What is Google Magenta?
Magenta is Google’s AI Wing that aims at developing new deep learning algorithms for generating songs, images, drawings, and alternative materials. Besides this, they also build tools and interfaces that enable artists and musicians to increase, not replace, their processes using these learning algorithm models. All their AI models and tools are open source and available on GitHub.
Google’s Magenta Wing explores the role of machine learning within the processes of making art and music. It was started by some researchers and engineers from the “Google Brain Team” along with several others who have contributed tirelessly to this project.
Magenta Studio (v0.1) Beta
Magenta Studio is a collection of music plugins designed on Magenta’s AI tools and models that use machine learning techniques for generating music. These tools are offered as standalone, as well as plugins for Ableton Live!
Plugins Included in Magenta Studio.
Continue
Continue uses the predictive power of recurrent neural networks (RNN) to generate notes that are likely to follow your drum beat or melody
Generate
Similar to Continue, Generate generates a 4 bar phrase with no input necessary. Generate uses a Variational Autoencoder (VAE) that has been trained on millions of melodies and rhythms to learn a summarised representation of musical qualities. Generate then chooses a random combination of these summarised qualities and decodes it back to MIDI to produce a new musical clip.
Interpolate
Interpolate takes two drum beats or two melodies as inputs. It then generates up to 16 clips which combine the qualities of these two clips. Useful for merging musical
Groove
Groove adjusts the timing and velocity of an input drum clip to produce the “feel” of an actual drummer. Quantized drum patterns are translated into human-like performances. Magenta has recorded 15 hours of real drummers performing on MIDI drum kits for this plugin.
Drumify
Drumify is used to generate a drum accompaniment to a bassline or melody, or to create a drum track from a tapped rhythm. It works best with performed inputs, but it can also handle quantized clips.
Read more about Magenta and the open-source research project that is exploring the role of machine learning as a tool in the creative process here