带有静音片段的 Pyaudio

Posted 2023-03-17

【中文标题】带有静音片段的 Pyaudio

【英文标题】：Pyaudio with silent segments

【发布时间】：2021-09-09 17:07:44

【问题描述】：

我正在尝试以下代码：

import pyaudio
from pydub import Audiosegment,generators
import time

silence = AudioSegment.silent(duration=125,frame_rate=32768)
sine_data = generators.Sine(1000).to_audio_segment().raw_data

chunk_number = 0
def preview_callback(in_data, frame_count, time_info, status):
    global chunk_number
    print("Preview callback")
    chunk_number +=1
    print(chunk_number)
    if chunk_number<10:
        return (silence.raw_data,pyaudio.paContinue)
    else:
        return (sine_data,pyaudio.paContinue)
#QoS settings
bit_rate = 128*1024 #128 kb/sec
packet_time = 125 #125 msec
packet_size = int(16384/4)
new_sample_rate = 32768

format = pyaudio.paInt16
channels = 2

p = pyaudio.PyAudio()

stream = p.open(format=pyaudio.paInt16,channels=2,rate=new_sample_rate,output=True,frames_per_buffer=packet_size,stream_callback=preview_callback)
stream.start_stream()

但Preview callback 消息只打印一次。 代码有什么问题？

编辑：使用正弦信号，预览回调消息打印正确，我可以听到正弦声音。

【参考方案1】：

import pyaudio
from pydub import AudioSegment,generators
import time

sine_segment = generators.Sine(1000).to_audio_segment()
sine_segment = sine_segment-200
sine_data = sine_segment.raw_data

def preview_callback(in_data, frame_count, time_info, status):
    print("Preview callback")
    return (sine_data,pyaudio.paContinue)
#QoS settings
bit_rate = 128*1024 #128 kb/sec
packet_time = 125 #125 msec
packet_size = int(16384/4)
new_sample_rate = 32768

format = pyaudio.paInt16
channels = 2

p = pyaudio.PyAudio()

stream = p.open(format=pyaudio.paInt16,channels=2,rate=new_sample_rate,output=True,frames_per_buffer=packet_size,stream_callback=preview_callback)
stream.start_stream()

while(True):
    print(stream.is_active())
    print(stream.is_stopped())
    time.sleep(2)

上面的代码不是物理静音，而是使用正弦信号。在这个信号中我们改变音量（-200db =~静音），然后代码似乎工作了。

警告：有更好的解决方案，所以在使用我的之前先搜索一下。