Localization of sound by humans has been shown to depend on a transformation of incident sounds by the pinnae, or external ear. The ears function as a computer-steerable array similar to an electronically swept radar antenna. The form of transformation is that of time delays. Autocorrelation of the time delays by mental function provides localization.nIt has been found that the ability to localize sounds in another environment may be reproduced using microphones adapted with ear replicas and high-quality condenser headphones. Extension of this technique to underwater use has been effectively demonstrated, despite some component shortcomings. The experimental highlights that support the theory and the field tests to evaluate the operational utility of localizing systems are discussed.nThe basic concept of autocorrelation of time delays introduced by the pinna has been extended to speech recognition problems. A new theory of human audition, which ascribes significance to the time domain rather than the frequency domain, has been developed that explains not only binaural and monaural localization, but also the "cocktail party effect," pitch discrimination, speech recognition, masking, intelligibility in reverberation, and other auditory phenomena.
展开▼
