The Magenta project is a 'neural synthesizer' that lets amateur musicians create new sounds for their work

Technology has seeped into every aspect of our lives, informing not just what we do but how we do it. Music is perhaps a prime example of this in terms of the mediums we use, alongside the very audio we create in and of itself. With the digitization of instruments, we’ve unlocked a suite of sounds never before possible, which bares the question—are there more sounds out there to be created and heard?

Under the Magenta project, brought to you by the same folks from Google Brain and DeepMind, the tech giant answers that question with a resounding ‘yes’ while pointing to its latest creation NSynth (Neural Synthesizer). Through this creation, even the most amateur of musicians will be capable of synthesizing new jingles (as the name suggests) for use in their personal work.

The platform works in an interesting way, morphing, blending and manipulating sounds of thousands of instruments to create original notes for use in original works. The team’s Helena Yeung explains:

“It uses a huge database of sounds, and the NSynth team then collects notes from approximately a thousand instruments, feeding them into a neural network. The neural net is able to run algorithms to learn about these notes, and then create a mathematical vector for each one. By doing so, it results in a machine that can mimic the sound of each instrument that can also combine these sounds.”

Complete with pre-installed sliders and two-dimensional interfaces that allow for seamless merging between two to four instruments, one can achieve some pretty funky new sounds with NSynth, which is of course the very purpose of the project. Billed as a means of ‘making music and art using machine learning,’ the creators hope the platform will be used to ‘explore the audible space between four different instruments all at once.’

While its unclear what the next steps are for the Magenta, its exciting to think of the possibilities bestowed to anyone with garage band or any of its software equivalents. The implications the tool has on emerging artists and world-renowned ones alike are huge; perhaps in the next few years we’ll see more compositions being generated with never-before-heard sounds, factoring in to the already large collection of digital-only musicians.

Magenta

Technology has seeped into every aspect of our lives, informing not just what we do but how we do it. Music is perhaps a prime example of this in terms of the mediums we use, alongside the very audio we create in and of itself. With the digitization of instruments, we’ve unlocked a suite of sounds never before possible, which bares the question—are there more sounds out there to be created and heard?

Under the Magenta project, brought to you by the same folks from Google Brain and DeepMind, the tech giant answers that question with a resounding ‘yes’ while pointing to its latest creation NSynth (Neural Synthesizer). Through this creation, even the most amateur of musicians will be capable of synthesizing new jingles (as the name suggests) for use in their personal work.