Raga Detector

A Python-based system for detecting and analyzing Carnatic classical music ragas from audio files using pitch extraction and pattern matching.

Overview

This project provides tools to:

• Extract pitch contours from audio using the pYIN algorithm
• Identify the tonic (Sa) frequency
• Analyze swara (note) distributions
• Match detected patterns to known Carnatic ragas
• Generate the complete 72 Melakarta raga system programmatically

Project Structure

├── raga_detector.py          # Main raga detection system
├── generate_melakartas.py    # Generates all 72 Melakarta ragas
└── requirements.txt          # Project dependencies

Installation

1. Clone or download this repository
2. Install dependencies:

   pip install -r requirements.txt

Requirements

• librosa (>=0.10.0) - Audio analysis and pitch extraction
• soundfile (>=0.12.0) - Audio file I/O
• scipy (>=1.11.0) - Scientific computing
• numpy (>=1.24.0) - Numerical operations
• scikit-learn (>=1.3.0) - Machine learning utilities
• matplotlib (>=3.7.0) - Data visualization
• seaborn (>=0.12.0) - Statistical visualization
• tqdm (>=4.65.0) - Progress bars

Quick Start

Run the raga detector on an audio file:

python raga_detector.py

Generate all 72 Melakarta ragas:

python generate_melakartas.py