Google’s Magenta Team Creates AI NSynth Super Synthesizer
Can artificial intelligence influence the arts? One Google research team certainly thinks so. Known as Magenta the team’s first project was an algorithm called Nsynth that was first unveiled last year. Now they’ve developed an interface that lets others get their hands on the technology.
Short for Neural Synthesizer, the NSynth algorithm is designed to learn from sounds it is presented with and generate new sounds from the information. These creations are not just blends but are entirely new sounds that people might otherwise struggle to create themselves.
In the latest development Magenta teamed up with Google Creative Lab to create NSynth Super. This interface lets users get hands on with NSynth. The interface can be played using any MIDI device such as a DAW, sequencer or keyboard.
In this video the team at Magenta talk about the creation of NSynth and the journey from a lot of heavy math, big computers and lots of code to something a musician could get their hands on and create music with.
To create this prototype Magenta recorded 16 sounds across 15 pitches and fed them into the algorithm. The algorithm used this information to create more than 100,000 unique sounds that were then loaded into the prototype. Each of the four dials on the Super was assigned four different sounds. Musicians can pick what sources they want their music to come from then manipulate the output using the touchscreen.
In the video below London-based producer Hector Plimmer shows off what can be achieved using the NSynth Super.
If you want an NSynth Super of your own there is a catch. Right now there are no plans to manufacture the NSynth Super. Instead the open source project is available on GitHub along with all the templates and instructions you need to create your own NSynth Super.