Unseen - for modular synthesiser and live visuals

Benjamin Carey

Details

Abstract

Unseen (2019) is a work for modular synthesiser and live visuals. The work was developed using images of Diffusion MRI scans of the human brain as source material, images that were kindly provided by Prof Fernando Calamante of the University of Sydney/Sydney Imaging. This work explores the concept of ‘networks’, both figuratively and literally, through a tight connection between a bespoke modular synthesis patch and dynamically processed MRI images. These source images show in great detail the connections between various parts of the brain. Such images allow researchers to understand the enormous complexity of the brain’s interconnected structure.

In this work, these source images are placed in a 3-dimensional scene and subjected to various forms of image processing and manipulation. These processed images are used training data for a neural network model trained in the Wekinator software, whose resultant outputs are converted into control voltages in response to analyses of the real-time movement of these images during a performance. These voltages are then used to control aspects of a modular synthesis patch. The sounds created by the synthesist are derived both from internal generative processes, as well as via the neural network outputs themselves, and the sound material is in turn analysed and used to manipulate the 3D scene which is being analysed by the neural network. This creates a tightly woven feedback loop between image, sound, digital information and control voltage. This interconnected feedback process is presented as a performative and compositional allegory for the complexity of the networked structure of the brain, enabling the performer to both affect, and be affected by a complex audio-visual network. Unseen was developed in response to the theme ‘Seeing the Unseen: From Brains to Black Holes’, and was commissioned by the University of Sydney as part of the 2019 Innovation Week program.

Bio

Benjamin Carey is a composer, performer and researcher. Ben’s research and practice is concerned with musical interactivity, generativity and the delicate dance between human and machine agencies in composition and performance. He has released four albums, the latest of which is an LP of modular synthesis works entitled ANTIMATTER (2019, Hospital Hill Records). Ben’s work has been performed and exhibited nationally and internationally at numerous festivals and academic conferences including the Huddersfield Festival of Contemporary Music (UK), IRCAM Live @ La Gaité Lyrique (France), the International Computer Music Conference (Australia) and the International Conference on New Interfaces for Musical Expression (USA, UK and Australia). Ben is currently Lecturer in Composition and Music Technology at the University of Sydney, Conservatorium of Music.