Is my audio uploaded?

No. The file is decoded and pitch-tracked by your browser inside the page, so unreleased material never leaves your device.

What material does it work on?

One line at a time: a sung vocal, a lead synth, a bassline, a whistled idea, a solo instrument. Chords, layered parts and full mixes are outside what a single-line pitch tracker can read.

Every 11 ms the audio is pitched with the YIN cumulative-mean-normalised difference function, readings are median-smoothed, and runs of the same semitone become notes at least 80 ms long. Each note reports its mean frequency, cents deviation and periodicity clarity.

Why is one note reported an octave off?

Autocorrelation methods can lock onto twice or half the true period, especially on a low or heavily filtered source. That is the classic failure of this family of algorithms, so the octave in the note name is worth confirming by ear.

Why did one held note become several rows?

Vibrato, a slide or a wide pitch drift crosses a semitone boundary, and the tracker honestly reports what it measured rather than guessing that you meant one note.

Can I get the key or the chords too?

The scale table ranks the scales that fit the notes actually heard. For tempo, key and Camelot code use the BPM & Key Finder, and for harmony from a full arrangement use the Chord Finder from Audio.

Free tool

Melody Finder from Audio

The notes of a single line, read straight from a file — with frequency, tuning offset and the clarity of each reading shown, so you can see what the tracker actually heard.

Your file is decoded and analysed in this browser tab. Nothing is uploaded, nothing is stored.

Your audio

Analysed in this tab. Nothing is uploaded or stored.

Choose an audio file

WAV, MP3, FLAC, AIFF, M4A, OGG — up to 250 MB. One line at a time reads best.

Detail

How this works

The audio is pitched roughly ninety times a second with YIN, a periodicity method that measures how well the waveform repeats at each candidate period rather than guessing from a spectrum. Readings are median-smoothed, then consecutive frames sitting on the same semitone are merged into one note of at least 80 milliseconds.

Each row keeps the raw measurement alongside the note name: mean frequency in hertz, the cents deviation from equal temperament — useful for spotting a flat vocal take or an instrument tuned away from A440 — and a clarity value from the periodicity itself.

It follows one voice at a time by design. Feed it a stem, a solo take or a bounced lead rather than a full mix, then use the Chord Finder from Audio for harmony and the BPM & Key Finder for tempo and key.

FAQ

Frequently asked questions

Toolbox

See all free ANYANO tools

Train your sound. Generate your song.

ANYANO is the bigger version of this: prepare music you own, train a Sound that stays private to your account, then generate from it.