@chordbook/tuner

A web-based library for pitch detection of stringed instruments. It uses the Web Audio API to capture audio from the microphone, and uses the pitchy to detect the predominant pitch.

Features

• ✅ Uses your devices microphone to detect the pitch of stringed instruments
• ✅ Filters noise from low and high frequencies outside the range of stringed instruments
• ✅ Does not try to detect when the volume is too low

Installation

npm install @chordbook/tuner

Usage

import { createTuner } from '@chordbook/tuner'

const tuner = createTuner({
  // The callback to call when a note is detected.
  onNote: note => {
    console.log('Note:', note)
  },

  // Here are some other settings you can fiddle with and their default values.
  // (let us know if you find values that work better).

  // The number of milliseconds between each pitch detection.
  updateInterval: 50,

  // The frequency of middle A. Defaults to 440Hz.
  a4: 440,

  // The minimum clarity threshold. Anything below this will be ignored
  clarityThreshold: 0.9,

  // The minimum volume threshold. -1000 means 1/1000th the volume of the loudest sound.
  minVolumeDecibels: -1000,

  // The minimum and maximum frequencies to detect. To reduce noise, everything else is
  // filtered out using a lowpass and highpass filter.
  minFrequency: 27.5, // A0, Lowest note on a piano
  maxFrequency: 4186.01, // C8, Highest note on a piano

  // The sample rate to use for the audio context.
  // https://developer.mozilla.org/en-US/docs/Web/API/BaseAudioContext/sampleRate
  sampleRate: 44100,

  // The size of buffer to use for frequency analysis, which maps to the `fftSize`:
  // https://developer.mozilla.org/en-US/docs/Web/API/AnalyserNode/fftSize
  bufferSize: 8192,

  // https://developer.mozilla.org/en-US/docs/Web/API/AnalyserNode/smoothingTimeConstant
  smoothingTimeConstant: 0.8
})

// Request access to the microphone and begin pitch detection
tuner.start()

// Stop listening
tuner.stop()

When a pitch is clearly detected, the onNote callback is called with an object containing the following properties:

{
  // The frequency of the detected note
  "frequency": 612.2498364209699,
  // The node name (e.g. "A", "C#")
  "name": "D♯",
  // The note number (0-89)
  "value": 75,
  // The number of cents the detected frequency is off from the nearest note
  "cents": -29,
  // The octave of the detected note
  "octave": 5,
  // The clarity of the detected note (0-1)
  "clarity": 0.9656978357299373
}

Contributing

Contributions are welcome!

1. Clone this repository: git clone https://github.com/chordbook/tuner.git
2. Install dependencies: npm install
3. Start the development server: npm run dev
4. Open http://localhost:5173/ in your browser

License

This project is licensed under the GPLv3.0 license.