Tutorial

This section covers the fundamentals of developing with audioviz, including a package overview and basic usage. We will assume basic familiarity with Python and NumPy/SciPy.

Overview

The audioviz package is structured as collection of submodules:

General Analysis

Analyze the waveform and spectrogram of an audio.

• plot_waveform

• signal_RMS_analysis

• plot_spectrogram

Pitch Analysis

Analyze the pitch and plot the pitch class of an audio.

• pitch_estimation

• plot_constant_q_transform

• pitch_class_histogram_chroma

• hzRevampHelper

• pitch_class_histogram_f0

Beat Analysis

Detect the onsets and track the beats of an audio.

• onsets_detection

• plot_onset_strength

• beat_tracking

• predominant_local_pulses

• static_tempo_estimation

• plot_tempogram

Chord Analysis

Analyze the chord and it's composition of an audio.

• compute_chromagram_from_filename

• chord_recognition_template

• chord_decoder

• plot_chord_recognition

• plot_binary_template_chord_recognition

• generate_chord_templates

• get_chord_labels

Structure Analysis

Analyze the structure of an audio.

• compute_sm_dot

• plot_feature_ssm

• SSM_chorma

• plot_self_similarity

• compute_kernel_checkerboard_gaussian

• compute_novelty_ssm

• SSM_Novelty

• SSM_Novelty_user_selection

Timbre Analysis

Analyze the timbre of an audio.

• spectral_centroid_analysis

• rolloff_frequency_analysis

• spectral_bandwidth_analysis

• harmonic_percussive_source_separation

Basic Usage Example

First, we need to import the module we'll need for the following analysis. Then load the audio we want to analysis. We load an example audio from librosa.

#import the module we needed
from audioviz import *
import librosa
import numpy as np

# load an example audio from librosa for analysis
y, sr = librosa.load('Mozart_Turkish_March.wav', duration = 45)

Or you can load your local file by the path of your audio.

filename = 'your audio file path'
y, sr = librosa.load(filename , duration = 45)

By calling the function in audioviz, we can simply get the values and their visualization of waveform, pitch, onset, chord, structure and timbre of the audio we to analyze.

from audioviz import GeneralAnalysis
GeneralAnalysis.plot_waveform(y = y, sr = sr, start_time=8, end_time=24)

from audioviz import PitchAnalysis
PitchAnalysis.pitch_estimation(y = y, sr = sr)

from audioviz import BeatAnalysis
BeatAnalysis.onsets_detection(y = y, sr = sr)

from audioviz import ChordAnalysis
ChordAnalysis.plot_binary_template_chord_recognition('./Mozart_Turkish_March.wav', 'anno.csv')

from audioviz import StructureAnalysis
StructureAnalysis.SSM_chorma('./Mozart_Turkish_March.wav', 'anno.csv')

from audioviz import TimbreAnalysis
TimbreAnalysis.spectral_centroid_analysis(y = y, sr = sr)

The module needed to import for each specific analysis can be referred to the API DOCUMENTATION for more details.