A New 3D Sound Reproduction Method Using an Auditory Filter

Abstract

SRS, which stands for Sound Retrieval System, is a 3D sound reproduction method that processes signals and retrieves their spatial information lost during the recording or playback process by applying head- related transfer functions. The SRS technique, however, has the problem of causing sharp noise, due to the extreme enhancement of the audible frequency band. This paper suggests a new 3D sound reproduction method that can address this problem. In order to prevent the audible frequency band from being over- enhanced, the algorithm suggested in this paper is designed to automatically restrict gain values calculated by the output ratio of the auditory filter designed after the human auditory system. From the results of computer simulations, it is confirmed that the suggested algorithm effectively prevents the audible frequency band from being over-enhanced, which is a problem of the conventional technique. In addition, it is verified in listening tests that the bass is evaluated to be comparatively inferior to the conventional SRS technique, while the clarity of sound was found to be superior.