Norwegian University of Science and

Technology
TTT4170 AUDIO TECHNOLOGY
Lab. Exercise 2
Room acoustics
Nicolò Antonante Jens Dankert Dmitrijs Jemeljanovs
October 17, 2010
Abstract
Contents
1 Introduction 1

2 Theory 1
2.1 Sound pressure level . . . . . . . . . . . . . . . . . . . . . . . 1
2.2 Octave bands . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2.3 Room acoustics and reverberation . . . . . . . . . . . . . . . . 2
2.4 Room absorption and absorption coefficients . . . . . . . . . . 2

3 Measurement method and equipment 3

3.1 Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
3.2 WinMLS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.3 Room . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.4 Positions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

4 Results 6
4.1 Average absorption factor . . . . . . . . . . . . . . . . . . . . 7
4.2 Fixed absorption factor . . . . . . . . . . . . . . . . . . . . . . 7
4.3 Decrease of reverberation time . . . . . . . . . . . . . . . . . . 8
4.4 Direct sound level . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.5 Surface materials . . . . . . . . . . . . . . . . . . . . . . . . . 10
4.6 Diffuse sound field . . . . . . . . . . . . . . . . . . . . . . . . 11

5 Conclusions 11
List of Figures
1 Illustration of an impulse response showing direct sound, early
reflections and reverberation . . . . . . . . . . . . . . . . . . . 3
2 Equipment used for measurements . . . . . . . . . . . . . . . . 4
3 Sketch of EL4 auditorium Elektroblokk B, Gløshaugen campus 5
4 Microphone positions at EL4 . . . . . . . . . . . . . . . . . . . 6
5 Relative SPL as a function of distance in the 125 Hz to 4 kHz
octave bands . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
6 Sketch of an absorber - front view and side view . . . . . . . . 10

List of Tables
1 Average absorption factor ᾱ for each octave band . . . . . . . 7
2 Average absorption factor αrs for the remaining surfaces . . . 7
3 Different positions and their distance to the loudspeaker / the
ceiling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4 Value table generated by WinMLS for position 1 . . . . . . . . 13
5 Value table generated by WinMLS for position 2 . . . . . . . . 14
6 Value table generated by WinMLS for position 3 . . . . . . . . 14
7 Value table generated by WinMLS for position a . . . . . . . . 14
8 Value table generated by WinMLS for position b . . . . . . . . 14
9 Value table generated by WinMLS for position c . . . . . . . . 15
10 Value table generated by WinMLS for position d . . . . . . . . 15
11 Value table generated by WinMLS for position e . . . . . . . . 15
12 Value table generated by WinMLS for position f . . . . . . . . 15
13 Value table generated by WinMLS for position g . . . . . . . . 15
14 Value table generated by WinMLS for position h . . . . . . . . 16
15 Value table generated by WinMLS for position i . . . . . . . . 16
1 Introduction
Room acoustics is an area of acoustic that deals with the behavior of sound
in an enclosed space. Reverberation time is an important parameter in room
acoustic analysis. It defines sound persistence in a particular space after the
source is removed. The optimum reverberation time for a room depends on
which use it is designed for.
The reverb can in fact have negative aspects, such as the risk of mask-
ing syllables during a speech, and positive such as the reinforcement of the
sources intensity. In this paper the measurement of the reverberation time
of an auditorium and the measurement of relative sound pressure level as a
function of distance are described. A brief introduction to basic concepts of
room acoustic analysis is also presented.

2 Theory
This section gives a basic overview of the theoretical background of recording
pressure levels, frequency selective analysis and room acoustic analysis.

2.1 Sound pressure level

The sound pressure p(x, t) is the variation in air pressure around the static
air pressure ps so that ptot (x, t) = ps + p(x, t). Sound pressure is measured in
Pascal (Pa). Sound pressure level (SPL) is defined as ten times the logarithm
of the ratio between the RMS square of the sound pressure prms and the square
of a reference value pref :

prms 2
   
prms
Lp = 10 log10 = 20 log10 dB (1)
pref 2 pref
The commonly used reference sound pressure in air is pref = 20 µPa.
SPL is measured in decibel (dB). In this paper, there is no calibration and
only relative sound pressure level is measured, as only the changes in sound
pressure are relevant.

