#!/usr/bin/python
# encoding:utf-8
# Filename : processAudio.py
# author by :morespeech
# python2.7
# platform:visual studio code, windows
# topic: practice every day
# detial: process audio
# !/usr/bin/env python
# -*- coding:utf-8 -*-
import os
import wave
import pyaudio
import pylab as pl
import numpy as np
class cProcessAudio:
def __init__(self):
pass
# public function
# read file to buffer
def readWav(self, filename, mode):
if not os.path.exists(filename):
return
else:
fileHandle = wave.open(filename, mode)
params = fileHandle.getparams()
nchannels, sampwidth, samplerate, nsamples = params[:4]
# read the data
str_data = fileHandle.readframes(nsamples)
fileHandle.close()
wave_data = np.fromstring(str_data, dtype=np.short)
if params[0] == 2:
wave_data.shape = -1, 2
wave_data = wave_data.T
return wave_data
# write buffer to file, only mono
def writeWav(self, outfilename, writemode, data, fs, nchannel):
fileHandle = wave.open(outfilename, writemode)
fileHandle.setnchannels(nchannel)
fileHandle.setsampwidth(2)
fileHandle.setframerate(fs)
fileHandle.writeframes(data.tostring())
fileHandle.close()
# record wav
def recordWav(self, filename, time, fs, nchannel):
pa = pyaudio.PyAudio()
save_buffer = ''
buffer_size = 1000
stream = pa.open(
format=pyaudio.paInt16,
channels=1,
rate=fs,
input=True,
frames_per_buffer=buffer_size)
read_time_per_second = fs / buffer_size
cnt = 0
while cnt < time * read_time_per_second:
str_data = stream.read(buffer_size)
save_buffer += str_data
cnt += 1
wave_data = np.fromstring(save_buffer, dtype=np.short)
self.writeWav(filename, "wb", wave_data, fs, nchannel)
save_buffer = ''
# play wav
def playWav(self, filename):
fileHandle = wave.open(filename, "rb")
p = pyaudio.PyAudio()
stream = p.open(
format=p.get_format_from_width(fileHandle.getsampwidth()),
channels=fileHandle.getnchannels(),
rate=fileHandle.getframerate(),
output=True)
data = fileHandle.readframes(1024)
while data != '':
stream.write(data)
data = fileHandle.readframes(1024)
stream.close()
p.terminate()
# plot wav
def plotWav(self, data, fs, nchannel):
if 2 == nchannel:
length = len(data[0])
time = np.arange(0, length) * (1.0 / fs)
pl.subplot(211)
pl.plot(time, data[0])
pl.subplot(212)
pl.plot(time, data[1], c="g")
pl.xlabel("time (seconds)")
pl.show()
else:
length = len(data)
time = np.arange(0, length) * (1.0 / fs)
pl.plot(time, data)
pl.show()
# demo
if __name__ == "__main__":
data = ''
data = cProcessAudio().readWav("1.wav", "rb")
# cProcessAudio().writeWav("new_1.wav", "wb", data, 8000, 1)
# cProcessAudio().plotWav(data, 8000, 2)
# cProcessAudio().playWav("1.wav")
cProcessAudio().recordWav("new_2.wav", 5, 8000, 1)