基于Python 的语音重采样函数解析

Lewa ·

更新时间:2024-09-21

· 925 次阅读

因为工作中会经常遇到不同采样率的声音文件的问题，特意写了一下重采样的程序。

原理就是把采样点转换到时间刻度之后再进行插值，经过测试，是没有问题的。


#!/usr/bin/env python
# -*- coding: utf-8 -*-
# @Time : 17-7-21 下午2:32
# @Author : Lei.Jinggui
# @Site : http://blog.csdn.net/lccever
# @File : Resample.py
# @Software: PyCharm Community Edition
# @contact: lccever@126.com
import numpy as np
def Resample(input_signal,src_fs,tar_fs):
 '''
 :param input_signal:输入信号
 :param src_fs:输入信号采样率
 :param tar_fs:输出信号采样率
 :return:输出信号
 '''
 dtype = input_signal.dtype
 audio_len = len(input_signal)
 audio_time_max = 1.0*(audio_len-1) / src_fs
 src_time = 1.0 * np.linspace(0,audio_len,audio_len) / src_fs
 tar_time = 1.0 * np.linspace(0,np.int(audio_time_max*tar_fs),np.int(audio_time_max*tar_fs)) / tar_fs
 output_signal = np.interp(tar_time,src_time,input_signal).astype(dtype)
 return output_signal
if __name__ == '__main__':
 import wave
 import pyaudio
 def playSound(audio_data_short, framerate=16000, channels=1):
  preply = pyaudio.PyAudio()
  # 播放声音
  streamreply = preply.open(format=pyaudio.paInt16,
         channels=channels,
         rate=framerate,
         output=True)
  data = audio_data_short.tostring()
  streamreply.write(data)
  streamreply.close()
  preply.terminate()
 wave_file = 'test.wav'
 audio_file = wave.open(wave_file, 'rb')
 audio_data = audio_file.readframes(audio_file.getnframes())
 audio_data_short = np.fromstring(audio_data, np.short)
 src_fs = audio_file.getframerate()
 src_chanels = audio_file.getnchannels()
 if src_chanels > 1:
  audio_data_short = audio_data_short[::src_chanels]
 tar_fs = np.int(src_fs * 0.5)
 playSound(audio_data_short,framerate=src_fs)
 audio_data_short0 = Resample(audio_data_short,src_fs,tar_fs)
 playSound(audio_data_short0,framerate=tar_fs)

补充知识：Python 多线程的退出/停止的一种是实现思路

在使用多线程的过程中,我们知道,python的线程是没有stop/terminate方法的,也就是说它被启动后,你无法再主动去退出它,除非主进程退出了,注意,是主进程,不是线程的父进程.

一个比较合理的方式就是把原因需要放到threading.Thread的target中的线程函数,改写到一个继承类中,下面是一个实现例子


import threading
import time
import os
# 原本需要用来启动的无线循环的函数
def print_thread():
 pid = os.getpid()
 counts = 0
 while True:
  print(f'threading pid: {pid} ran: {counts:04d} s')
  counts += 1
  time.sleep(1)
# 把函数放到改写到类的run方法中，便可以通过调用类方法，实现线程的终止
class StoppableThread(threading.Thread):
 def __init__(self, daemon=None):
  super(StoppableThread, self).__init__(daemon=daemon)
  self.__is_running = True
  self.daemon = daemon
 def terminate(self):
  self.__is_running = False
 def run(self):
  pid = os.getpid()
  counts = 0
  while self.__is_running:
   print(f'threading running: {pid} ran: {counts:04d} s')
   counts += 1
   time.sleep(1)
def call_thread():
 thread = StoppableThread()
 thread.daemon = True
 thread.start()
 pid = os.getpid()
 counts = 0
 for i in range(5):
  print(f'0 call threading pid: {pid} ran: {counts:04d} s')
  counts += 2
  time.sleep(2)
 # 主动把线程退出
 thread.terminate()
if __name__ == '__main__':
 call_thread()
 print(f'==========call_thread finish===========')
 counts = 0
 for i in range(5):
  counts += 1
  time.sleep(1)
  print(f'main thread:{counts:04d} s')

以上这篇基于Python 的语音重采样函数解析就是小编分享给大家的全部内容了，希望能给大家一个参考，也希望大家多多支持软件开发网。

您可能感兴趣的文章:python中resample函数实现重采样和降采样代码Python基于opencv的图像压缩算法实例分析使用python实现语音文件的特征提取方法Python实现语音识别和语音合成功能

采样函数重采样 Python

1024 个赞