摘要
From the point of spatial sampling, spatial interpolation of HRTFs （head-related transfer functions） and signal mixing for multichannel （surround） sound are analyzed. First, it is proved that they are mathematically equivalent. Different methods for HRTFs interpolation are equivalent to different signal mixing methods for multichannel sound. Then, a stricter derivation for the signal mixing of multichannel sound and the law of sine for stereophonic sound is given. It is pointed out that trying to reconstruct lateral HRTFs by adjacent linear interpolation is wrong. And for accurate sound image localization, the conventional equation of adjacent linear interpolation of HRTFs is revised. At last, it is also pointed out that some methods used in the analysis of HRTFs and multichannel sound can be used for reference mutually.
From the point of spatial sampling, spatial interpolation of HRTFs （head-related transfer functions） and signal mixing for multichannel （surround） sound are analyzed. First, it is proved that they are mathematically equivalent. Different methods for HRTFs interpolation are equivalent to different signal mixing methods for multichannel sound. Then, a stricter derivation for the signal mixing of multichannel sound and the law of sine for stereophonic sound is given. It is pointed out that trying to reconstruct lateral HRTFs by adjacent linear interpolation is wrong. And for accurate sound image localization, the conventional equation of adjacent linear interpolation of HRTFs is revised. At last, it is also pointed out that some methods used in the analysis of HRTFs and multichannel sound can be used for reference mutually.
基金
This work was supported by the National Natural Science Foundation of China (No.10374031).
