Sound field reproduction for reduction of pre-echo artifacts based on time-domain weighted least-squares method프리에코 감소를 위한 시간 영역의 가중 최소 제곱법을 적용한 음장 재현 기법

Cited 0 time in webofscience Cited 0 time in scopus
  • Hit : 758
  • Download : 0
One of the major problems of sound field reproduction techniques is that by approximating continuous array of loudspeakers to discretely spaced loudspeakers spatial aliasing of driving functions occurs. When spatial aliasing conditions are not satisfied, sound field synthesized to reproduce that of virtual source produces aliasing artifacts in front of the desired wave front, called “pre-echoes”. Pre-echoes become problematic because due to precedence effect, if listener hears pre-echo artifacts first, perception error of virtual sources, such as localization error, and coloration effects occur. A new sound field reproduction method using least-squares method, was suggested by Noh in 2011. The goal of this method is to reproduce target sound field induced by virtual source, using discrete loudspeaker arrays by minimizing sound field error in least-squares sense. Least-squares method is advanta-geous is since this method is a time domain approach, sound field can be weighted spatially and temporally, which can be applied to reduce pre-echo artifacts To verify the suggested method, time domain analysis was performed through numerical simulations. However, due to some assumptions made to derive the cost func-tion and due to the nature of least-squares method, calculated optimal control filter showed high frequency components, and large input, which was impractical in real life. To reduce large input and eliminate high fre-quency components, such constraints was designed and applied to the cost function. These modifications showed that control filter calculated with modified cost function resolved existing problems. Now that means of sound field reproduction technique is modified, temporal and spatial error weighting function was designed to reduce pre-echo artifacts. The error weighting function took form as re-verse of step function, weighing much more to space and time of pre-echo artifact, and less to space and time of target sound field. The cost function was...
Advisors
Park, Young-Jinresearcher박영진Park, Youn-Sik박윤식
Description
한국과학기술원 : 기계공학전공,
Publisher
한국과학기술원
Issue Date
2013
Identifier
514893/325007  / 020113465
Language
eng
Description

학위논문(석사) - 한국과학기술원 : 기계공학전공, 2013.2, [ ii, 106 p. ]

Keywords

pre-echo; binaural sound; sound field reproduction; wave field synthesis; 음장재현; 프리에코; 에일리어싱; 입체음향; 가중 최소 제곱법; least-squares method

URI
http://hdl.handle.net/10203/181751
Link
http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=514893&flag=dissertation
Appears in Collection
ME-Theses_Master(석사논문)
Files in This Item
There are no files associated with this item.

qr_code

  • mendeley

    citeulike


rss_1.0 rss_2.0 atom_1.0