Sound Field Reproduction System Using Simultaneous Perturbation Method

Kazuya TSUKAMOTO; Yoshinobu KAJIKAWA; Yasuo NOMURA

doi:10.1093/ietfec/e91-a.3.801

Sound Field Reproduction System Using Simultaneous Perturbation Method

Kazuya TSUKAMOTO, Yoshinobu KAJIKAWA, Yasuo NOMURA

Full Text Views

0

Share
Cite this

Summary :

In this paper, we propose a novel sound field reproduction system that uses the simultaneous perturbation (SP) method as well as two fast convergence techniques. Sound field reproduction systems that reproduce any desired signal at listener's ear generally use fixed preprocessing filters that are determined by the transfer functions from loudspeakers to control points in advance. However, control point movement results in severe localization errors. Our solution is a sound field reproduction system, based on the SP method, which uses only an error signal to update the filter coefficients. The SP method can track all control point movements but suffers from slow convergence. Hence, we also propose two methods that offer improved convergence speeds. One is a delay control method that compensates the delay caused by back-and-forth control point movements. The other is a compensation method that offsets the localization error caused by head rotation. Simulations demonstrate that the proposed methods can well track control point movements while offering reasonable convergence speeds.

Publication: IEICE TRANSACTIONS on Fundamentals Vol.E91-A No.3 pp.801-808

Publication Date: 2008/03/01

Publicized

Online ISSN: 1745-1337

DOI: 10.1093/ietfec/e91-a.3.801

Type of Manuscript: PAPER

Category: Digital Signal Processing

Cite this

Copy

Kazuya TSUKAMOTO, Yoshinobu KAJIKAWA, Yasuo NOMURA, "Sound Field Reproduction System Using Simultaneous Perturbation Method" in IEICE TRANSACTIONS on Fundamentals, vol. E91-A, no. 3, pp. 801-808, March 2008, doi: 10.1093/ietfec/e91-a.3.801.
Abstract: In this paper, we propose a novel sound field reproduction system that uses the simultaneous perturbation (SP) method as well as two fast convergence techniques. Sound field reproduction systems that reproduce any desired signal at listener's ear generally use fixed preprocessing filters that are determined by the transfer functions from loudspeakers to control points in advance. However, control point movement results in severe localization errors. Our solution is a sound field reproduction system, based on the SP method, which uses only an error signal to update the filter coefficients. The SP method can track all control point movements but suffers from slow convergence. Hence, we also propose two methods that offer improved convergence speeds. One is a delay control method that compensates the delay caused by back-and-forth control point movements. The other is a compensation method that offsets the localization error caused by head rotation. Simulations demonstrate that the proposed methods can well track control point movements while offering reasonable convergence speeds.
URL: https://globals.ieice.org/en_transactions/fundamentals/10.1093/ietfec/e91-a.3.801/_p

Copy

@ARTICLE{e91-a_3_801,
author={Kazuya TSUKAMOTO, Yoshinobu KAJIKAWA, Yasuo NOMURA, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Sound Field Reproduction System Using Simultaneous Perturbation Method},
year={2008},
volume={E91-A},
number={3},
pages={801-808},
abstract={In this paper, we propose a novel sound field reproduction system that uses the simultaneous perturbation (SP) method as well as two fast convergence techniques. Sound field reproduction systems that reproduce any desired signal at listener's ear generally use fixed preprocessing filters that are determined by the transfer functions from loudspeakers to control points in advance. However, control point movement results in severe localization errors. Our solution is a sound field reproduction system, based on the SP method, which uses only an error signal to update the filter coefficients. The SP method can track all control point movements but suffers from slow convergence. Hence, we also propose two methods that offer improved convergence speeds. One is a delay control method that compensates the delay caused by back-and-forth control point movements. The other is a compensation method that offsets the localization error caused by head rotation. Simulations demonstrate that the proposed methods can well track control point movements while offering reasonable convergence speeds.},
keywords={},
doi={10.1093/ietfec/e91-a.3.801},
ISSN={1745-1337},
month={March},}

Copy

TY - JOUR
TI - Sound Field Reproduction System Using Simultaneous Perturbation Method
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 801
EP - 808
AU - Kazuya TSUKAMOTO
AU - Yoshinobu KAJIKAWA
AU - Yasuo NOMURA
PY - 2008
DO - 10.1093/ietfec/e91-a.3.801
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E91-A
IS - 3
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - March 2008
AB - In this paper, we propose a novel sound field reproduction system that uses the simultaneous perturbation (SP) method as well as two fast convergence techniques. Sound field reproduction systems that reproduce any desired signal at listener's ear generally use fixed preprocessing filters that are determined by the transfer functions from loudspeakers to control points in advance. However, control point movement results in severe localization errors. Our solution is a sound field reproduction system, based on the SP method, which uses only an error signal to update the filter coefficients. The SP method can track all control point movements but suffers from slow convergence. Hence, we also propose two methods that offer improved convergence speeds. One is a delay control method that compensates the delay caused by back-and-forth control point movements. The other is a compensation method that offsets the localization error caused by head rotation. Simulations demonstrate that the proposed methods can well track control point movements while offering reasonable convergence speeds.
ER -