Question

我正在使用Pyaudio同时播放和录制wav文件。我在Ubuntu 16.04上使用Python 2.7.12。主要步骤是：

播放wav文件并通过扬声器输出。
使用麦克风录制声音。
每隔一定数量的帧实时同时处理wav和录制的声音，例如512帧。

这是我的代码：

chunk1 = 2048
chunk2 = 2048
format_ = pyaudio.paInt16
wf = wave.open(file_, 'rb')
p1 = pyaudio.PyAudio()
p2 = pyaudio.PyAudio()

# stream1 for playing the wav file
stream1 = p1.open(format=format_,
                  channels=1,
                  rate=16000,
                  output=True,
                  frames_per_buffer=chunk1)

# stream1 for recording the sound coming from speaker
stream2 = p2.open(format=format_,
                  channels=1,
                  rate=16000,
                  input=True,
                  frames_per_buffer=chunk2)

frames1 = []
frames2 = []

stream1.start_stream()

data_read = wf.readframes(chunk1)
while data_read != '':
    # play wav part
    stream1.write(data_read)
    data_read = wf.readframes(chunk1)
    temp1 = np.fromstring(data_read)
    print "source:", temp1.shape
    frames1.append(data_read)

    # record sound part
    data_write = stream2.read(chunk2)
    temp2 = np.fromstring(data_write)
    print "recorded:", temp2.shape
    frames2.append(data_write)

stream1.stop_stream()
stream1.close()
p1.terminate()
stream2.stop_stream()
stream2.close()
p2.terminate()

wf1 = wave.open('input.wav', 'wb')
wf1.setnchannels(wf.getnchannels())
wf1.setsampwidth(p1.get_sample_size(format_))
wf1.setframerate(wf.getframerate())
wf1.writeframes(b''.join(frames1))
wf1.close()

wf2 = wave.open('output.wav', 'wb')
wf2.setnchannels(wf.getnchannels())
wf2.setsampwidth(p2.get_sample_size(format_))
wf2.setframerate(wf.getframerate())
wf2.writeframes(b''.join(frames2))
wf2.close()

我这样做有些困难。我得到的错误如下：

source: (512,)
recorded: (512,)
source: (512,)
ALSA lib pcm.c:7963:(snd_pcm_recover) underrun occurred
recorded: (512,)
source: (512,)
ALSA lib pcm.c:7963:(snd_pcm_recover) underrun occurred
recorded: (512,)
ALSA lib pcm.c:7963:(snd_pcm_recover) underrun occurred
source: (512,)
ALSA lib pcm.c:7963:(snd_pcm_recover) underrun occurred
recorded: (512,)
source: (512,)
recorded: (512,)
ALSA lib pcm.c:7963:(snd_pcm_recover) underrun occurred

我也不明白为什么必须将chunk = 2048设置为frame = 512。

你能帮我解决这个问题吗？如果有人有其他建议，请告诉我。

当我发布这个问题时，我无法打开PyAudio文档：

https://people.csail.mit.edu/hubert/pyaudio/docs/

该网站已关闭。

干杯。

如何播放wav并同时录制来自扬声器的声音

0 个答案: