Proc. 1999 IEEEWorkshop on Applications of Signal Processing to Audio and Acoustics, New Paltz, New York, Oct. 17-20, 1999 UNIFORM SPREADING OF AMPLITUDE PANNED VIRTUAL SOURCES Ville Pulkki Laboratory of Acoustics and Audio Signal Processing Helsinki University of Technology P.O.Box 3000, FIN-02015 HUT, Finland Ville.Pulkki@hut.fi ABSTRACT The perceived spatial spread of amplitude panned virtual sources is dependent on the number of loudspeakers that are used to pro- duce them. When pair-wise or triplet-wise panning is applied, the number of active loudspeakers varies as a function of the panning direction. This may cause unwanted changes in spatial spread and coloration of a virtual source if it is moved in the sound stage. In this paper a method is presented to make the directional spread of amplitude panned virtual sources independent of their panning direction. This is accomplished by panning the sound signal to multiple directions near each other simultaneously. This forms a single virtual source with constant directional spread as a function of direction. 1. INTRODUCTION In many theaters and auditoriums there exists loudspeaker systems that include a large number of loudspeakers. Audio systems with multiple loudspeakers are also becoming more common in domes- tic use. In domestic use the speakers are most often placed to a horizontal plane around the listener (two-dimensional speaker se- tups). In theaters there are also systems in which also elevated or descended loudspeakers exist (three-dimensional setups). Virtual sources can be positioned to such loudspeaker systems using various methods. In two-dimensional setups the panning is most often performed using pair-wise panning methods [1]. Pair- wise panning can be generalized to triplet-wise panning for three- dimensional loudspeaker setups [2]. Pair-wise panning and triplet- wise panning yields an acceptable virtual source quality in rela- tively large listening area. However, the virtual source quality is dependent on panning direction because the number of loudspeak- ers emanating the same sound signal varies in different directions. This may be perceived inconvenient when moving virtual sound sources are applied. In matrixing sound reproduction techniques, like Ambisonics [3], sound signals are encoded analogically to few audio channels. In the decoding stage loudspeaker signals are decoded using some matrixing operations. In these techniques the same sound signal is existent in all loudspeakers, which may degrade the virtual source quality. However, when moving sound sources are applied, the number of loudspeakers producing a virtual source is not depen- dent on panning direction, thus the directional spread does not vary prominently. Basically the matrixing systems and pair-wise or triplet-wise panning methods are quite similar techniques. In both methods a sound signal is applied to a number of loudspeakers with different amplitudes. The goal of this study is to form methods for virtual source positioning that would be something in between the matrix- ing techniques and pair-wise or triplet-wise panning. This paper is organized as follows: At first some basics of spatial hearing and amplitude panning are reviewed. The method for spreading amplitude-panned virtual sources is then presented. The directional spread of pair-wise panned virtual sources and spreaded virtual sources are computed using a binaural model. 2. SPATIAL HEARING Spatial and directional hearing has been studied intensively; for overviews, see e.g. [4]. The duplex theory of sound localization states that the two main cues of sound source localization are the interaural time difference ITD and the interaural level difference ILD which are due to wave propagation time difference (primarily below 1.5 kHz) and the shadowing effect by the head (primarily above 1.5 kHz), respectively. Both cues depend on frequency. In the median plane where the distances to both ears are equal, the ITD and ILD values are close to zero. Other effects, such as spectral cues and head movements, are considered to carry eleva- tion and front-back information in the median plane. Spatial dis- crimination is difficult also in so called cones of confusion where ILD and ITD vary only slightly due to the unsymmetry of the head. A cone of confusion can be approximated with a cone which has symmetry axis in the line passing through listener’s ears. If the cues of an auditory object are conflicting, the listener may perceive the object in multiple directions simultaneously. The directional spread denotes how well the auditory object is situated in one direction: what smaller directional spread, that more point- like auditory object. 3. AMPLITUDE PANNING Amplitude panning [5] is most often applied to two loudspeak- ers which are in a standard stereophonic listening configuration, as depicted in Fig. 1. Loudspeakers 1 and 2 are placed in front of the listener with aperture of . A signal is applied to each loudspeaker with different amplitudes, this can be formulated as (1) where is the signal to be applied to loudspeaker , is the gain factor of the corresponding channel, and is the time param- eter. The sound signals arrive to listener’s both ears from both loud- speakers. If the wave propagation time difference is taken into ac- count, but the shadowing effect of the head is neglected, we may W99-1 Copyright © 1999 IEEE. Reprinted from Proc. of the 1999 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics. Mohonk Mountain House, New Paltz, New York. This material is posted here with permission of the IEEE. Such permission of the IEEE does not in any way imply IEEE endorsement of any of Helsinki University of Technology's products or services Internal or personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution must be obtained from the IEEE by sending a blank email message to pubs-permissions@ieee.org. By choosing to view this document, you agree to all provisions of the copyright laws protecting it.