[Sound Analysis] Sound-data Preprocessing : Librosa

Librosa ? 

Librosa is a python package for music and audio analysis. It provides the building blocks necessary to create music information retrieval systems.

Install

pip install librosa

Example Code 

1) Import library 

import librosa
import librosa.display
import numpy as np
import os
os.environ['LIBROSA_DATA_DIR'] = './sound_example/'

2) Load sound data 

# Load 'libri1' example file.

## (example of library) Get file name
filename = librosa.example('libri1')
print(f"filename:{filename}")   # ./sound_example/librosa/5703-47212-0000.ogg

## Load an audio file as a floating point time series.
y, sr = librosa.load(filename)   
#### # y, sr = librosa.load(filename, sr=16000)  # sr=none : native sampling rate
print(f" {y}, {sr}")

y : sound time serias. [np.ndarray]

sr : sampling rate. [number>0]

 

3) Extract feature : MFCC (Mel-frequency cepstral coefficients) 

# Generate mfccs from a time series
librosa.feature.mfcc(y=y, sr=sr)
''' Output 
array([[-565.919, -564.288, ..., -426.484, -434.668],
       [  10.305,   12.509, ...,   88.43 ,   90.12 ],
       ...,
       [   2.807,    2.068, ...,   -6.725,   -5.159],
       [   2.822,    2.244, ...,   -6.198,   -6.177]], dtype=float32)
'''     


# Use a pre-computed log-power Mel spectrogram
S = librosa.feature.melspectrogram(y=y, sr=sr, n_mels=128, fmax=8000)
librosa.feature.mfcc(S=librosa.power_to_db(S))

''' Output 
array([[-559.974, -558.449, ..., -411.96 , -420.458],
       [  11.018,   13.046, ...,   76.972,   80.888],
       ...,
       [   2.713,    2.379, ...,    1.464,   -2.835],
       [   2.712,    2.619, ...,    2.209,    0.648]], dtype=float32)
'''

# Get more components
mfccs = librosa.feature.mfcc(y=y, sr=sr, n_mfcc=40)
print( len(mfccs) )  # 40

4) Visualize the MFCC series

import matplotlib.pyplot as plt

fig, ax = plt.subplots(nrows=2, sharex=True)

S = librosa.feature.melspectrogram(y=y, sr=sr, n_mels=128, fmax=8000)

img = librosa.display.specshow(librosa.power_to_db(S, ref=np.max),
                               x_axis='time', y_axis='mel', fmax=8000,
                               ax=ax[0])
fig.colorbar(img, ax=[ax[0]])
ax[0].set(title='Mel spectrogram')
ax[0].label_outer()
img = librosa.display.specshow(mfccs, x_axis='time', ax=ax[1])
fig.colorbar(img, ax=[ax[1]])
ax[1].set(title='MFCC')

Output