1
2.2 Octave bands
In order to analyse the spectrum of the noise measured, Octave band filtering
can be used. This operation consists of using different filters that each have
an octave’s width and a central frequency at the corresponding octave centre.
An octave is the interval between one musical pitch and another with half or
double its frequency. The central frequencies of the different octaves are:
63 Hz, 125 Hz, 250 Hz, 500 Hz, 1 kHz, 2 kHz, 4 kHz, 8 kHz
In this way the signal can be decomposed and analysed, for example in
order to see in which frequency band the noise has its main components.

2.3 Room acoustics and reverberation

One of the main aims of room acoustics is to study reflection and absorption
of sound to improve the hearing experience. An important aspect to take
into account is reverberation. An impulsive sound generated in a room will
persist for a certain time depending on room size, walls materials and shapes.

Reverberation time measures how long the sound remains audible and there-
fore gives and indication of how it will mix with the following sound. In an
auditorium, for example, a short reverberation time is desirable in order to
avoid the masking of syllables. If the persistance of a syllable is too long,
it can mask the sound of the following one making the speech difficult to
understand.

Reverberation time is generally defined as the time required for reflections

to decay by 60dB below the level of the direct sound. Its usually indicated
as T60 . In order to calculate T60 Sabines formula can be used:
0.161 · V
T60 ≈ , (2)
A
where V is the volume of the room and A is the room absorption.

2.4 Room absorption and absorption coefficients

The absorption properties are strictly dependent on the materials used. The
absorption coefficient of a material is a number between 0 and 1 which in-
dicates the ratio between the absorbed sound and the incident sound. It is

2
Figure 1: Illustration of an impulse response showing direct sound, early
reflections and reverberation

generally indicated with α. According to Sabines formula, the total absorp-

tion ᾱ in a room can be calculated as:

A = 4mV + S·ᾱ, (3)

where is the average absorption coefficient and S is the total surface of
the room. The average absorption coefficient can be described as:
P
Si αi
i
ᾱ = , (4)
Stot
where Si are the different surfaces of walls ceiling and floor, and αi are the
respective absorption coefficients. To modify the absorption of a room, and
therefore the reverberation time, panels with different absorption coefficients
can be used to cover ceiling and walls.

3 Measurement method and equipment

The laboratory’s task is to perform measurements in an auditorium using the
WinMLS measurement system. The recordings are done using microphone,

3
an amplifier, a laptop computer and a loudspeaker. Analysis of the recordings
is performed using the WinMLS software.

3.1 Equipment
For recording, a microphone, an amplifierand a laptop computer are used.
The microphone shown in figure 2(a) is a Norsonic Type 1206 microphone,
which is connected to the amplifer. All measurements are performed with
the microphone facing upwards. Figure 2(b) shows the Norsonic Type 336
amplifer that is used to amplify the microphone’s output. All measurements
except for one are performed with a 30 dB gain in order to achieve a high
signal-to-noise ratio.

(a) Norsonic Type 1206 microphone (b) Norsonic Type 336 amplifier

(c) Genelec Active Monitor Model 1029A

Figure 2: Equipment used for measurements

The loudspeaker shown in figure 2(c) is a Genelec Active Monitor Model

4
1029A and has a built-in amplifier. The laptop computer is used to control
both microphone input and loudspeaker output with the WinMLS measure-
ment software.

3.2 WinMLS
WinMLS measures impulse or frequency responses and transfer functions
using a computer and a sound card. For this exercise, predefined settings are
loaded. The measurement type is swept sine with a duration of 20 seconds per
measurement. In addition, WinMLS is used to calculate the reverberation
time and the relative sound pressure level.

3.3 Room
The room chosen for the measurements is the EL4 auditorium Elektroblokk
B, Gløshaugen campus. It is a small auditorium with seven rows of seven
chairs each. The higher the distance from the rows to the lectern, the more
elevated the rows are, as shown in figure 3.

