Mesa (2016)

Recreating a live electronics performance from 1966.

Platforms: Python
Read the scientific article // Get the source code

The Mesa project, led by Jonathan Goldman (Associate Professor of Musicology in the Faculty of Music of the University of Montreal), focused on the digital recreation of the eponymous 1966 live electronic work by American experimentalist composer Gordon Mumma.

Mesa was originally a duo for bandoneon and home-made electronics, performed by composer and pianist David Tudor and Mumma himself. Shortly after its composition, it also became the soundtrack to the Merce Cunningham‘s choreography Place.

The project’s goal was to reproduce the electronics used in the performance of Mesa using modern digital technology. The resulting software would be used in a live performance with Goldman on the bandoneon and composer Ofer Pelz operating the digital electronics. This new digital version was presented at the MontrĂ©al/Nouvelles Musiques festival on February 27th 2017, at the Agora Hydro-Quebec (MontrĂ©al).

Jonathan Goldman and Ofer Pelz rehearsing for Mesa.
Ofer Pelz using the MESA software and Jonathan Goldman on the bandoneon (Photo: Francis Lecavalier, 2017)

For this project, I programmed the electronics of Mesa using the Python language, as well as the pyo sound library and wxPython for the user interface. I worked closely with Goldman and Pelz to create an instrument that was historically accurate and easy to operate in a live setting. I also had the tremendous luck to be able to communicate with Mumma directly, in order to get precisions on the inner workings of the original Mesa electronics.

Draft of the signal processing path of Mesa (Photo: Francis Lecavalier, 2016)

The MESA software allowed the operator to modulate a signal coming from one of the four contact microphones placed on the bandoneon with the signal from another contact microphone, using ring modulation. Additional sources were available for modulation as well (sine and noise generators) and the resulting output was spatialized over four discrete outputs. The user could also easily control the application using USB MIDI devices, by defining his own custom mapping.

Screenshot of the MESA software (Photo: Francis Lecavalier, 2017)