The binaural display of clouds of point sources

We propose a binaural display that auralizes clouds of nearby point sources by filtering a monaural source with multiple HRTF pairs. To overcome the computational burden that this paradigm introduces, we propose a two-stage filter implementation in which the first filter is common to all spatial locations in the cloud, and the second filter provides the spectral cues unique to each location within the cloud. We find that much of the computational burden can be shifted to the first filter, in which case the net computational complexity of the display reduces to that of a conventional binaural display