Advanced I/O Use Cases

This section covers advanced use cases for input and output which go beyond the I/O functionality currently provided by librosa.

Read specific formats

librosa uses soundfile and audioread for reading audio. As of v0.7, librosa uses soundfile by default, and falls back on audioread only when dealing with codecs unsupported by soundfile (notably, MP3, and some variants of WAV). For a list of codecs supported by soundfile, see the libsndfile documentation.

Note

See installation instruction for PySoundFile here.

Librosa’s load function is meant for the common case where you want to load an entire (fragment of a) recording into memory, but some applications require more flexibility. In these cases, we recommend using soundfile directly. Reading audio files using soundfile is similar to the method in librosa. One important difference is that the read data is of shape (nb_samples, nb_channels) compared to (nb_channels, nb_samples) in librosa.core.load. Also the signal is not resampled to 22050 Hz by default, hence it would need be transposed and resampled for further processing in librosa. The following example is equivalent to librosa.load(librosa.util.ex(‘trumpet’)):

import librosa
import soundfile as sf

# Get example audio file
filename = librosa.ex('trumpet')

data, samplerate = sf.read(filename, dtype='float32')
data = data.T
data_22k = librosa.resample(data, samplerate, 22050)

Blockwise Reading

For large audio signals it could be beneficial to not load the whole audio file into memory. Librosa 0.7 introduced a streaming interface, which can be used to work on short fragments of audio sequentially. librosa.stream cuts an input file into blocks of audio, which correspond to a given number of frames, which can be iterated over as in the following example:

import librosa

sr = librosa.get_samplerate('/path/to/file.wav')

# Set the frame parameters to be equivalent to the librosa defaults
# in the file's native sampling rate
frame_length = (2048 * sr) // 22050
hop_length = (512 * sr) // 22050

# Stream the data, working on 128 frames at a time
stream = librosa.stream('path/to/file.wav',
                        block_length=128,
                        frame_length=frame_length,
                        hop_length=hop_length)

chromas = []
for y in stream:
   chroma_block = librosa.feature.chroma_stft(y=y, sr=sr,
                                              n_fft=frame_length,
                                              hop_length=hop_length,
                                              center=False)
   chromas.append(chromas)

In this example, each audio fragment y will consist of 128 frames worth of samples, or more specifically, len(y) == frame_length + (block_length - 1) * hop_length. Each fragment y will overlap with the subsequent fragment by frame_length - hop_length samples, which ensures that stream processing will provide equivalent results to if the entire sequence was processed in one step (assuming padding / centering is disabled).

For more details about the streaming interface, refer to librosa.stream.

Read file-like objects

If you want to read audio from file-like objects (also called virtual files) you can use soundfile as well. (This will also work with librosa.load and librosa.stream, provided that the underlying codec is supported by soundfile.)

E.g.: read files from zip compressed archives:

import zipfile as zf
import soundfile as sf
import io

with zf.ZipFile('test.zip') as myzip:
    with myzip.open('stereo_file.wav') as myfile:
        tmp = io.BytesIO(myfile.read())
        data, samplerate = sf.read(tmp)

Download and read from URL:

import soundfile as sf
import io

from six.moves.urllib.request import urlopen

url = "https://raw.githubusercontent.com/librosa/librosa/master/tests/data/test1_44100.wav"

data, samplerate = sf.read(io.BytesIO(urlopen(url).read()))

Write out audio files

PySoundFile provides output functionality that can be used directly with numpy array audio buffers:

import numpy as np
import soundfile as sf

rate = 44100
data = np.random.uniform(-1, 1, size=(rate * 10, 2))

# Write out audio as 24bit PCM WAV
sf.write('stereo_file.wav', data, samplerate, subtype='PCM_24')

# Write out audio as 24bit Flac
sf.write('stereo_file.flac', data, samplerate, format='flac', subtype='PCM_24')

# Write out audio as 16bit OGG
sf.write('stereo_file.ogg', data, samplerate, format='ogg', subtype='vorbis')