The Google-owned company says that while only a small percentage of Android users currently have the feature, a brand new update will be rolled out to everyone soon.
Finding tunes is now easier.
Last week, Google revealed that it will begin testing a brand new function that will let Android users search for music on YouTube just by humming or singing.
“We’re experimenting with the ability for people to search for a song on YouTube or by recording the song they’re currently playing,” according to a statement on YouTube’s help page.
Users who are part of the test program can activate the new function from YouTube’s Voice Search and “we or record the song you’re looking for for 3 seconds or more,” the page says. can.”
The video-sharing website will send the user all “relevant official music content, user-generated videos, and/or shorts featuring songs discovered in the YouTube app” as soon as the software recognizes a song.
Although only a small percentage of Android users currently have this capability, the Google-owned company has promised that everyone will get the brand-new upgrade soon.
If the upgrade is successful, the new function will compete directly with the Shazam app, which is owned by Apple.
Beta group members can go to the new voice search option on YouTube and “we or record the song you’re looking for for 3 seconds or more,” the page says.
Google says they want to “make it easier for users to find and engage with content on YouTube” in addition to relieving artists of the burden of frequent uploads.
Google has previously incorporated voice search technologies into several of its platforms.
The video-sharing website will send the user all “relevant official music content, user-generated videos, and/or shorts featuring songs discovered in the YouTube app” as soon as the software recognizes a song.
The functionality is apparently coming to the Google app, Google Search widget, and Google Assistant in 2020.
The key distinction is the need to hum the tune for 10 to 15 seconds as opposed to three.
The search engine said: “When you hum a chord in search, our machine learning models convert the audio into a number-based sequence that represents the song’s lyrics.” “Our models are trained to recognize songs from studio recordings and multiple sources, including human singing, whistling, or humming.”
“All other information is similarly removed by the algorithm, including the timbre and tone of the voice and any accompanying music. The song’s numerical fingerprint, or number-based sequence, remains,” the statement said.
“We compare these sequences to tens of thousands of songs from around the world and identify potential matches in real-time.”