pyauido中的奇怪的白噪声

Question

我正在尝试通过与Numpy阵列混乱来动态修改歌曲的音量。我做到了，但是除了主歌之外，我还可以听到一些奇怪的声音。我认为问题是“ numpy.int8”。

import pyaudio
import wave
import numpy


def audio_datalist_set_volume(datalist, volume):
    """ Change value of list of audio chunks """
    sound_level = (volume / 100.)

    chunk = numpy.frombuffer(datalist, numpy.int8)

    chunk = chunk * sound_level

    return chunk.astype(numpy.int8)


CHUNK = 1024

# if len(sys.argv) < 2:
#     print("Plays a wave file.\n\nUsage: %s filename.wav" % sys.argv[0])
#     sys.exit(-1)

wf = wave.open("your_audio.wav", 'rb')

# instantiate PyAudio (1)
p = pyaudio.PyAudio()

# open stream (2)
stream = p.open(format=p.get_format_from_width(wf.getsampwidth()),
                channels=wf.getnchannels(),
                rate=wf.getframerate(),
                output=True)

# read data
data = wf.readframes(CHUNK)

# play stream (3)
while len(data) > 0:
    nda = audio_datalist_set_volume(data, 20)
    stream.write(nda)
    data = wf.readframes(CHUNK)

# stop stream (4)
stream.stop_stream()
stream.close()

# close PyAudio (5)
p.terminate()

更新：我认为我发现了问题：numpy.int8应该为“ B”或“ I1”更改。这是一个适当的决定吗？

Answer 1

如果要处理任意WAV文件，则应确保Numpy数据类型与流的格式匹配：

def audio_datalist_set_volume(datalist, volume, numpy_type):
    """ Change value of list of audio chunks """
    sound_level = (volume / 100.)
    chunk = numpy.frombuffer(datalist, numpy_type)
    chunk = chunk * sound_level
    return chunk

# ...

numpy_type = {
    pyaudio.paFloat32: np.float32,
    pyaudio.paInt32: np.int32,
    # numpy can't do 24-bit ints :(
    pyaudio.paInt16: np.int16,
    pyaudio.paInt8: np.int8,
    pyaudio.paUInt8: np.uint8,
}[stream._format]