🎵 YIN Pitch Detection in Rust

Implementation of the YIN algorithm for estimating the fundamental frequency of an audio signal in Rust. Optimized for speed, and with parabolic interpretation for refining the result.

Background

The YIN algorithm (Cheveigné & Kawahara, 2002) estimates the fundamental frequency (f₀) of a signal by minimizing the Cumulative Mean Normalized Difference Function (CMNDF).

This implementation follows the standard YIN steps:

1. Difference function (DF)
2. Cumulative mean normalization (CMNDF)
3. Absolute thresholding
4. (Parabolic interpolation for refined lag estimate)

The resulting best lag gives the estimated period, and therefore the frequency:

$$ f = \frac{f_s}{\text{lag}} $$

De Cheveigné, A., & Kawahara, H. (2002). YIN, a fundamental frequency estimator for speech and music. The Journal of the Acoustical Society of America, 111(4), 1917–1930. https://doi.org/10.1121/1.1458024

@article{de-cheveigne-2002,
	author = {De Cheveigné, Alain and Kawahara, Hideki},
	journal = {The Journal of the Acoustical Society of America},
	month = {4},
	number = {4},
	pages = {1917--1930},
	title = {{YIN, a fundamental frequency estimator for speech and music}},
	volume = {111},
	year = {2002},
	doi = {10.1121/1.1458024},
	url = {https://doi.org/10.1121/1.1458024}
}

Example Usage

use yin_rs::Yin;

fn main() {
    let sample_rate = 44100.0;
    let frequency = 440.0;
    let duration = 2.0;

    let signal: Vec<f64> = (0..(duration as usize * sample_rate as usize))
        .map(|n| (2.0 * std::f64::consts::PI * frequency * n as f64 / sample_rate).sin())
        .collect();

    let yin = Yin::init(0.1, 50.0, 2000.0, sample_rate);

    let result = yin.yin(&signal);
    println!(
        "Detected frequency: {:.2} Hz",
        result.get_frequency() // 441
    );
    println!(
        "Using parabolic interpolation, detected frequency: {:.2} Hz",
        result.get_frequency_with_interpolation() // 440.02
    );
}