How does Vocal Highway detect my vocal pitch in real time?

Vocal Highway samples your microphone input via the Web Audio API and runs an in-browser YIN difference algorithm with parabolic sub-sample interpolation to determine the fundamental frequency of your voice up to 60 times per second.

Do I need headphones to play Vocal Highway?

Headphones are strongly recommended when using microphone input. Headphones prevent computer speaker audio from looping back into your microphone, guaranteeing zero false triggers.

Can I play without a microphone?

Yes! Vocal Highway features an on-screen synth keyboard mapped to keys A through K on computer keyboards and touch surfaces on mobile phones.

Vocal Highway | Real-Time Vocal Pitch Trainer & Sight-Singing Arcade

The Science of Real-Time Pitch Tracking & The YIN Algorithm

Singing accurately requires immediate neurological biofeedback: your brain must compare the acoustic sensation in your inner ear to the intended frequency target within milliseconds. Vocal Highway runs an in-browser implementation of the YIN fundamental frequency algorithm, developed by de Cheveigné and Kawahara.

Unlike naive zero-crossing algorithms or computationally heavy Fourier transforms (FFT) that suffer from harmonic octave-jumping, YIN analyzes the cumulative mean normalized difference function of the incoming audio buffer. This extracts the exact fundamental frequency (f₀) of human vocal fold vibrations, allowing Vocal Highway to evaluate your singing pitch in real time without server latency.

Understanding Cent Deviation and Micro-Intonation

In modern equal temperament (12-TET), each octave is divided into 1,200 cents (100 cents per semitone). The human ear can detect pitch differences as small as 5 to 10 cents in isolated tones:

Perfect (0 to ±14 cents): Studio-grade intonation requiring no digital pitch correction.
Great (±15 to ±35 cents): Musical vocal pitch with natural vibrato and acoustic drift.
Acceptable (±36 to ±85 cents): Unambiguously the correct note, but noticeably sharp or flat.

Frequently Asked Questions

When using computer speakers, synthesized backing tracks and drum beats play into the room and enter your microphone. Headphones isolate your voice, ensuring the YIN algorithm only analyzes your vocal folds.

Each phrase follows a 16-beat cycle: during the Listen phase, the synthesizer demonstrates the melody while you memorize the notes. When the gate switches to the green Sing phase, you match the pitches on beat.

Authored & Engineered by Austin Ross

Austin Ross is an audio DSP developer and music producer behind Austin Aux. He designs precision DSP algorithms, web audio engines, and creative tools for artists and audio engineers worldwide.