Atmospheres of Inference

Jean Brazeau's MFA Project
Friday, September 9 & Saturday, September 10, 2022
Doors at 8:00 PM | Performance at 8:30 PM | FREE
Studio T – SFU Goldcorp Centre for the Arts
149 West Hastings Street, Vancouver

There is a persisting affection for analog audio equipment developed over the second half of the 20th century. Unlike instruments based on digital technologies, early analog instruments exhibit unpredictable behaviours contingent on their environmental conditions and the components which furnish their circuit boards. These components and their acute material properties, drastically affect an instruments’ performance and functionality. In a world organised around the binary logic of the digital (discrete processes of quantification, capture, validation, and storage), the analog (continuous, unstable, and entropic) demarcates a space of exception laden with contingency.

To this end, a significant issue emerges when considering how early analog instruments are to be preserved and historicized within institutional collections, as the more signal that flows through their circuits, the closer these instruments step towards failure; exhausted capacitors, dead diodes, spent fuses, etc. render such instruments unusable according to their designated musical use. Institutions which still possess these machines will eventually be confronted with having to choose whether these objects are more valuable as silent cultural artefacts with the integrity of their electronics intact, or as functional musical instruments furnished with inauthentic components. As the conditions of industrial electronics manufacturing change and components become increasingly inaccessible, it is inevitable that the finality of all analog synthesisers will eventually be silence. Taking this into account, an archival imperative emerges: how do we proceed with not only archiving the material instrument, but also the processual functions of the instrument?

Over the past ten-years, the advent of recursive strategies in digital signal processing, ie. machine-learning, has established new plateaus for computing continuity from nonlinear behaviour. As such, the divide which formerly separated the two regimes of electronic technics is becoming more opaque, as the contingency which formerly textured analog technologies can now be inferred with sufficient computational resources. Such inferences can be highly unpredictable, yet this attribute may make such methods uniquely suited for capturing the technical essence of analog synthesis.

Through research-creation, Jean Brazeau’s Masters in Fine Arts has been focused on archival strategies for the analog synthesisers dwelling in Simon Fraser University’s collection. With a particular focus on the Serge Modular Music System and its unique architecture that promotes a nonlinear approach to composition and performance, his thesis concert Atmospheres of Inference positions the failure of analog-based electronic instruments alongside the folly of contemporary machine-listening technologies— namely, their challenged capacity to adequately hear and reproduce the formal characteristics of the media they are trained on.

For this performance, Brazeau has trained a series of deep-learning algorithms (VAEs) on his extensive archive of recordings conducted over the course of his MFA, and considers the failure of both technologies as an enriching site of musical production. Performing alongside the high-dimensional models on a seven-channel loudspeaker system, Brazeau attends to the fading division between the analog and the digital.

This performance is presented in partial fulfilment of the requirements of the degree of  Masters in Fine Arts at Simon Fraser University.