Recording Audio
Now just switching to the recording mode of the article. Few of the above-mentioned libraries are used for the same, both playing and recording can be explained collectively but for many, it might get a bit confusing. So different dedicated sections are preferred here.
Note- Before using any library for recording, make sure the microphone of your device is actually connected and is ON also not Muted. one can check the same using operating system features and settings.
Method 1. Using python-sounddevice
This library allows you to play (explained above) and record NumPy arrays containing audio signal information. This module requires scipy or wavio to save the recorded audio, this means scipy or wavio library should be pre-installed along with Numpy before using this package for recording.
Python3
# import required librariesimport sounddevice as sdfrom scipy.io.wavfile import writeimport wavio as wv# Sampling frequencyfrequency = 44400# Recording duration in secondsduration = 3.5# to record audio from# sound-device into a Numpyrecording = sd.rec(int(duration * frequency), samplerate = freq, channels = 2)# Wait for the audio to completesd.wait()# using scipy to save the recording in .wav format# This will convert the NumPy array# to an audio file with the given sampling frequencywrite("recording0.wav", freq, recording)# using wavio to save the recording in .wav format# This will convert the NumPy array to an audio# file with the given sampling frequencywv.write("recording1.wav", recording, freq, sampwidth=2) |
Output:
As stated in the Audio section sounddevice had a default option to specify the channel and frequency for the repeated use. After that no need to pass this option as arguments in sd.rec() method. Following code represents the same along with one can also change the data type of recorded array from default type of float32 to others.
Python3
import sounddevice as sdsd.default.samplerate = 4400sd.default.channels = 2myrecording = sd.rec(int(duration * fs))# change the data type: pass a new argument in .rec() of dtype# myrecording = sd.rec(int(duration * fs), dtype='float64')sd.wait() |
Simultaneous Playback & Recording
To Play an array named my_arr and Record at the same time. Here sample rate is smpl_rate
Python3
import sounddevice as sdimport numpy as npsmpl_rate = 44100my_arr = np.random.uniform(-1,1,smpl_rate)recordd= sd.playrec(my_arr, smpl_rate, channels=2)sd.wait() |
Method 2: Using pyaudio.
As above, we played audio using pyaudio by reading the pyaudio.Stream(). To record audio we must write to this same stream. Following is the code to record a few seconds of audio and save the same to a .wav file:
Python3
import pyaudioimport wave# Record in chunks of 1024 sampleschunk = 1024 # 16 bits per samplesample_format = pyaudio.paInt16 chanels = 2# Record at 44400 samples per secondsmpl_rt = 44400 seconds = 4filename = "path_of_file.wav"# Create an interface to PortAudiopa = pyaudio.PyAudio() stream = pa.open(format=sample_format, channels=chanels, rate=smpl_rt, input=True, frames_per_buffer=chunk)print('Recording...')# Initialize array that be used for storing framesframes = [] # Store data in chunks for 8 secondsfor i in range(0, int(smpl_rt / chunk * seconds)): data = stream.read(chunk) frames.append(data)# Stop and close the stream stream.stop_stream()stream.close()# Terminate - PortAudio interfacepa.terminate()print('Done !!! ')# Save the recorded data in a .wav formatsf = wave.open(filename, 'wb')sf.setnchannels(chanels)sf.setsampwidth(pa.get_sample_size(sample_format))sf.setframerate(smpl_rt)sf.writeframes(b''.join(frames))sf.close() |
Output:
How to Play and Record Audio in Python?
As python can mostly do everything one can imagine including playing and recording audio. This article will make you familiar with some python libraries and straight-forwards methods using those libraries for playing and recording sound in python, with some more functionalities in exchange for few extra intended python lines.
Most of the audio files are in MP3 and WAV file formats. WAV audio files are the simplest digital audio format with lossless high recording rates as a result WAV files are large compared to other formats. For the same reason, MP3 formats are used which are small in size and compresses files with very little difference to overall sound quality. Also, it is very easy to convert WAV to MP3 with open-source and free software which are widely available over the internet.