👍️[Update] Controller : Deviceの自動変更処理を実装

requirements.txt

@@ -16,5 +16,5 @@ psutil==5.9.8
 pykakasi==2.3.0
 pycaw==20240210
 translators @ git+https://github.com/misyaguziya/translators@5.9.2.1
-SpeechRecognition @ git+https://github.com/misyaguziya/custom_speech_recognition@3.10.4
+SpeechRecognition @ git+https://github.com/misyaguziya/custom_speech_recognition@3.10.4.1
 tinyoscquery @ git+https://github.com/cyberkitsune/tinyoscquery@0.1.2

src-python/model.py

@@ -426,13 +426,8 @@ class Model:
         return result
 
     def startMicTranscript(self, fnc):
-        if config.AUTO_MIC_SELECT is True:
+        mic_host_name = config.SELECTED_MIC_HOST
-            default_device = device_manager.getDefaultMicDevice()
+        mic_device_name = config.SELECTED_MIC_DEVICE
-            mic_host_name = default_device["host"]["name"]
-            mic_device_name = default_device["device"]["name"]
-        else:
-            mic_host_name = config.SELECTED_MIC_HOST
-            mic_device_name = config.SELECTED_MIC_DEVICE
 
         mic_device_list = device_manager.getMicDevices().get(mic_host_name, [{"name": "NoDevice"}])
         selected_mic_device = [device for device in mic_device_list if device["name"] == mic_device_name]
@@ -453,7 +448,7 @@ class Model:
             device=mic_device,
             energy_threshold=config.MIC_THRESHOLD,
             dynamic_energy_threshold=config.MIC_AUTOMATIC_THRESHOLD,
-            record_timeout=record_timeout,
+            phrase_time_limit=record_timeout,
         )
         # self.mic_audio_recorder.recordIntoQueue(self.mic_audio_queue, mic_energy_queue)
         self.mic_audio_recorder.recordIntoQueue(self.mic_audio_queue, None)
@@ -560,13 +555,8 @@ class Model:
         if isinstance(fnc, Callable):
             self.check_mic_energy_fnc = fnc
 
-        if config.AUTO_MIC_SELECT is True:
+        mic_host_name = config.SELECTED_MIC_HOST
-            default_device = device_manager.getDefaultMicDevice()
+        mic_device_name = config.SELECTED_MIC_DEVICE
-            mic_host_name = default_device["host"]["name"]
-            mic_device_name = default_device["device"]["name"]
-        else:
-            mic_host_name = config.SELECTED_MIC_HOST
-            mic_device_name = config.SELECTED_MIC_DEVICE
 
         mic_device_list = device_manager.getMicDevices().get(mic_host_name, [{"name": "NoDevice"}])
         selected_mic_device = [device for device in mic_device_list if device["name"] == mic_device_name]
@@ -620,7 +610,7 @@ class Model:
             device=speaker_device,
             energy_threshold=config.SPEAKER_THRESHOLD,
             dynamic_energy_threshold=config.SPEAKER_AUTOMATIC_THRESHOLD,
-            record_timeout=record_timeout,
+            phrase_time_limit=record_timeout,
         )
         # self.speaker_audio_recorder.recordIntoQueue(speaker_audio_queue, speaker_energy_queue)
         self.speaker_audio_recorder.recordIntoQueue(speaker_audio_queue, None)

src-python/models/transcription/transcription_recorder.py

@@ -91,10 +91,12 @@ class SelectedSpeakerEnergyRecorder(BaseEnergyRecorder):
         # self.adjustForNoise()
 
 class BaseEnergyAndAudioRecorder:
-    def __init__(self, source, energy_threshold, dynamic_energy_threshold, record_timeout):
+    def __init__(self, source, energy_threshold, dynamic_energy_threshold, phrase_time_limit, phrase_timeout, record_timeout):
         self.recorder = Recognizer()
         self.recorder.energy_threshold = energy_threshold
         self.recorder.dynamic_energy_threshold = dynamic_energy_threshold
+        self.phrase_time_limit = phrase_time_limit
+        self.phrase_timeout = phrase_timeout
         self.record_timeout = record_timeout
         self.stop = None
 
@@ -117,20 +119,29 @@ class BaseEnergyAndAudioRecorder:
         self.stop, self.pause, self.resume = self.recorder.listen_energy_and_audio_in_background(
             source=self.source,
             callback=audioRecordCallback,
-            phrase_time_limit=self.record_timeout,
+            phrase_time_limit=self.phrase_time_limit,
-            callback_energy=energyRecordCallback if energy_queue is not None else None)
+            callback_energy=energyRecordCallback if energy_queue is not None else None,
+            phrase_timeout=self.phrase_timeout,
+            record_timeout=self.record_timeout)
 
 class SelectedMicEnergyAndAudioRecorder(BaseEnergyAndAudioRecorder):
-    def __init__(self, device, energy_threshold, dynamic_energy_threshold, record_timeout):
+    def __init__(self, device, energy_threshold, dynamic_energy_threshold, phrase_time_limit, phrase_timeout:int=1, record_timeout:int=5):
         source=Microphone(
             device_index=device['index'],
             sample_rate=int(device["defaultSampleRate"]),
         )
-        super().__init__(source=source, energy_threshold=energy_threshold, dynamic_energy_threshold=dynamic_energy_threshold, record_timeout=record_timeout)
+        super().__init__(
+            source=source,
+            energy_threshold=energy_threshold,
+            dynamic_energy_threshold=dynamic_energy_threshold,
+            phrase_time_limit=phrase_time_limit,
+            phrase_timeout=phrase_timeout,
+            record_timeout=record_timeout,
+        )
         # self.adjustForNoise()
 
 class SelectedSpeakerEnergyAndAudioRecorder(BaseEnergyAndAudioRecorder):
-    def __init__(self, device, energy_threshold, dynamic_energy_threshold, record_timeout):
+    def __init__(self, device, energy_threshold, dynamic_energy_threshold, phrase_time_limit, phrase_timeout:int=1, record_timeout:int=5):
 
         source = Microphone(speaker=True,
             device_index= device["index"],
@@ -138,5 +149,12 @@ class SelectedSpeakerEnergyAndAudioRecorder(BaseEnergyAndAudioRecorder):
             chunk_size=get_sample_size(paInt16),
             channels=device["maxInputChannels"]
         )
-        super().__init__(source=source, energy_threshold=energy_threshold, dynamic_energy_threshold=dynamic_energy_threshold, record_timeout=record_timeout)
+        super().__init__(
+            source=source,
+            energy_threshold=energy_threshold,
+            dynamic_energy_threshold=dynamic_energy_threshold,
+            phrase_time_limit=phrase_time_limit,
+            phrase_timeout=phrase_timeout,
+            record_timeout=record_timeout,
+        )
         # self.adjustForNoise()