Google Can Help You Find A Song If You Hum

google search for a song

Yesterday, Google announced that it is adding a new hum to search feature to its search tools. This would allow the users to just hum the tune and search for any song that is stuck in their heads.

Notably, this could be very beneficial for those who could not remember the lyrics of the song. You can just hum the tune of the song and Google will try to identify the song with that tune and show you the results.

It is a common case scenario, where a nice song playing at the coffee shop, stucks into our head. Due to some of the other reasons, we could not recall the lyrics of the songs.


You can literally search for the song that goes “da daaaa da da daaaa na naa naa ooohh yeah?” and Google will use the machine learning capabilities to search and identify the song the resembles the hum that you searched in the first place.

Besides, users would be able to hum, whistle, or sing a melody and let Google do its job for searching the song. However, you need to make sure that you have the latest version of the Google app installed on your device.

Google hum to search feature is available on the Google app for both iOS and Android

Good thing is that Google has not limited this feature only to the Google App on Android. Instead, the iOS users can also take advantage of this Google hum to search feature on their Apple devices.


You simply need to tap on the mic icon and say “what’s this song?” or click the newly added “Search a song” button and then hum the tune you remember for 10-15 seconds.

Google will make use of the inbuilt machine learning and will try to closely show the result that is likely a match. Users will be able to listen to the search results and take note of the song’s name if that is perfect to your hum.

Once the search is complete, you can view the matching videos, listen to the audio that is correct to your hum, find the lyrics of the song, and much more.


To go into the details, Google explains this feature using machine learning models to “transform the audio into a number-based sequence representing the song’s melody.” It then compares the melody with the existing songs.

The machine learning models are trained on a variety of sources. For instance, humans singing, whistling, or humming, as well as studio recordings.

Apparently, this feature is currently available in English on iOS and in more than 20 languages on Android. Google will expand this to more languages as the days go on.