Figure 3: Sketch of EL4 auditorium Elektroblokk B, Gløshaugen campus

5
3.4 Positions
The measurements are performed at nine different microphone positions
which are shown in figure 4. At each position, one measurement was per-
formed for 20 seconds. All measurement results can be found in appendix
A.

Figure 4: Microphone positions at EL4

4 Results
Preliminary to other calculations, the auditorium’s geometrical properties
have to be calculated. The dimensions are taken from figure 3.

Stot = 218, 867 m2 (5)

Schair = 33, 422 m2 (6)

V = 174, 286 m3 (7)

6
4.1 Average absorption factor
Table 1 shows the average absorption factors in the room for each octave
band. WinMLS calculates only T30 values, as T60 are difficult to obtain in
praxis. Nevertheless, T60 values can be attained by a multiplication with the
factor 2. This results in a slightly altered version of Sabine’s equation:
0, 0805 ms ·V
T30 = (8)
Stot · ᾱ
0, 0805 ms ·V
ᾱ = (9)
Stot · T30

Table 1: Average absorption factor ᾱ for each octave band

Octave Band 125 Hz 250 Hz 500 Hz 1 kHz 2 kHz 5 kHz
Average absorption factor ᾱ 0.078 0.121 0.132 0.134 0.135 0.140

4.2 Fixed absorption factor

Asumming that the chair (sloped) surface has an absorption factor α of
0.7, the average absorption factor for the remaining surfaces is calculated.
The calculations are performed for each octave band according to Sabine’s
formulad. The resulting absorption factors can be found in table 2.
P
Si αi
i
ᾱ = (10)
Stot
ᾱ·Stot − αchair ·Schair
αrs = (11)
Srs

Table 2: Average absorption factor αrs for the remaining surfaces

Octave Band 125 Hz 250 Hz 500 Hz 1 kHz 2 kHz 5 kHz
Average absorption factor αrs -0.034 0.017 0.029 0.031 0.034 0.04

7
4.3 Decrease of reverberation time
In the 1 kHz octave band, the average reveberation time T60 is 0.96 seconds.
This reverberation time is to be lowered to 0.5 seconds using sound absorbers
with different absorption coefficients α from 0.3 to 0.9. It is assumed that the
chair surface still has an α of 0.7 and that sound absorbers are used for the
ceiling only. Everything else is assumed to have a fixed absorption value of
0.031, which is the average absorption factor of the room in the 1 kHz octave
band according to table 1.
0.161 ms ·V
T60 = (12)
Stot · ᾱ
P
Si αi
i
ᾱ = (13)
Stot

0.161 ms ·V 0.161 ms ·V
⇒ T60 = P = (14)
Si αi Schair · αchair + Srest · αrest + Sceiling · αceiling
i

s
 
0.161 m ·V
T60
− Schair · αchair − Srest · αrest
⇔ αceiling = (15)
Sceiling

s
·174.286 m3
 
0.161 m
0.5 s
− 33, 422 m2 · 0.7 − 121.325 m2 · 0.031
αceiling = = 0.452
64.12
(16)
Increasing only the ceiling’s absorption is on of the easiest solutions to in-
crease the average absorption. Of course, there are countless other solutions,
for example covering parts of surface sections or use of multiple, varying
absorber material for a single section.

4.4 Direct sound level

Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy
eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam
voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita
kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit amet.

8
Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy
eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam
voluptua. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam
nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat,
sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum.
Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor
sit amet. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam
nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat,
sed diam voluptua.

Figure 5: Relative SPL as a function of distance in the 125 Hz to 4 kHz octave

bands

At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd
gubergren, no sea takimata sanctus est Lorem ipsum dolor sit amet.

