Philip Coleman Philip J B Jackson Marek Olik pdcolemansurreyacuk Centre for Vision Speech and Signal Processing University of Surrey Guildford Surrey GU2 7XH UK Jan Abildgaard ID: 489075
Download Presentation The PPT/PDF document "NUMERICAL OPTIMIZATION OF LOUDSPEAKER CO..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
NUMERICAL OPTIMIZATION OF LOUDSPEAKER CONFIGURATION FOR SOUND ZONE REPRODUCTION
Philip Coleman, Philip J. B. Jackson, Marek Olikp.d.coleman@surrey.ac.ukCentre for Vision, Speech and Signal Processing,University of Surrey, Guildford, Surrey, GU2 7XH, UK Jan Abildgaard PedersenBang & Olufsen A/S (now with Dynaudio A/S, Sverigesvej 15, 8660 Skanderborg, DK)
15th July 2014
Paper
#219,
Session
SS06ASlide2
Introduction
Personal sound is an active research topicSlide3
Introduction
Personal sound is an active research topicA number of control strategies proposed [1][1] P. Coleman, P. J. B. Jackson, M. Olik, M. Møller, M. Olsen, and J. Pedersen, “Acoustic contrast, planarity and robustness of sound zone methods using a circular loudspeaker array,” J. Acoust. Soc. Am. 135(4), p.1929-1940, 2014. Slide4
Introduction
Loudspeaker arrays for personal audio:Compact line array [e.g. 2,3][2] J.-H. Chang, C.-H. Lee, J.-Y. Park, and Y.-H. Kim, “A realization of sound focused personal audio system using acoustic contrast control,” J. Acoust. Soc. Am. 125(4), p. 2091–2097, 2009[3] Simón-Gálvez, M. F., Elliott, S. J., & Cheer, J. “The effect of reverberation on personal audio devices.” J. Acoust. Soc. Am. 135(5), 2654-2663, 2014.Slide5
Introduction
Loudspeaker arrays for personal audio:Compact line array Circular array [e.g. 4,5]
[4]
F. Jacobsen, M. Olsen, M.
Møller
, and F.
Agerkvist
,
“A comparison of two strategies for generating sound zones in a room,” in
Proc. 18
th
ICSV,
Rio de Janeiro, Brazil, 10-14 July 2011
.
[5]
M. Shin, S. Q. Lee, F. M.
Fazi
, P. A. Nelson, D. Kim, S. Wang, K. H. Park, and J.
Seo
(2010), “Maximization of acoustic energy difference between two spaces,
”
.
Acoust
. Soc. Am.
128(1)
, p.
121-131, 2010Slide6
Introduction
Loudspeaker arrays for personal audio:Compact line arrayCircular arrayBoth array types may have benefitsUsers may have some freedom to position loudspeakersWe investigate optimal loudspeaker placementSlide7
Introduction
Best positions for N loudspeakers ?Can optimized arrays give…Improved cancellation?Better control of target sound field?Reduced power consumption?Increased robustness?Improved compensation for room?
?
?
?Slide8
Introduction
Previous workCrosstalk cancellation [6,7]Sound zones [8]???
[6] M. R. Bai, C.-W. Tung, and C.-C. Lee, “Optimal design of loudspeaker arrays for robust cross-talk cancellation using the taguchi method and the genetic algorithm,”
J.
Acoust
. Soc. Am. 117(5)
,
p
. 2802–
2813, 2005
[
7
] T
. Takeuchi and P. A.
Nelson,
“Optimal source distribution for binaural synthesis over
loudspeakers”,
J.
Acoust
. Soc. Am.
112(6), p.
2786–
2797, 2002[8] P. Coleman, M. Møller
, M. Olsen, M. Olik, P. J. B. Jackson, and J. Pedersen (Abstract), “Performance of optimized sound field control techniques in simulated and real acoustic environments,” in J. Acoust. Soc. Am., 131(4),
p. 3465, 2012. Presented at Acoustics 2012, Hong Kong, 13-18 May 2012, available viawww.posz.orgSlide9
Approach
Sound zone source weights calculated with acoustic contrast control [9,10]constraint on source weightsdark zone energybright zone energy
[
9
]
J-W. Choi and Y-H Kim, “Generation of an acoustically bright zone with an illuminated region using multiple sources”, J.
