_wave(self):
 		"""Writes a .wav file.
 
 		Takes path, PCM audio data, and sample rate.
 		"""
+
 		audio_data = self.to_mono16000_data()
 		path = temp_file(suffix='.wav')
 		# print(f'temp_file={path}')