Skip to main content

Music identification through audio fingerprinting

Project description

Music Identification Through Audio Fingerprinting

PyPI PyPI - Python Version PyPI - License
Identifies a song from a small recording (say 30 seconds). The song has to be a part of your created database, of course.

This project is essentially a simplified version of what Shazam does.

The Idea

The flowchart on the right describes the series of steps.

Below is a brief summary. For detailed explanation with analysis, check out our Report.

We decimate the audio signal by a factor of 4 after passing it through a low pass filter (to smartly avoid aliasing). Thereafter, the signal is converted to frequency domain using the famous Fast Fourier Transform.

We take small chunks of the sample (roughly 0.3 seconds) and take the peak frequencies along a logarithmic scale. Those values are then associated with a hash value. We do so for all the songs and hence create a database.

We perform similar steps for the recorded sample. The answer is the song with the highest number of matches for a particular offset value.

Requirements

  • Python 3+.
  • pip (package installer for Python). See here for installation.
  • ffmpeg. See here for installation.
  • PortAudio. Only for Linux/OSX users. Check your distribution's repos for latest builds. Instead you can also build it from source, see here.

Installation

  • Install using pip (preferred in a virtual environment).
pip install Presto-Chango

Usage

  • On the first run, create your database by specifying the location of your songs directory.
presto-chango create-db <songs-directory>
  • Identify a song by either recording in real time or using a pre-recorded sample.
# Record in real time
presto-chango identify

# Use a pre-recorded sample
presto-chango identify --file=samples/sample1.wav
  • The algorithm returns the top five matches and the number of offsets that matched for each of them. Example
$ presto-chango identify --file="samples/sample_GAY.wav"
  Loading database
  .
  .
  .
  Database loaded

  Processing...

  Results:
  Kane Brown - Good as You (Official Music Video)_mS3TeZEp_PE.wav 41
  Katy Perry - Never Really Over (Official)_aEb5gNsmGJ8.wav 39
  Ed Sheeran - Perfect (Official Music Video)_2Vv-BfVoq4g.wav 37
  Cody Johnson - On My Way To You (Official Music Video)_RKUENGsDXBA.wav 24
  Jason Aldean - Rearview Town_WEUUvntknTI.wav 23

Building the source code

  • Clone the repository
git clone https://github.com/yashrajkakkad/presto-chango.git
cd presto-chango
  • Create a virtual environment
python -m venv venv
source venv/bin/activate
  • Install the package. The --editable flag makes it so that we don't have to reinstall everytime we make some change.
pip install --editable .

Testing

You can run the tester code if you're too lazy to record songs. It will cut random 30 second samples from songs and run the algorithm.

python tester.py

References

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for Presto-Chango, version 1.0.2
Filename, size File type Python version Upload date Hashes
Filename, size Presto_Chango-1.0.2-py3-none-any.whl (11.9 kB) File type Wheel Python version py3 Upload date Hashes View hashes
Filename, size Presto Chango-1.0.2.tar.gz (9.8 kB) File type Source Python version None Upload date Hashes View hashes

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page