Google now lets you search for songs by humming the tune

19 Oct 2020

Image: © Davizro Photography/Stock.adobe.com

Google Search has introduced a new tool to find music that rivals platforms such as Midomi and Shazam.

Google is introducing a new feature that enables users to search for songs that they don’t know the name of by humming, whistling or singing the tune.

The new feature was announced in a company blogpost by Krishna Kumar, senior product manager at Google Search. It is available on the latest version of the mobile Google app or the Google Search widget.

“We all know how frustrating it is when you can’t remember the name of a song or any of the words but the tune is stuck in your head,” Kumar wrote. “Google can now help you figure it out – no lyrics, artist name or perfect pitch required.”

Google’s new tool for song searching rivals technologies such as SoundHound’s Midomi and Apple-owned Shazam. While Midomi allows users to search for music by singing or humming a tune, Shazam currently only allows users to find songs if they are played by the original artist – not hummed or sang by users.

Using the new feature

Outlining how users can access the new hum-to-search feature, Kumar said: “Open the latest version of the Google app or find your Google Search widget, tap the mic icon and say ‘what’s this song?’ or click the ‘search a song’ button. Then start humming for 10 to 15 seconds.”

The feature can also be used with Google Assistant if users say ‘Hey Google, what’s this song?’ and then hum the tune. It is initially available in English on iOS and in more than 20 languages on Android. Kumar said that Google hopes to expand this to more languages in the future.

After a song is hummed, Google’s machine learning algorithm helps to identify potential song matches. Users can then select the best match and explore information on the song and artist, view any accompanying music videos, find the lyrics, read analysis or listen to the song through their streaming app of choice.

“A song’s melody is like its fingerprint: they each have their own unique identity,” Kumar added. “We’ve built machine learning models that can match your hum, whistle or singing to the right ‘fingerprint’.”

He explained that when songs are hummed into the search feature, Google’s machine learning models transform the audio into a number-based sequence representing the song’s melody. The models are trained to identify songs based on a variety of sources, including humans singing, whistling or humming, as well as studio recordings.

Kelly Earley was a journalist with Silicon Republic

editorial@siliconrepublic.com