Taking advantage of geometric acoustics modeling using metadata

Virtual reality has recently experienced resurgence in popularity, and thanks to the advances in processor technology over the last decade it has become feasible to use geometric acoustics as a method of simulating the acoustics of a virtual environment using the same scene model data. Usually, the highest priority is realism, however this is not always favourable, especially for spatial audio. The acoustics of a virtual environment may not be pleasing. This paper discusses the development of a custom geometric acoustics program, a data structure (Raw Impulse Vector) produced by the program that provides open access to all of the captured rays including metadata about each ray, and finally the types of processing that can be performed on this data in order to perceptually optimise the acoustics. This can be of great benefit to interactive audio systems that employ a geometric acoustics algorithm as a method of artificially simulating reverb where the characteristics of the reverb can be optimised based on the listener’s preferences.