A Subject-Independent 3D Sound Using Adaptive Head and Pinna Models

• Main Authors: Cheng-Ta Chang, 詹承達
• Other Authors: Oscal T.-C. Chen
• Format: Others
• Language: zh-TW
• Published: 1999
• Online Access: http://ndltd.ncl.edu.tw/handle/23376173578365927513

碩士 === 國立中正大學 === 電機工程研究所 === 87 === With a progress of the 3D animation for virtual-reality applications, the 3D sound system has become an important research topic in recent year. Unlike the AC-3 audio system, the 3D sound system based on the head-related transfer function requires less hardware facilities. It can provide users easily to engage various multimedia applications in a low- cost entertainment system. In this thesis, we present an efficient 3D sound system by using head and pinna models. Only two loud speakers are used to generate the sound localization. The interaural time difference and interaural intensity difference are modeled by the head shadow. In addition, the pinna is constructed by calculating the coordinate positions of the polygon surfaces of its mesh structure which is utilized to physically characterize sound reflection. These models for 3D sound generation require low computation complexities and are easily to fit the sound-perceptual characteristics of an individual listener. In addition to constructing models, we realized an efficient measurement system of the HRTF by using a stretched pules waveform. The measured results with high signal-to-noise ratios can be obtained. The nonlinearity and distortion of the measurement equipment are minimized. By using this measurement system, we can easily and precisely get the results of the HRTF for 3D sound applications.