FORGET GATE

Google Brain’s Magenta, a research project exploring the role of machine learning in the process of creating art and music, was the basis of the compositional process. A dataset consisting of a large body of MIDI data drawn from existing Metaludios was used to train several Magenta models.

The training lasted for approximately 50 hours. Subsequently, hundreds of MIDI files were generated, of which three were finally selected. To improve readability, these MIDI files were then re-notated in Sibelius (notation software), adding phrasing, pedalling (based on the MIDI sustain information) and some expression marks. The music has also been redistributed between the hands in a more pianistic, practical way. Other than that, no modifications have been made to the original machine-generated MIDI files in terms of rhythm, tempo, pitch, dynamics, and pedal (sustain) information.  

As a composer, I am both fascinated and mystified by the result. I can clearly hear reminiscences of other Metaludios, although nothing is ever repeated literally. It is evident that the algorithm has learnt from the dataset and it is able to generate material “in my own style”. I believe that machine learning has a great potential as a creative tool for the composer.