The Tech Behind Shazam

US How they ID over 23,000 songs per minute      31/07/23

Shazam is one of the most impressive apps created to identify what track or song you're listening to. This video explains how the service takes small fragments of music recorded by users for identification and references a phenomenal catalog of existing productions.

It works using scatter plots, which identify the transients in the song - rather than the actual harmonic content, as shown on a spectrogram. Genius! 

https://www.shazam.com/gb/home

 

About the author [midierror]: midierror is co-founder of Potato Audio, he makes nifty Max For Live devices, award-winning Sample Packs and hosts the 'midierror meets...' interview series speaking with all kinds of people working in sound & music. 


Get the latest headlines, offers and more in our fantastic sonicstate newsletter - delivered to your in box every week

More News: Like This
Even more news...


 

Get Our Newsletter


More...

SonicLAB: Behringer FLOW4V 32-Bit Float Recorder 

And FLOW4VIO


SonicLAB: Korg Nu:Tekt NTS-4 Performance Mixer 

With FX and 2 Ch USB audio


SonicLAB: Genki Katla With Venus Theory 

We look at the volcanic synth


Lifeforms SV-2 - Pittsburgh Modular's Redesigned Semi-Modular 

SV-1 successor with thru-zero FM


SonicLAB: Pittsburgh Modular Lifeforms SV-2 

Classy analogue synth reviewed


AmiGUS mini: 24-bit Sound for Amiga 1200 & 600 

32 Channel PCMCIA Soundcard in Beta