Perceptually-Motivated Sonification of Spatiotemporally-Dynamic CFD Data

Lucas Temor,Daniel E. MacDonald,Thangam Natarajan,Peter W. Coppin,David A. Steinman

DOI: https://doi.org/10.21785/icad2021.039

2021-06-01

Abstract:Everyday perception and action are fundamentally multisensory. Despite this, the sole reliance on visualization for the representation of complex 3D spatiotemporal data is still widespread. In the past we have proposed various prototypes for the sonification of dense data from computational fluid dynamics (CFD) simulations of turbulent-like blood flow, but did not robustly consider the perception and associated meaning-making of the resultant sounds. To reduce some of the complexities of these data for sonification, in this work we present a feature-based approach, applying ideas from auditory scene analysis to sonify different data features along perceptually-separable auditory streams. As there are many possible features in these dense data, we followed the analogy of “caricature” to guide our definition and subsequent amplification of unique spectral and fluctuating features, while effectively minimizing the features common between simulations. This approach may allow for better insight into the behavior of flow instabilities when compared to our previous sonifications and/or visualizations, and additionally we observed benefits when some redundancy was maintained between modalities.