How to Use Google Hum to Search to Find Any Tune

Catch That Tune: How Google Learned to Hear the Music in Your Head

We have all been there: a stray melody loops endlessly in your brain. You don’t know the lyrics, you don’t know the artist, and you certainly don’t know what album it came from. For decades, describing a tune meant awkwardly tapping a table or singing random syllables to an equally confused friend.

Then, Google changed the game by letting users hum, whistle, or sing a melody straight into their search bar. Often referred to colloquially as searching by tune, this feature transformed how we interact with audio technology, turning a vague mental itch into an exact match in seconds.

The Anatomy of an Earworm

To understand why “hum to search” or tuning into a melody is a technical marvel, you have to look at what happens when you hum. Everyone’s pitch, tone, and vocal accuracy vary wildly. Some people are pitch-perfect; others sound like a dying engine.

When you hum or whistle into the Google app or ask Google Assistant “What’s this song?”, your audio input isn’t matched like a traditional fingerprint. Instead, Google’s machine learning models convert your audio stream into a number-based sequence representing a sort of numerical “song profile” or melody line.

Think of it like a fingerprint for the contour of a tune—the distinct ups and downs of the notes, regardless of whether you sang it in the right key or used the right instrument. The system then compares this numerical sequence against a vast database of recorded songs, instrumental versions, and live performances to find the closest statistical matches.

Step-by-Step: How to Search by Tune

Using Google’s melody-matching feature takes just a few seconds across mobile devices: