Day #213 - Working with Sound and Python

	#pip install librosa
	import librosa
	import matplotlib.pyplot as plt
	import librosa.display
	import numpy as np
	path = r'D:\PetProject\Audio_Analytics\UrbanSound.tar\UrbanSound\data\air_conditioner\204240.wav'
	#waveform `y`
	#Store the sampling rate as `sr`
	#By default, this uses resampy’s high-quality mode (‘kaiser_best’).
	y,sr = librosa.load(path,res_type='kaiser_fast')
	plt.figure(figsize=(10,5))
	librosa.display.waveplot(y,sr=sr)
	#Mel frequency cepstral coefficients (MFCCs)
	#small set of features
	mfcc = librosa.feature.mfcc(y=y, sr=sr)
	print(mfcc.shape)
	#numpy.ndarray of size (n_mfcc, T)
	mfccs=np.mean(librosa.feature.mfcc(y=y,sr=sr,n_mfcc=40).T,axis=0)
	print(mfccs)
	print(mfccs.shape)
	#Load 5 seconds of a wav file, starting 15 seconds in
	y,sr = librosa.load(path,res_type='kaiser_fast', offset=15.0, duration=5.0)
	plt.figure(figsize=(10,5))
	librosa.display.waveplot(y,sr=sr)
	#Load a wav file and resample to 11 KHz
	y,sr = librosa.load(path,res_type='kaiser_fast',sr=11025)
	plt.figure(figsize=(10,5))
	librosa.display.waveplot(y,sr=sr)

view raw sound.py hosted with ❤ by GitHub

Happy Mastering DL!!!