我做了类似的事情，但使用python-sounddevice。这是我的回购

    编辑：这是特定的录音课程，以便澄清

    1. import threading
    2. import queue
    3. import numpy
    4. import sounddevice as sd
    5. import soundfile as sf
    7. class AudioRecorder():
    9.     def __init__(self):
    11.         self.open = True
    12.         self.file_name = 'name_of_file.wav'
    13.         self.channels = 1
    14.         self.q = queue.Queue()
    16.         # Get samplerate
    17.         device_info = sd.query_devices(2, 'input')
    18.         self.samplerate = int(device_info['default_samplerate'])
    20.     def callback(self, indata, frames, time, status):
    22.         # This is called (from a separate thread) for each audio block.
    23.         if status:
    24.             print(status, file=sys.stderr)
    25.         self.q.put(indata.copy())
    27.     def record(self):
    28.         with sf.SoundFile(self.file_name, mode='x', samplerate=self.samplerate, channels=self.channels) as file:
    29.             with sd.InputStream(samplerate=self.samplerate, channels=self.channels, callback=self.callback):
    31.                 while(self.open == True):
    32.                     file.write(self.q.get())

    编辑2：代码是一个Python类，使用类似于问题中显示的图像的I2S麦克风创建音频文件。值为self.opentrue时，sounddevice会将音频数据写入队列（def callback），然后将数据写入文件。您所要做的就是切换self.open开始和停止录制。

    作者: Kyle Polley 发布者: 27.11.2018 05:40