Reproduction of simulated acoustic scenes for clinical and hearing research in a reverberant room using four horizontal loudspeakers

When reproducing a virtual acoustic environment over loudspeakers within a reverberant playback room, the acoustics of the playback space can modify the spectral and spatial properties of the virtual environment considerably. Traditionally, optimal loudspeaker rendering of such scenes requires dedicated loudspeaker setups positioned in an anechoic room, employing techniques like Vector-BaseAmplitude Panning (VBAP) to render virtual reverberant sources, or Higher Order Ambisonics (HOA) to render spherical harmonics each using a large number of loudspeakers. In this study we evaluate to what extent it is possible to reproduce virtual acoustic environments using only a limited number of loudspeakers placed within a normal echoic room. Recently, we proposed a perceptually-based method using only four loudspeakers that specifically aims to compensate the detrimental effects of reverberation of the playback room by separately reproducing optimized versions of the direct and reverberant sound fields [Fallah et al., Reproduction of simulated acoustic scenes with limited number of loudspeakers in a reverberant room (2023)]. In this study, this approach is explained in more detail, and specifically an important parameter that controls the power ratio between direct and reverb sounds is investigated in detail using objective and subjective evaluations. In listening tests, the similarity of this proposed Acoustic Room Transformation (ART) method is compared to that of a reference rendering of the virtual acoustic environment within an anechoic room. The results of listening tests show significant improvements in the timbral and spatial characteristics of reproduced sound using the ART method compared to conventional playback without room compensation and show a closer match to the reference simulated environment.