HomeHomeUpUpSearchSearchE-mailMail
NEW

Acoustic measurements of several impulse sound sources

Introduction

Several acoustic parameters defined in ISO 3382 seem to be important for room acoustic measurements. According to Bradley [2010] several acoustic measures are closely correlated, so it is important to pick the right ones to make sure one can make clear conclusions. Also the Just Noticeable Difference (JND) for these parameters is important (so one can determine when something is different). The JND is not yet fully determined in many real life situations. The Readily Detectable Improvement (RDI) seems to be a factor 2 to 3 higher than the JND (Bradley [1999]).

In the below table one can see the least correlated parameters with their JND (from literature, Bradley [2010], Ahearn [2009] and Cross&Watson [2006]) and RDI (calculated by web master):
Type of measure
 Measures
 JND
 RDI
Sound strength
 G: strenGth
 1 dB
 3 dB
Clarity measures
 C50: Clarity speech
 1.5 dB
 3 dB
C80: Clarity music
 1.5 dB
 3 dB
Decay times EDT: Early Decay Time 4%
 10%
RT60 (T30, T20): Reverberation Time 4%
 10%
Spacial impression
 IACC: inter Aural Cross Correlation
 0.075
 0.15
LF: Lateral energy Fraction
 0.05
 0.1
GLL: Late Lateral sound level
 ?
 ?

Sources to make a impulse sound

Several sources have been tested for:and re-usability; 'A' weighted sound strenGth; and reproducibility of strenGth (1sigma). On this web page strenGth has not been calibrated!
Source type
 		Reusable
 		G [dB]
 		G (1sigma) [dB]
 		Size
 		Area/Volume
 		Distance
 		Number of
activations
Balloon1
 No
 12
 1.3
 spheric ~10 cm (r)
 V= ~8000 cm3
 2 m
 5
Banger1
 Yes
 5
 3.8
 paper cone 21*15 cm (r*h)
V= ~15000 cm3 1 m
 4
Planks
 Yes
 10
 1.2
 2 planks each 38*5 cm (l*w)
 A= ~200 cm2
 1 m
 4
Rat trap
 Yes
 -4
 0.1
 plastic 12*0.2cm (l*w)
on plank 38*5 cm
 A= ~2.5 cm2 1 m
 4
Hand clap
 Yes
 -8
 3.9
 15*10 cm (l*w)
 A= ~150 cm2 1 m
 4
Banger2
 Yes
 13
 1.2
 paper cone 20*30 cm (r*h) V= ~25000 cm3 2 m
 3

Rat trap sound level at 1m is 90 dB SPL ('A' weighted and measured with RadioShack Sound Level Meter).
The reproducibility of Banger1's and of Hand clap's strenGth is outside the RDI (~3dB).
In the below pictures once can get an idea of the reproducibility (in this case the signal strength [dB] is given).
Signal strength of different sources in a room

Frequency response of the above sources

Each of the above sources have a slightly different frequency response for the strength:
Difference in stenGth of the sources

Some observations:

Room acoustics derived with the above sources

The above sources have been used to determine the acoustical parameters of a room (conservatory 4.75*3.05*2.6m (l*w*h)=~37m3: 3 glass walls, 1 gypsum wallboard, slanted polycarbonate roof and vinyl tiled concrete floor). The 'A' weighted values and their standard variation (1sigma) of the measures can be seen below:
Source type
 C50 [dB]
 C50 (1sigma) [dB]
 C80 [dB]
 C80 (1sigma) [dB]
 EDT [sec]
 EDT (1sigma) [sec]
 T30 [sec]
 T30 (1sigma) [sec]
Balloon1
 9.1
 2.3
 15.9
 2.0
 0.34
 0.05
 0.33
 0.01
Banger1
 12.6
 1.4
 17.8
 1.2
 0.23
 0.02
 0.34
 0.01
Planks
 12.5
 0.4
 20.1
 0.4
 0.23
 0.01
 0.25
 0.01
Rat trap
 13.4
 0.3
 20.8
 0.3
 0.26
 0.01
 0.24
 0.01
Hand clap
 13.2
 0.7
 20.4
 0.7
 0.23
 0.01
 0.24
 0.01
Banger2
 10.3
 0.7
 16.6
 0.4
 0.26
 0.02
 0.30
 0.01

None of the standard deviations of the C50 and C80 are larger than the RDI (~3dB). The Balloon1's EDT has a standard deviation larger than its RDI (10%: ~0.03 sec).
Some observations;

Planned activities

Following are planned:

Software used

Acknowledgments

I would like to thank the following people for their help and constructive feedback: Steve Marshall, Rupert Till, umashankar and all other unmentioned people. Any remaining errors in methodology or results are my responsibility of course!!! If you want to provide constructive feedback, let me know.

Literature

Disclaimer and Copyright
HomeHomeUpUpSearchSearchE-mailMail
Major content related changes: September 12, 2015