TorchAudio Load Audio with Specific Sampling Rate – TorchAudio Tutorial

In this tutorial, we will introduce you how to read an audio with specific sampling rate in torchaudio. It is very useful for audio processing.

How to load an audio with specific sampling rate in torchaudio?

Here are two steps:

Step 1: load an audio

We can use torchaudio.load() function to load an audio. Here is the tutorial:

Understand torchaudio.load(): Read Audio with Examples – TorchAudio Tutorial

Step 2: resampling an audio

We can use torchaudio.functional.resample() to resample an audio. Here is the detail.

TorchAudio Audio Resampling Tutorial for Beginners

Here we will create a function to implement it.

import torchaudio
wav_file = "008554.wav"

def read_audio(wav_file, sample_rate = 8000):
    wav_data, sr = torchaudio.load(wav_file)
    print(wav_data.shape)
    if sr != sample_rate:
        wav_data_2 = torchaudio.functional.resample(wav_data, sr, sample_rate)
        print(wav_data_2.shape)
    return wav_data_2
wav_data_2 = read_audio(wav_file)

Run this code, we will see:

torch.Size([1, 230496])
torch.Size([1, 38416])