Processing, Sound
Leave a comment

Automatic Orchestra – Algorithmic sound composition and networked music

The Automatic Orchestra is an audio installation exploring algorithmic composition and networked music. A common set of rules distributed among a network of MIDI devices opens up a melodic space orchestrated by automatic logic. The perpetual interaction among the devices and the interpretation of encoded musical messages blurs the distinctiveness between structured composition and performative improvisation.

The setup consists of 12 pods. One pod consists of a microcontroller mounted on a custom PCB board which is attached to a set of speakers. All pods are wired together to form a circular network transmitting musical data. Therefore the data travels through each unit before it is passed on to its neighbour. The synthesis of the audio parameters depends on the application running on each pod. A shared framework provides a synchronized basis for the exchange of encoded musical messages but each pod will interpret and alter the data based on its individual algorithmic rule set.


Hardware included Teensy 3.1 on CFO BODYSEQ, Monopod, Speakers, multiple plug for power and 3.5mm audio jack cables to connect MIDI ports. Software included Processing IDE (for developing prototypes) with the libraries including rwmidi, oscP5, audio synthesis with JSyn, and Arduino IDE CFO library.

In 2015 the project was first presented at the Resonate festival in Belgrade, Serbia. Algorithmic compositions were developed in situ at the conference in a three-day closed workshop where visitors were invited to observe the development process and engage with the artists. The workshop ended in a 10 minute-long performance. Afterwards the installation was presented in the entrance hall of the conference.

The project was a collaborative effort between students of the University of the Arts, Bremen and Copenhagen Institute of Interaction Design headed by Dennis P. Paul and Jakob Bak. Participating students were: David Beermann, Hendrik Heuer, Irena Kukric, Julian Hespenheide and Thomas Hoheisel.

Project Page


Leave a Reply

Your email address will not be published. Required fields are marked *