Abstract
The present invention provides a method of composition operation of high efficiency head related transfer function (HRTF), which uses head related transfer functions to perform spatial locating of a sound source, in order to address the problem of consuming a great amount of resources in using the conventional convolution algorithm, which leads to poor efficiency of processing. Spatial data matrix of the head related transfer function is decomposed with singular value decomposition and its order is reduced to decrease the amount of data to be processed, and to obtain high precision locating of the sound source, in which interpolation is employed to calculate the spatial locating information. The present invention is able to achieve the effect of generating continuously movable sound image for target sound source by featuring effectiveness of high-speed operation and simulation of omni-directional spatial data.