|
|
@@ -815,6 +815,8 @@ class AudioToTextRecorder:
|
|
|
0.3)
|
|
|
)
|
|
|
self.frames = []
|
|
|
+ self.new_frames = mp.Event()
|
|
|
+ self.new_frames.set()
|
|
|
|
|
|
# Recording control flags
|
|
|
self.is_recording = False
|
|
|
@@ -1267,11 +1269,11 @@ class AudioToTextRecorder:
|
|
|
audio_array = np.frombuffer(b''.join(self.frames), dtype=np.int16)
|
|
|
self.audio = audio_array.astype(np.float32) / INT16_MAX_ABS_VALUE
|
|
|
self.frames.clear()
|
|
|
+ self.new_frames.set()
|
|
|
|
|
|
# Reset recording-related timestamps
|
|
|
self.recording_stop_time = 0
|
|
|
self.listen_start = 0
|
|
|
-
|
|
|
self._set_state("inactive")
|
|
|
|
|
|
except KeyboardInterrupt:
|
|
|
@@ -1453,6 +1455,7 @@ class AudioToTextRecorder:
|
|
|
self.wakeword_detected = False
|
|
|
self.wake_word_detect_time = 0
|
|
|
self.frames = []
|
|
|
+ self.new_frames.set()
|
|
|
self.is_recording = True
|
|
|
self.recording_start_time = time.time()
|
|
|
self.is_silero_speech_active = False
|
|
|
@@ -1816,7 +1819,8 @@ class AudioToTextRecorder:
|
|
|
# Add the buffered audio
|
|
|
# to the recording frames
|
|
|
self.frames.extend(list(self.audio_buffer))
|
|
|
- self.audio_buffer.clear()
|
|
|
+ self.new_frames.set()
|
|
|
+ self.audio_buffer.clear()
|
|
|
|
|
|
if self.use_extended_logging:
|
|
|
logging.debug('Debug: Resetting Silero VAD model states')
|
|
|
@@ -1987,6 +1991,7 @@ class AudioToTextRecorder:
|
|
|
if self.use_extended_logging:
|
|
|
logging.debug('Debug: Appending data to frames')
|
|
|
self.frames.append(data)
|
|
|
+ self.new_frames.set()
|
|
|
|
|
|
if self.use_extended_logging:
|
|
|
logging.debug('Debug: Checking if not recording or speech end silence start')
|
|
|
@@ -2029,6 +2034,8 @@ class AudioToTextRecorder:
|
|
|
|
|
|
# Check if the recording is active
|
|
|
if self.is_recording:
|
|
|
+ self.new_frames.wait()
|
|
|
+ self.new_frames.clear()
|
|
|
|
|
|
# Sleep for the duration of the transcription resolution
|
|
|
time.sleep(self.realtime_processing_pause)
|
Kolja Beigel