This work presents a method to control the perceived distance of an auditory object by changing the directivity pattern of a loudspeaker and consequently the direct-to-reverberant ratio at the listening spot. Control of the directivity pattern is achieved by beamforming using a compact multi-driver loudspeaker unit. A small-sized cubic array consisting of six drivers is assembled, and per driver beamforming filters are derived from directional measurements of the array. The proposed method is evaluated using formal listening tests. The results show that the perceived distance can be controlled effectively by directivity pattern modification.
IntroductionPerceiving the location of a sound source is a fundamental property of the human hearing system. In addition to the perceived source direction, human hearing is sensitive to the source distance. Hence spatial sound rendering systems should be able to reproduce the perception of distance convincingly. Although auditory distance perception has been studied to some degree (see Zahorik et al., 2005 for a thorough review), there are not many studies on how to actually render auditory distance with loudspeakers in normal rooms. The aim of this study is to propose such a method. A number of acoustic cues are known to be important for the perception of source distance (Zahorik et al., 2005): sound power variation following the inverse squared-distance law, the direct-to-reverberant energy ratio (D/R-ratio), spectral changes, and binaural cues. Spectral cues result from high-frequency sound attenuation further away from the listener, and they are perceived at distances greater than about 15 m. Binaural cues result from near-field excess inter-aural differences at distances smaller than about 1 m. As these cues are significant only at very small and large distances, they are not deemed crucial for typical spatial-sound rendering systems.Controlling the sound power in a distance-dependent manner is trivial; a simple gain modification following the inverse-distance law between a point source and the listening spot can be employed. On the contrary, controlling the D/R-ratio in normal rooms is not straightforward because the room reverberation itself dominates the resulting D/R-ratio. In theory, dereverberation techniques could be applied (Mourjopoulos, 1994), but typically they work only at a single point, and they are sensitive to changes in the room acoustics. Moreover, it is not fully understood how hearing actually detects D/R-ratio changes (Zahorik et al., 2005). Both monaural (Lounsbury and Butler, 1979) and binaural (Bronkhorst, 2002) features have been proposed to be dominant in distance perception. Monaural processing could utilize phase coherence of partials over frequency (Griesinger, 2010;Laitinen et al., 2013), whereas binaural processing could utilize, e.g., fluctuations of the inter-aural level differences over time (Catic et al., 2013). Furthermore, it has been proposed that directions and delays of early reflections provide meaningful cues (Pellegr...