Google is now in the process of developing a new program to synthesize sound using machine learning techniques — aka artificial intelligence. The tech giant has dubbed the program “NSynth” (Neural Synthesizer) and is the brainchild of Google Magenta, a group of AI researchers who were charged with finding out if AI is capable of creating music. The program is fed samples from various instruments, which are then analyzed by the software in order to create a set of new sounds by tapping into the mathematical characteristics of each instrument and their respective note values.
Google Magenta explains on their official website:
“Unlike a traditional synthesizer which generates audio from hand-designed components like oscillators and wavetables, NSynth uses deep neural networks to generate sounds at the level of individual samples. Learning directly from data, NSynth provides artists with intuitive control over timbre and dynamics and the ability to explore new sounds that would be difficult or impossible to produce with a hand-tuned synthesizer”.
Listen to a few samples below:
Google will be showcasing the new technology this week at Moogfest, an annual celebration of music and technology held in Durham, North Carolina.