YouTube is testing a search by hum feature on Android devices

Users will need to hum or record a playing song for three or more seconds to be suggested official music content which YouTube thinks matches up.

When you purchase through affiliate links on MusicTech.com, you may contribute to our site through commissions. Learn more
YouTube app icon (a red play button) on a smartphone device

Image: NurPhoto / Getty

YouTube is currently testing a new search tool which allows users to track down music by humming or recording a song that is currently being played.

The search tool rivals Apple-owned music search tool Shazam, which the tech giant acquired in 2018. Shazam, however, can’t recognise a song you hum yourself, but only a song being played.

YouTube’s search by hum tool is currently available to Android users only, and is confirmed by the video sharing and music app on its official help page, as of 22 August.

YouTube says that those in the experiment can toggle from the voice search feature to the new song search feature, where here you can record a playing track or hum a song. You’ll need to do this for at least three seconds, from which it will then suggest relevant music content, user-generated videos and possibly even YouTube Shorts (its rival to TikTok).

The company says it is rolling out this feature to a “small percentage” of users across the globe who use Android devices.

Google, the parent company of YouTube, launched the hum-to-search capability across the Google app, Google Search widget and Google Assistant back in 2020. YouTube’s offering appears to be slightly more advanced though, as Google’s version required a much longer 10-15 seconds of humming to identify a song.

It has not been specified exactly how the new YouTube search tool works, but Google rolled out its own AI tool, Music LM, back in May this year. Users can feed Music LM text prompts such as “soulful jazz” or “a calming violin melody backed by a distorted guitar riff” for it to create two versions of a song for you. Users can then vote on which one of the tracks they like better, which serves to “help improve the model”.

logo

Get the latest news, reviews and tutorials to your inbox.

Subscribe
Join Our Mailing List & Get Exclusive DealsSign Up Now
logo

The world’s leading media brand at the intersection of music and technology.

© 2024 MusicTech is part of NME Networks.