How The Heck Does Shazam Work?

Shazam identifies songs by converting audio captured from a phone’s microphone into a spectrogram using the Fast Fourier Transform (FFT), which breaks sound into frequency components over time. It then extracts the loudest peaks from this spectrogram to create a sparse "constellation map," forming unique fingerprint hashes by pairing nearby peaks based on frequency and timing. These hashes are matched against a massive database for near-instant identification, making the system robust against noise but less effective for humming or covers. The process allows Shazam to recognize songs quickly and accurately with only a few seconds of audio.