A High - Quality and Low - Complexity Audio Codec Using Simplified Psychoacoustic Model

• Main Authors: Tsung-Chih Liao, 廖崇志
• Other Authors: Oscal T.-C. Chen
• Format: Others
• Language: zh-TW
• Published: 2000
• Online Access: http://ndltd.ncl.edu.tw/handle/40774502614017513931

碩士 === 國立中正大學 === 電機工程研究所 === 88 === In this era when multimedia technology and communication network develop at a rush speed, digital signal processing plays an important role in compressing and transmitting multimedia data more efficiently. Nowadays, audio signal processing technology has been widely applied to various multimedia systems in our daily life. In order to decrease audio encoded data and remain excellent quality at the same time, developers have created various audio codecs with high compression rates and high fidelity. However, those audio codecs also employ too much implementation complexity. In this thesis, we have developed a simplified psychoacoustic model by analyzing the properties of the human auditory system, and apply this model to the audio codecs. The structure of this codec bases on the discrete wavelet transform with perfect reconstruction. We calculate a masking threshold by the simplified psychoacoustic model, and also implement an appropriate bit allocation and quantization. Compared with the common psychoacoustic model, the proposed simplified psychoacoustic model not only decreases computational the complexity, but also improves the audio quality at a low bit rate.