Residency at NOVARS: Machine learning & Musique concrete

I’m joining NOVARS Research Centre for Innovation in Sound at the University of Manchester as artist in residence to explore systems for machine learning and musique concrete 2020-2021. The residency is part of European Art Science Technology Network for Digital Culture a partnership with 14 EU institutions. My research question is “How can concrete materials and neural networks project future sonic realities?”

Building upon my conceptual research into ML and music, the residency will allow me time, space and support to develop my technical skills in order to create my own systems to realise the idea for the AURA MACHINE neural network. The AURA MACHINE architecture takes inspiration from Russian projectionism, thinking around Walter Benjamin’s aura of an object, musique concrete and sample culture.

I’ll be creating a training dataset of concrete materials and echoes of industry and processing through generative neural network architectures, considering process, acousmatic sound and issues of bias, labour and automation at each stage of system development.

I’ll be using Python and Max to create ML systems for sound sculpture and live electronics and am really excited to be collaborating with the wonderful people at PRiSM, Centre for Practice & Research in Science & music at the Royal Northern College of Music through our new ML and music working group across the two institutions. I am really looking forward to being part of this new creative community for machine learning in Manchester and gaining support and insight from academic specialists in electroacoustic music and data science.

I will be starting a research website/blog for the project, until this is ready, here is a talk I gave setting out the research framework residency at ZKM inSonic Festival in December 2020.

inSonic talk at ZKM