Working with sounds in Python

Working with sounds in Python

I didn't understand how to handle sounds in Python, so I summarized it roughly.

Basically, I referred to this site Playing and Recording Sound in Python.

Regeneration

An example of a library that plays audio

• Playsound A simple package just for playing WAV and MP3 files
• Simpleaudio Library that can play WAV files and NumPy arrays
• winsound Library for Windows that can play WAV files and beeps
• [python-sounddevice](# python-sounddevice), pyaudio Providing a PortAudio library for playing WAV files
• pydub pyaudio is also required, but by installing ffmpeg you can play a wide range of audio formats with just a few lines of code.

playsound

Only the playsound function is implemented, and a WAV file or MP3 file is specified and played.

install

pip install playsound

usage

from playsound import playsound

playsound("sample.wav")

simpleaudio

Play a WAV file or NumPy array. It is also possible to determine whether it is playing.

install

pip install simpleaudio

usage

import simpleaudio

wav_obj = simpleaudio.WaveObject.from_wave_file("sample.wav")
play_obj = wav_obj.play()
play_obj.wait_done()

#Check if it is playing
if play_obj.is_playing():
	print("still playing")

winsound

For Windows, is winsound available by default? WAV files and beeps can be played.

usage

import winsound

winsound.PlaySound("sample.wav", winsound.SND_FILENAME)

#Beep playback
import winsound

winsound.Beep(1000, 100) #Play 1000Hz beep for 100ms

python-sounddevice

Provides binding for the PortAudio library. Provides functionality for playing / recording audio signals and NumPy arrays.

install

If you are using Anaconda, you can install it with the conda package

conda install -c conda-forge python-sounddevice

usage

import sounddevice as sd
import wave
import numpy as np

wf = wave.open("sample.wav")
fs = wf.getframerate()
data = wf.readframes(wf.getnframes())
data = np.frombuffer(data, dtype='int16')

sd.play(data, fs)
status = sd.wait()

The above handles the entire song at once, but when performing real-time processing, it is necessary to handle the audio signal in short units by callback processing.

import sounddevice as sd

duration = 5.5

def callback(indata, outdata, frames, time, status):
    if status:
        print(status)
    outdata[:] = indata

with sd.Stream(channels=2, callback=callback):
    sd.sleep(int(duration * 1000))

pydub

install

pip install pydub

When using non-WAV files such as MP3, you also need to install ffmpeg or libav.

In addition, you need to install one of the following to play audio (simpleaudio is recommended)

• simpleaudio
• pyaudio
• ffplay (included with ffmpeg)
• avplay (included with libav)

usage

from pydub import AudioSegment
from pydub.playback import play

sound = AudioSegment.from_file("sample.wav", format="wav")
play(sound)

pyaudio

Provides binding for the PortAudio library. Easy audio playback and recording on cross-platform.

Playing simple audio is more complicated than other methods, so if you just want to play a file, you should choose another method.

However, it is possible to set input / output devices and check latency for control at a lower level. Also, since it can be described in callback, it is suitable for performing relatively complicated processing.

install

pip install pyaudio

#Also possible with conda
conda install pyaudio

usage

import pyaudio
import wave
import sys

chunk = 1024

wf = wave.open("sample.wav", "rb")
p = pyaudio.PyAudio()
stream = p.open(format=p.get_format_from_width(wf.getsampwidth()), channels=wf.getnchannels(), rate=wf.getframerate(), output=True)
data = wf.readframes(chunk)

while data != '':
    stream.write(data)
    data = wf.readframes(chunk)

stream.stop_stream()
stream.close()
p.terminate()

recording

An example of a library that can record audio

• [python-sounddevice](# python-sounddevice) Record from a microphone and keep it as a NumPy array
• pyaudio Record from a microphone and keep it as a binary

python-sounddevice

usage

import sounddevice as sd
from scipy.io.wavfile import write

record_second = 3
fs = 44100

myrecording = sd.rec(int(record_second * fs), samplerate=fs, channels=2)

write('output.wav', fs, myrecording)

pyaudio

usage

import pyaudio
import wave

chunk = 1024
format = pyaudio.paInt16
channels = 2
fs = 44100
record_second = 3

p = pyaudio.PyAudio()
stream = p.open(format=format, channels=channels, rate=fs, input=True, frames_per_buffer=chunk)

print("* recording")

frames = []

for i in range(int(fs / chunk * record_second)):
    data = stream.read(chunk)
    frames.append(data)

print("* done recording")

stream.stop_stream()
stream.close()
p.terminate()

wf = wave.open("output.wav", "wb")
wf.setnchannels(channels)
wf.setsampwidth(p.get_sample_size(format))
wf.setframerate(fs)
wf.writeframes(b''.join(frames))
wf.close()

device

Since you may want to specify the playback / recording device, check how to change it in each library.

python-sounddevice

Show available devices and selected devices

import sounddevice as sd
sd.query_devices()

Execution example

You can also use the following command on the terminal.

python -m sounddevice

The device can be selected by setting the device ID to default.device or assigning it toplay ()andStream ()as a device argument.

import sounddevice as sd
sd.default.device = 1, 5