Display of Three-dimensional Virtual Acoustic Sound Fields for Human Listeners: Value of Discrete Multichannel Simulation

In contrast to two other multichannel sound-field synthesis techniques, Higher Order Ambisonics (HOA) and Wavefield Synthesis (WFS), Discrete Multichannel Simulation (DMS) has distinct advantages in the display of three-dimensional virtual acoustic sound fields for human listeners. In particular, it is the impact of listener head movements on reproduction fidelity that is the primary concern, although the influence of static head acoustics on the displayed sound field is another critical factor. Whereas movement of the listener's head outside of the 'sweet spot' for HOA can result in dramatic failure in terms of both spatial and timbral performance of a sound-field display system, given a large number of loudspeakers, a value of the DMS system is that the configuration of simulated sources and reflections is fixed on each of the respective loudspeakers even though the absolute angles vary with head translation and rotation. Experiences in the design and evaluation of three distinct types of multichannel sound-field synthesis systems are presented in this paper, these three being examples of HOA, WFS, and DMS. The emphasis of this report is upon the results of listening tests that were undertaken in these three case studies, which reveal that listener head movements produce relatively less problems for the DMS system, which also showed a distinct advantage in manipulating the apparent distance of displayed virtual sources, albeit limited to virtual sources positioned outside of a listener's peripersonal space.