Dhwani

Dhwani is a vocal tuner designed specifically for Hindustani classical music. It helps singers practice and improve their intonation by providing real-time feedback on the swar (note) being sung. Dhwani is available as both a web application and a reusable JavaScript library.

Features (Web App & Library)

• Real-time Pitch Detection: Analyzes audio input and identifies the closest Hindustani swar.
• Just Intonation: Uses a just intonation tuning system based on common Hindustani classical frequencies (starting with Sa at 240 Hz).
• Multiple Octaves: Handles multiple octaves, including Mandra, Madhya, Taar, and Ati-Taar saptaks.
• Noise Reduction: Includes basic noise reduction.

Web App Features

• Swar Dial Display: A visually intuitive dial shows the current swar, with neighboring swaras fading out to the sides.
• Flat/Natural/Sharp Indicators: Indicates whether the sung note is flat (komal), natural (shuddh), or sharp (tivra) relative to the ideal swar frequency.
• Detailed Information: Displays the octave, detected frequency (in Hz), and the difference in cents from the ideal swar frequency.
• Responsive Design: Works well on desktops, tablets, and mobile devices.
• Lightweight and Fast: Built with just HTML, CSS, and JavaScript.

How to Use (Web App)

1. Open dhwani.mgks.dev in a modern web browser (Chrome, Firefox, Edge, or Safari recommended).
2. Grant the website permission to access your microphone.
3. Begin singing a Hindustani swar. The tuner will display the detected swar and provide visual feedback.
4. Humming is suggested for better accuracy.

Using the Dhwani Library

The Dhwani library provides a simple API for pitch detection.

Installation

You can include dhwani.js directly in your HTML:

<script type="module">
    import { Dhwani } from './dhwani.js'; // Adjust path if needed

    // Your code using the Dhwani library here
</script>

Or, you can use it within a module bundler (like Webpack, Rollup, or Parcel):

#  This project doesn't have a package.json, so direct npm install isn't applicable
#  You would typically use:  npm install <package-name>
#  But for this project, you just include the dhwani.js file directly.

API

Dhwani Class

• constructor(sampleRate: number, threshold?: number)

  • Creates a new Dhwani instance.
  • sampleRate: The sample rate of the audio context (e.g., audioContext.sampleRate).
  • threshold (optional): The threshold for pitch detection (between 0 and 1). Lower values are more sensitive but may be more prone to noise. Defaults to 0.1. Adjust this for optimal performance.

• getPitch(buffer: Float32Array): number | null

  • Analyzes an audio buffer and returns the detected pitch in Hz.
  • buffer: A Float32Array containing the audio data (e.g., from AnalyserNode.getFloatTimeDomainData()).
  • Returns the frequency in Hz if a pitch is detected, or null if no pitch is detected.

Example (Library Usage)

import { Dhwani } from './dhwani.js'; // Adjust path if necessary

async function startPitchDetection() {
    try {
        const audioContext = new AudioContext();
        const stream = await navigator.mediaDevices.getUserMedia({ audio: true });
        const source = audioContext.createMediaStreamSource(stream);
        const analyser = audioContext.createAnalyser();
        analyser.fftSize = 2048;
        source.connect(analyser);
        const buffer = new Float32Array(analyser.fftSize);

        const dhwani = new Dhwani(audioContext.sampleRate, 0.15); // Adjust threshold

        function updatePitch() {
            analyser.getFloatTimeDomainData(buffer);
            const frequency = dhwani.getPitch(buffer);

            if (frequency) {
                console.log("Detected Frequency:", frequency.toFixed(2), "Hz");
            } else {
                console.log("No pitch detected");
            }

            requestAnimationFrame(updatePitch);
        }

        updatePitch();

    } catch (error) {
        console.error("Error:", error);
    }
}

startPitchDetection();

Technology

Dhwani is built using:

• HTML: For the structure of the web page (web app).
• CSS: For styling and visual presentation (web app).
• JavaScript: For real-time audio processing, pitch detection, and user interface updates (web app and library).
• Web Audio API: For accessing the microphone and analyzing audio.

Development

The project is completely free and open-source.

Contributions, bug reports, and feature requests are welcome!

License

This project is licensed under the MIT License - see the LICENSE file for details.

Support the Project

GitHub Sponsors: Support this project and my other work by becoming a GitHub sponsor, it means a lot :)

Follow Me on GitHub | Add Project to Watchlist | Star the Project