9
 Figure 6: Sketch of an absorber - front view and side view

4.5 Surface materials

In auditorium EL4 there are two main absorbing surfaces - chairs and ab-
sorber on the ceiling. Absorber on the ceiling as shown in figure 6 is perfo-
rated material which is typical resonant absorber. Resonant absorbers are
usually materials with holes and slots or perforated materials, like in the case
of auditorium EL4.
A Resonant absorber is an absorber which is designed to absorb sound in a
narrow frequency range. Resonance frequency can be calculated as follows:
s
c Shole
f0 = (17)
2π d · Sperhole · lef f
π
lef f = l + ·b (18)
2
Shole - cross section area of the hole
Sperhole - area behind each hole
d - distance from the wall
lef f - effective length of a hole
b - radius of a hole

10
 Having in mind above formulae, f0 is calculated:
π π 1.5
lef f = l + ·b=3+ · = 4.1781 · 10−3 (m)
2 2 2
πd2
Shole = = 1.77 · 10−6 (m2 )
4
π · 6.52
Sperhole = = 33.183 · 10−6 (m2 )
4
r
344 1.77 · 10−6
f0 = ≈ 357.15(Hz)
2π 0.3 · 33.183 · 10−6 · 4.1781 · 10−3

4.6 Diffuse sound field

Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy
eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam
voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet
clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit
amet. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam
nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat,
sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum.
Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit
amet.

5 Conclusions
Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy
eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam
voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet
clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit
amet. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam
nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat,
sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum.
Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor
sit amet.Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam
nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat,

11
sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum.
Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor
sit amet. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam
nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat,
sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum.
Stet clita kasd gubergren, no sea takimata sanctus est Lorem ipsum dolor sit
amet.

12
 Position Distance to speaker gain Distance to ceiling

1 7,05 30 1,1
2 4,12 30 1,7
3 1,03 30 1,95

a 7,05 30 1,1
b 6,07 30 1,4
c 5,09 30 1,61
d 4,12 30 1,7
e 3,03 30 1,92
f 2,01 30 1,95
g 1,03 30 1,95
h 0,5 30 1,95
i 0,25 20 1,95

Table 3: Different positions and their distance to the loudspeaker / the ceiling

F[Hz] 63 125 250 500 1000 2000 4000 8000 16000

snr[dB] 15,6 26,6 28,5 40,6 47,8 51,2 57,3 56,1 53
edr[dB] 27,1 38,1 43,5 56,4 63,7 67 71,7 72,9 70,3
EDT[s] 1,23 0,7 0,38 0,27 0,27 0,11 0,08 0,09 0
T30[s] 1,33 0,9 0,52 0,47 0,44 0,43 0,42 0,36 0,29
corr -1 -0,99 -0,99 -1 -1 -1 -1 -1 -1
T20[s] 1,35 0,72 0,44 0,43 0,43 0,45 0,4 0,36 0,31
corr -1 -1 -0,98 -1 -1 -0,99 -1 -1 -0,98

Table 4: Value table generated by WinMLS for position 1

13
F[Hz] 63 125 250 500 1000 2000 4000 8000 16000
snr[dB] 10,6 20,9 25,7 36,2 42,2 44,9 47 50 43,8
edr[dB] 23,1 34,3 42,7 51,3 59,4 60,6 64 65,7 60,2
EDT[s] 1,28 0,84 0,38 0,49 0,43 0,47 0,36 0,32 0,17
T30[s] 1,37 0,77 0,57 0,48 0,49 0,49 0,47 0,39 0,29
corr -0,99 -1 -0,99 -1 -1 -1 -1 -1 -1
T20[s] 1,37 0,78 0,53 0,46 0,47 0,48 0,46 0,39 0,27
corr -0,99 -1 -0,98 -1 -1 -1 -1 -1 -1
r[m] 4,2

Table 5: Value table generated by WinMLS for position 2

F[Hz] 63 125 250 500 1000 2000 4000 8000 16000

snr[dB] 16 21,9 22,3 31 39,3 43,1 47,6 47,9 42,1
edr[dB] 29,6 36,7 37,6 47,2 54,6 59 63,4 64,4 59,3
EDT[s] 1,08 0,79 0,45 0,5 0,34 0,3 0,3 0,21 0,14
T30[s] 1,38 0,79 0,5 0,51 0,51 0,5 0,48 0,4 0,31
corr -1 -1 -1 -1 -1 -1 -1 -1 -1
T20[s] 1,36 0,84 0,51 0,49 0,51 0,49 0,45 0,38 0,29
corr -1 -1 -0,99 -1 -1 -1 -1 -1 -0,99
r[m] 7,2

