Tom Mudd

Algorithms Playing Algorithms: Coding Music via the Algorithmic Creation of Score and Instrument Files for the NESS Physical Model

{kind: title}

This talk will reflect on my recent explorations of the Next Generation Sound Synthesis set of physical models created by researchers at the University of Edinburgh. These models are sophisticated numerical simulations of a range of acoustic situations, notably guitar/fretboard interactions, brass instruments, bowed strings, 2D membranes, and diffraction models of sound propagation in acoustic spaces.

My use of the models is as a musician interested in sonically exploring via algorithmic processes. My process has generally involved the iterative creation of a range of algorithms that attempt to play the NESS instrument algorithms: simulations of performers playing simulations of instruments. A notable aspect of this process is that the results are very difficult to hear in terms of digital synthesis processes. Due to the convincing nature of the simulations the results are much more suggestive of the logic and culture of the instruments being simulated. For example, if the model is of a guitar, the listener will likely picture a human performer, or at the very least, a physical instrument, as the source of the music, even if they are very much aware of the digital nature of the process. Along with this comes the whole culture of strumming and plucking string instruments: the variations in the simulated instrument and the simulated performance conjure different waypoints from the histories of guitar cultures and guitar performances. Even for me as the creator, it is very difficult to conceive of the results as digital synthesis rather than instrumentally created music, opening up a strange aesthetic space.

Despite their simulated nature, I will elaborate the material-oriented nature of my own use of the algorithms, in the mould of Keep's notion of instrumentalising (Keep, 2009). Comparing my own work with that of others who have used the NESS models is instructive in trying to unpick just how the nature of the models—and the nature of the interaction with these models—shapes the outcome, shedding light on the capacities and tendencies that emerge when composers engage with the models.

Excerpts of my work with the models can be heard here (guitar):
and here (brass):

The models themselves can be explored at

Keep, A (2009) Improvising with sounding objects in experimental music. In The Ashgate Research Companion to Experimental Music, page 113 – 130. Ashgate Publishing Limited.


meta: true

event: Simulation

author: Tom Mudd