A recently released Google Magenta project enables a user to mix lo-fi and hip-hop music tracks to create a customized music room in the browser. The feature doesn’t require any musical ability. The Magenta is designed to automatically sync and use Google’s machine learning systems for creating art and music, and the Lo-Fi Player is a creative example of what Magenta can do.

When Lo-Fi Player opens, users are automatically directed to a pixellated virtual “room” where one can click various objects, like a clock, cat or a piano in a room. These sounds can be blended to change the different tracks, like the bass line and the song melody. “When the view outside the window relates to the music playing in the background of a track, a person can change both visuals and the music by clicking on the window,” Lo-Fi Player creator Vibert Thio stated. He further states that the creative team chose the format of a music-generating room instead of the contemporary composition tool or a musical instrument as the music generation aspect is new and “a popular genre with a relatively simple music structure.” The feature is powered by magenta.js, the open-source JavaScript API to use in Magenta in-browser.

Google Magenta’s Lo-Fi Player

Virtual music room, Virtual music studio

Goggle Magenta Lo-Fi player image

Lo-Fi Player virtual music room also has an excellent interactive YouTube stream, which is a “shared space” where people from various places can listen and be in the same music room together. However, at that time, instead of clicking on elements present in the room, the players would be able to type commands into the live chat window to arrange and blend the tracks.

Virtual Music Room

Google’s open-source TensorFlow system powers the Magenta. This is a part of a current research project titled “Exploring the role of machine learning as an effective tool in the creative process.” Other related Magenta projects include the Piano Genie, which is an AI program that lets any person “play” the piano, and NSynth, which is a machine learning algorithm that collaborates with a neural network to learn and create unique sounds.

The Magenta Lo-Fi Player is customizable with its source code available on GitHub. Creator Thio states that the team also developed a detailed tutorial named “Play, Magenta!” where users can mix and edit sounds and canvas live on their web browsers. Thio also stated that Lo-Fi Player could not replace human producers or the existing lo-fi hip hop streams. “You can think of the Magenta Lo-Fi Player as a virtual music studio prototype for creating an interactive music piece or an introduction to the lo-fi genre to help people appreciate the art more,” Thio stated.

About Magenta Lo-Fi Player

Through the Lo-Fi Player, one can interact with elements of the room and create custom music room. Listeners can also share their music room with others. Machine learning models from magenta.js power the new experience.  “Lo-Fi Player” is a virtual music room in a browser that let users play with music beats. The feature choose Lo-Fi Hip Hop genre as it a popular one with simple music structure. The Lo-fi feature also enables the creation of new music that makes sense to listeners.