1993 Fiscal Year Final Research Report Summary
Basic Research on Effective Coding Scheme of Wide band Audio Signal based on the Auditory Characteristics
| Project/Area Number |
04452185
|
|---|
| Research Category |
Grant-in-Aid for General Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
物理学一般
|
|---|
| Research Institution | Kumamoto University |
|---|
Principal Investigator |
EBATA Masanao Kumamoto University, Engineering, Professor, 工学部, 教授 (40005319)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
UEDA Yuichi Kumamoto University, Engineering, Associate Professor, 工学部, 助教授 (00141961)
USAGAWA Tsuyoshi Kumamoto University, Engineering, Associate Professor, 工学部, 助教授 (30160229)
SONODA Yorinobu Kumamoto University, Engineering, Professor, 工学部, 教授 (70037836)
MIYAZONO Hiromitsu Yatsushiro National Technical College, Research Assistant, 助手 (00229858)
WATANABE Akira Kumamoto University, Engineering, Professor, 工学部, 教授 (50040382)
|
|---|
| Project Period (FY) |
1992 – 1993
|
| Keywords | Compression of Wide Band Signal / Low Bit Rate Coding / Masking Characteristics / Mathematical Model of Auditory System / Pychoacoustical Measurement / Model of Basilar Membrane |
|---|
| Research Abstract |
In this research, a measurement of simultaneous masking for short duration tone is carried out in order to improve the compression ratio on subband coding technique using masking effect. And a subband coding system based on the result is composed. The composed coding system shows good coding quality. However, to improve the coding scheme more, the collections of psychoacoustical data on masking phenomenon are very hard to realize. And as a new approach for subband coding, we propose a subband coding system based on cochlea model, which basically does not required any masking data. This coding system also shows good quality of coding. As a result of this research project, we can conclude followings. As a result of phychoacoustical experiments, masking data which are directly applicable to subband coding are obtained. The data show the threshold of detection raises when the duration of sound becomes larger than certain length. This phenomenon can be understood as the difference of spectral structures of shorter and longer duration pure tone. And two kind of coding systems are composed. The first one is based on the measured masking data with heuristic tuning, and the second one is based on the cochlea model. The second system does not require any masking data but it can code at 96 kbit/s with high quality sound.
|
Research Products
(12 results)
