YouTube tests a search feature where users hum to identify songs

YouTube announced a new experiment on Android devices that determines a song via humming — which seems like a major step up from Apple’s music recognition app Shazam.

As noted on YouTube’s support page, the video-sharing platform is testing a search-by-song capability on the Android version of the app that allows users to figure out a song on YouTube by humming, singing or recording a song.

Users who have access to the experiment can toggle from YouTube voice search to the new song search feature and hum, sing or record a song for three or more seconds. The platform then identifies the tune and directs the user to relevant YouTube videos featuring the searched song, whether that be the official music video, user-generated content or Shorts.

The search-by-song capability is only available to a small portion of Android users. If the feature rolls out more widely, we can see it being helpful for many, as YouTube is a popular destination for looking up songs.

YouTube’s latest experiment probably sounds familiar to some users. In 2020, YouTube’s parent company Google first launched the capability on the Google app, Google Search widget and Google Assistant, letting users figure out a song by humming, whistling or singing into the microphone icon. However, the main difference appears to be that Google’s feature requires users to hum for 10-15 seconds in order to identify the song.

As Google previously explained, its feature is built on machine learning models that can match a person’s hum to a song’s “fingerprint” or signature melody. The new YouTube test uses the same technology as the Google feature, the company confirmed to TechCrunch.

Other music recognition apps like SoundHound and MusixMatch can also identify songs by singing or humming the tune, but they aren’t as popular compared to YouTube and Google. (Still, we recommend checking them out as well).