Acoust
. Soc. Am. 111, 1695–1700, 2002
.
[10] Elliott
, S. J., Cheer, J., Choi, J. W., & Kim,
Y.
Robustness and regularization of personal audio systems. IEEE
Trans. ASLP,
20(7), 2123-
2133, 2012.Slide10
Evaluation metrics
Generalizable set of metricsEvaluation metricLinked characteristicsContrastMinimal interferencePlanarity [11]Spatial sound distributionControl effortRobustness, low electrical power[11] P. J. B. Jackson, F. Jacobsen, P. Coleman and J. Pedersen, “Sound field planarity characterized by
superdirective beamforming”, in Proc. 21st ICA, Montreal, 2-7 June 2013.Slide11
Evaluation metrics
Evaluation metricLinked characteristicsContrastMinimal interferencePlanaritySpatial sound distributionControl effortRobustness, low electrical powerobserved sound pressures in zone Aobserved sound pressures in zone Bnumber of observation microphones in zone B
number of observation microphones in zone AGeneralizable set of metricsSlide12
Evaluation metrics
energy coincident with the principal plane wave directiontotal energy in the zoneEvaluation metricLinked characteristicsContrastMinimal interferencePlanaritySpatial sound distributionControl effortRobustness, low electrical powerGeneralizable set of metricsSlide13
Evaluation metrics
sum of squared loudspeaker weightsreference loudspeaker weightEvaluation metricLinked characteristicsContrastMinimal interferencePlanaritySpatial sound distributionControl effortRobustness, low electrical power
Generalizable set of metricsSlide14
Objective function
Defined optimization cost function based on physical metricsWhereChoose or optimize weighting coefficientsCould use perceptual model [12][12] J. Francombe, P. Coleman, M. Olik, K. Baykaner, P. J. B. Jackson, R. Mason, M. Dewhirst, S Bech and J. Pedersen, "Perceptually optimized loudspeaker selection for the creation of personal sound zones, in Proc. 52nd AES Int. Conf., Guildford, UK, 2-4 Sept. 2013.Slide15
Approach
Sequential Forward-Backward Search [13]+2, -1Applied each element in turnFocus here on contrast-only caseOther results included in paperSelected between 6 and 30 optimal positions based on predicted performance (mean at 100, 200, ..., 4000 Hz for both zones)[13] P. A. Devijver and J. Kittler (1982), Pattern recognition: A statistical approach. Englewood Cliffs, NJ: Prentice/Hall International., p.220 Slide16
Reproduction setup
60 channel circular candidate arrayTwo 25 × 35 cm zonesIndependent performance measurement setSlide17
Results
Array configurations10 loudspeaker exampleContrast-onlyArcCircleSlide18
Results
Acoustic contrast (average over freq.)Circle worst over frequencyOptimal set best for 6 loudspeakersContrast-onlySlide19
Results
10 loudspeakers over frequencyContrast-only
6 dB
?Slide20
Results
Sound pressure level2650 Hz notch, simulated in free-fieldContrast-only
Dark zone
Bright zoneSlide21
Summary
Loudspeaker array geometries not previously investigated for sound zonesProposed objective function based on physical metricsImproved min. contrast by 6 dB compared to reference arrays (10 loudspeaker example)Further work should investigate:Weighting of cost functionExtended loudspeaker setsAdvanced numerical search methodsSlide22
Stereophonic personal audio reproduction using planarity control optimization
Paper #558Did you see my last talk?
Numerical optimization of loudspeaker configuration for sound zone reproductionSlide23
Acknowledgements
www.linkedin.com/in/philipcolemanaudiop.d.coleman@surrey.ac.ukThanks to Alice Duque who made RIR measurementsSlide24
Results
Optimal 10 channel arrays for other weights:
Contrast-only
Effort-
only
Condition-
only
Planarity-onlySlide25
Results
Optimal 10 channel arrays:
Contrast-only
Effort-
only
Cond-only
Planarity-onlySlide26
Implementation
Measure room responses (60 × 768)Select optimal loudspeakersCalculate optimal source weights for each frequencyInverse FFT/shift to make FIR filters (×60)Independent performance measurement set