Evaluation  of  a  New  Active  Acoustics  System  in  Music  Performance  of  

String  Quartets  

Doyuen  Ko  and  Wieslaw WoszczykBelmont  University,  Nashville,  USAMcGill  University,  Montreal,  Canada

Objectives

• Evaluate musicians’ overall impression of anew active acoustic system

• Evaluate the effects of an active acousticsystem on musicians’ perception on acoustics

• Investigate the relationship between objectiveroom acoustic parameters and musicians’preference on acoustics

Impulse  response  based  active  acoustics

Real-­time  convolution  engine

M M

Capture Reproduction

16  omni-­‐directional  dodecahedral  speakers  and  4  sub-­‐woofers  (196  drivers  in  total)

Previous  publications• Woszczyk, W., Ko, D., & Leonard, B. (2009).Convolution-­‐Based Virtual Concert Hall Acoustics UsingAural Segmentation and Selection of MultichannelImpulse Responses (pp. 1–8). Presented at the Inter-­‐Noise, Ottawa,Canada.

• Woszczyk, W., Ko, D., & Leonard, B. (2010). AConvolution-­‐Based System for Virtual Acoustic Supportof Performing Musicians. Presented at the 129th AudioEngineering Society Convention, San Francisco, USA.

• Woszczyk, W., Ko, D., & Leonard, B. (2012). VirtualAcoustics at the Service of Music Performance andRecording. Archives of Acoustics, 37(1), 109–113.doi:10.2478/v10168-­‐012-­‐0015-­‐6

Evaluation• 11 professional stringquartets (44 musicians)

• Average 21.3 years of formalclassical music training

• Played a short excerpt oftheir choice in each givenacoustic condition

• Filled out the questionnaireafter playing in eachcondition

3  acoustic  conditions

• Condition  1 – VAT  system  off  • Condition  2 – VAT  system  on  (with  ER,  mid  and  late  reverberation)• Condition  3 – VAT  system  on  with  ER,  mid  and  extended late  reverberation    

Condi&on'1' Condi&on'2' Condi&on'3'

T30'(sec)' 1.38' 1.74' 2.08'

0'

0.5'

1'

1.5'

2'

2.5'

sec$

T30$(Reverb$-me)$

Condi&on'1' Condi&on'2' Condi&on'3'

ST1'(dB)' 013.8' 013.2' 013.1'

014'

013.8'

013.6'

013.4'

013.2'

013'

012.8'

012.6'

dB#

ST1#(Stage#support)#

Condi&on'1' Condi&on'2' Condi&on'3'

Ltotal'(dBSPL)' 88.2' 88.2' 88.3'

87.5'

88'

88.5'

89'

dBSPL&

Ltotal&(Level)&

Condi&on'1' Condi&on'2' Condi&on'3'

IACC'(full)' 0.456' 0.43' 0.414'

0.39'

0.4'

0.41'

0.42'

0.43'

0.44'

0.45'

0.46'

correla'

on)

IACC)(Spa'al)impression))

Objective  measurements  (ISO3382-­‐1)

Ease of hearing selvesEase of hearing others

Ease of maintaining tempoEase of hearing dynamics

Amount of reverberationQuality of reverberation

EnvelopmentHeight sensation

Tonal balanceClarity

Preference

11  questions  for  musicians

Preference

Factor  analysis

Quality   of  reverberation

Multiple  Regression  Analysis  (MRA)Dependent  variable:  preference

Quality  of  reverberationAmount  of  reverberationHearing  other  musicians

(R2 =  0.49)

Quality  of  reverberationHearing  other  musicians

Height  sensation(R2 =  0.542)

Conclusion• The results show a strong preference for enhancedacoustics over inherent acoustics of the space.

• Factor analysis revealed musicians are able to reliablydistinguish and describe the acoustic differencesbetween three acoustic conditions. Three primaryunderlying perceptual dimensions are: Stagesupport, spatial impression and tonal balance.

• ‘Quality of reverberation,’ ‘Amount of reverberation,’‘Hearing other musicians’ and ‘Height sensation’were four most salient attributes explainingmusicians’ preferences on acoustic conditions.