SciTech

Shazam phone app redefines music

Credit: Paola Mathus/ Credit: Paola Mathus/

In years past, if you heard an amazing tune playing, you would rush to the closest person to you, with one question: “What’s the name of this song?” Others would alternatively type as many lyrics as they could discern into Google’s search bar, only to realize with much dejection that they had heard the wrong words altogether. Nonetheless, technology, as it so often does, has solved this recurring problem of the unknown song with an app called Shazam.

When a song is playing and a person wants to discover its name, they pull out their phone, launch Shazam, and start recording. After five to ten seconds, the title of the song pops up on the screen, accompanied by other information pertaining to it such as the artist, album, artwork, and year of release. Admittedly, this process elicits awe every time, but it is very much grounded in the realm of computer science.

Shazam was founded in 1999 by Chris Barton and Phillip Inghelbrecht, both University of California, Berkley students, along with Avery Wang, who was a digital signal processing Ph.D. student at Stanford University. Dhiraj Mukherjee also joined the team from a firm located in the UK. In 2002, before smartphones picked up, a user had to dial "2580" on their phone while holding it up to the music, and the user then received an SMS text with the song title and artist name. In 2008, Shazam became one of the first apps on Apple's new App Store.

Shazam employs a four-step process to detect audio picked up by a phone’s microphone. Firstly, according to their website, Shazam Entertainment Ltd. keeps a large database of 11 million songs and their unique "fingerprints." A fingerprint is the distinct sound wave pattern of an audio excerpt. Next, the user records whatever it is he or she wishes to be identified, then the sound wave of the recording is compared to the database of stored fingerprints until a match is found. When a match is found, the information connected to that correlation in the database is returned to the user. If a match is not found, as is often the case in noisy environments and for obscure music, an error is returned.

The fingerprint of a song's audio is saved as a spectrogram, which is a time-frequency graph of a wave. This is a crucial step in converting the analog sound wave into a digital format that Shazam's scripts can read and compare.

Shazam is not the only nor is it the first app to perform this function. Several competitors exist, such as Google Sound Search, SoundHound — which detects songs by having users hum — and Amazon's FireFly. Shazam is, however, the most successful and ubiquitous of the bunch. Its service has expanded beyond detection of only music, to movies, TV shows and adverts.
Shazam offers a glimpse into an exciting future for portable technology. Security, the internet of things, user interfaces, and voice commands are areas that could benefit from Shazam's model. Today, similar technology is already being used to unlock phones using voice.

Shazam is an indispensable utility on phones for millions of users. It has been downloaded ten million times, and correctly identified songs over 15 billion times. The modern marvel that is Shazam exemplifies and is a reminder of the power of ingenious computing. Now, there's no need to ask that nearby person for the name of a catchy song: one can just Shazam it.