Medusa - A Distributed Sound Environment

This paper introduces Medusa, a distributed sound environment that allows several machines connected in a local area network to share multiple streams of audio and MIDI, and to replace hardware mixers and also specialized multi-channel audio cables by network communication. Medusa has no centralized servers: any computer in the local environment may act as a server of audio/MIDI streams, and as client to remote audio/MIDI streams. Besides allowing audio and MIDI communication, Medusa acts as a distributed sound environment where networked sound resources can be transparently used and reconfigured as local resources. We discuss the implementation of Medusa in terms of desirable features, and report user experience with a group of composers from the University of São Paulo/Brazil.