%0 Generic
%A Bavister, P
%D 2023
%F discovery:10190486
%I European Acoustics Association (EAA)
%K Evolutionary Computation, Biometric Sensing,  Neural Networks, Virtual Acoustics
%P 6411-6418
%T Biometrically Evolved Site-Specific Music as a Response to Localised Acoustic Conditions
%U https://discovery.ucl.ac.uk/id/eprint/10190486/
%X This paper describes a series of experiments to engage  intelligent systems and biometric sensing in a reciprocally  creative relationship between computer composed music  and physical space, accessing correlations between spatial  volume, materiality, and performance.  The paper will review tests undertaken at UCL in 2022, in  evolving site-specific music using virtual acoustics,  evolutionary programming, and biometric sensing. The  paper will describe and define the toolsets involved, neural  networks to determine note generation and periodicity,  evolutionary processes to determine sequencing and  emotional response as a fitness function and summarises  how these were applied. The analysis of the outputs is  compared against the room’s acoustic data for correlations  and relationships. Metrics to seek correlations against  tempo fluctuations are T30, T20, EDT and C80.  The output of the tests gives us clues as to what future  music is likely to appeal emotionally in such spaces for  differing listeners’ demographics. As the spaces chosen  were not typical acoustic musical venues, there are no  preconceived ideas about what would or should not  sound acceptable in each. If music can evolve to suit a  space, then surely each space, however acoustically  detrimental, can host something that can be viewed as  aesthetically pleasing and site specific.
%Z © 2023 First author et al. This is an open-access article  distributed under the terms of the Creative Commons Attribution  3.0 Unported License, which permits unrestricted use, distribution,  and reproduction in any medium, provided the original author and  source are credited.