Table 6: Value table generated by WinMLS for position 3

F[Hz] 63 125 250 500 1000 2000 4000 8000 16000

snr[dB] 16 21,9 22,3 31 39,3 43,1 47,6 47,9 42,1
edr[dB] 29,6 36,7 37,6 47,2 54,6 59 63,4 64,4 59,3
G[dB] 101,6 101,2 99,2 100,3 105,2 106,4 108,1 108,1 103,4
r[m] 7,2

Table 7: Value table generated by WinMLS for position a

F[Hz] 63 125 250 500 1000 2000 4000 8000 16000

snr[dB] 11,6 22,6 24,9 34,4 41,2 43,2 48,5 48,6 40,5
edr[dB] 27,2 33,1 38,2 50,2 58,1 58,8 63,3 65 58,4
G[dB] 97,2 100,3 99,7 103,3 106,7 106,8 108,1 109,3 103,2
r[m] 6,2

Table 8: Value table generated by WinMLS for position b

14
F[Hz] 63 125 250 500 1000 2000 4000 8000 16000
snr[dB] 11,3 21,8 24,9 36,9 42,4 44,4 49,7 48,7 42,8
edr[dB] 23,5 34,3 40 53,3 58,4 60,8 64,7 65,4 59,4
G[dB] 96,1 100,8 100,3 104,9 107,1 107,3 109,3 109,1 104
r[m] 5,2

Table 9: Value table generated by WinMLS for position c

F[Hz] 63 125 250 500 1000 2000 4000 8000 16000

snr[dB] 11 20,1 26,3 36,1 42,9 43,6 47,9 49,7 44,3
] edr[dB] 24,1 34 43 52,1 58,2 60,6 64,1 65,5 61,2
G[dB] 94,7 100,7 102 104 106,9 108,2 110,2 110,1 105,9
r[m] 4,2

Table 10: Value table generated by WinMLS for position d

F[Hz] 63 125 250 500 1000 2000 4000 8000 16000

snr[dB] 13,8 21,1 26,5 37,7 43,4 47,2 51,8 51,8 47,2
edr[dB] 28,6 34,6 41,9 54 60 62,9 66,9 68,3 64,2
G[dB] 99,2 101,6 102,1 105,5 108,5 109,1 112,3 112,6 108,7
r[m] 3,2

Table 11: Value table generated by WinMLS for position e

F[Hz] 63 125 250 500 1000 2000 4000 8000 16000

snr[dB] 15,1 21,5 29 38,9 45,5 45,4 52,3 53,5 48,8
edr[dB] 28,9 34,7 44,9 56 63,3 62,4 63,7 68,2 65,5
G[dB] 99,6 101,6 105,6 107,4 110,8 112,2 114 113,7 110,5
r[m] 2,1

Table 12: Value table generated by WinMLS for position f

F[Hz] 63 125 250 500 1000 2000 4000 8000 16000

snr[dB] 17,1 25,3 29,7 42,5 50,5 54,7 58 57,6 54,4
edr[dB] 27,9 38 44,4 55,4 65,7 69,7 73,2 74 70,8
G[dB] 101,5 105,2 108,5 111,2 115,3 117,2 118,9 119 116,1
r[m] 1,2

Table 13: Value table generated by WinMLS for position g

15
F[Hz] 63 125 250 500 1000 2000 4000 8000 16000
snr[dB] 20,1 32,6 38,8 50,8 55,8 56,4 59,8 63,6 60,5
edr[dB] 33,2 43,6 50,6 62,1 66,3 72,9 74,6 79,5 77,3
G[dB] 105,6 111,2 113,9 116,9 120,6 123,2 124,4 126 123,9
r[m] 0,2

Table 14: Value table generated by WinMLS for position h

F[Hz] 63 125 250 500 1000 2000 4000 8000 16000

snr[dB] 24,3 38 43 54,5 60,7 64,1 63,6 65,4 61,7
edr[dB] 39,7 50,3 60,2 68,1 77,2 75,9 73,5 81,3 78,8
G[dB] 101,7 108,6 111,3 113,8 117,5 119,1 117,6 120,1 118,7
r[m] 0,4

Table 15: Value table generated by WinMLS for position i

16