ASSESSMENT OF THE ACOUSTICAL PERFORMANCE

IN MALACCAMOSQUE, MALAYSIA: ACASE STUDY ON

THE SAYYIDINA ABU BAKAR MOSQUE
Azma Putra, Dayang H. Kassim, Mohd Y. Yaakob, Y. Abdullah,
Nor S. Muhammad and Ruztamreen Jenal
Faculty of Mechanical Engineering, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya 76100, Durian Tunggal, Melaka, Malaysia,
e-mail: [email protected]
Mohd Jailani M. Nor
Department of Mechanical and Materials Engineering,
Faculty of Engineering & Built Environement
Universiti Kebangsaan Malaysia, 43600 UKM, Bangi Selangor, Malaysia
As a building for worship and one of the important symbols in Islam, a mosque is usually
designed with its architectural grandeur. Among of the characteristics is its spacious inte-
rior to create a perception so that people will feel small when entering the house of God.
Most mosques therefore have high ceiling and are usually constructed with dome shape. In
Malacca, Malaysia however, the architectural of the mosque resembles the traditional Malay
house where the ceiling follows a pyramidal shape. This paper discusses the acoustical con-
dition of the mosque in Malacca. The Sayyidina Abu Bakar Mosque is presented here for a
case study. Measurement of its reverberation time was conducted and the results are validated
with simulation data from CATT room acoustics software. Other important acoustical param-
eters are evaluated such as the early decay time, clarity, denition, lateral fraction and speech
transmission index.
1. Introduction
Mosques are the most important house of worhsip in Islamic religion. Through century, mosque
played an important role from religious to political, from a place to spread religious teachings of Islam
as well as a centre of knowledge development. Mosque can be described as a multifunction public
space where the moslems perform various activities including praying, preaching, and Quran recita-
tion. In contrast to churches which require the harmonic of music, mosque activities therefore can be
seen clearly concentrated only on speech. Unfortunately, the architectural design of a mosque often
neglects the effect on its acoustical performance especially on the speech intelligibility. Considera-
tion of its acoustical quality usually comes after the completion of mosque and is mainly focused on
providing the sound system equipment.
1, 2
ICSV20, Bangkok, Thailand, July 7-11, 2013 1
20th International Congress on Sound and Vibration (ICSV20), Bangkok, Thailand, 7-11 July 2013
Several works have been published which studied of the acoustical parameter of mosques. Ham-
mad
3
performed acoustics measurement in different mosques in Amman, Jordan including the King
Abdullah mosque where RASTI, Early Decay Time (EDT) and Reverberation Time (RT) were as-
sessed. The acoustical performance was found to be poor for most of the mosques, including the
central King Abdullah mosque. The later was then treated by changing the wall materials and carpets
to improve the articulation index.
4
Similar study was also conducted by Abdou
2
for mosques in Saudi Arabia employing the im-
pulse response measurement. Out of 21 mosques under study, 20 mosques shows poor RT below
1 kHz. Abdou
5
then continued with computer simulation to investigate the effect of mosque geome-
try on the acoustical performance.
S u and Yilmazer
6
studied the interior acoustics of Kocatepe Mosque in Ankara, Turkey. They
found that the mosque is a good place for reciting the musical version of the Holy Quran, as it creates
a ritual and tranquil acoustical atmosphere. However, for the prayer mode, the mosque has inadequate
intelligibility of speech.
This paper presents the study of acoustical performance of mosque in Melacca, Malaysia whose
ceiling geometry in particular, is different compared to the other mosques across Malaysia. The
Sayyidina Abu Bakar mosque is chosen for the case study and where there are also complaints from
the worshippers on the difculty of listening sermons. A simulation model is developed using CATT
indoor acoustics software and RT measurement was also conducted.
2. Modeling and Validation
2.1 Architectural design of mosques in Malacca
Most of mosques across the globe usually have a standard architectural design with rectangular-
like geometry and with ceiling and roof having dome shape to create spatial and sacred perception.
This dome is the inuence from the mosques in Turkey dated from the Ottoman Empire. However,
while the parallel walls enclosing the large prayer hall in its interior are almost similar for most
mosques, in some places the design of the ceiling or roof corresponds to the local culture, such as in
Malaysia, especially in the state of Malacca where pyramidal roof is used.
The pyramidal roof, even until now, is commonly found in traditional Malay houses. This type
of pyramidal roof is often used by the modern government and public buildings to preserve the unique
identity of Melaka and so much so, it is also applied in the local mosque. Figure 1 shows examples
of mosques in Melaka applying the pyramidal roof design.
(a) (b)
Figure 1. Mosques in Melaka, Malaysia with pyramidal roof: (a) Al-Alami mosque and (b) Sayyidina Abu
Bakar mosque.
ICSV20, Bangkok, Thailand, July 7-11, 2013
2
20th International Congress on Sound and Vibration (ICSV20), Bangkok, Thailand, 7-11 July 2013
2.2 CATT model of Sayyidina Abu Bakar Mosque
The Sayyidina Abu Bakar mosque in Melaka has a volume of approximately 19,500 m
3
with
6,000 m
2
surface area. It can accommodate 400 jamaah at one particular time. The roof has small
dome as the top part, but supported with sloped wall structure that consists of two tier roofs each
with angle of 27 degree and 30 degree from a horizontal plane. Most of the surface area are made of
concrete material followed by a thick carpet on the oor and large area of glasses on the side walls
and back wall.
These geometry details are then modeled using the CATT indoor acoustic software version 9.
The absorption coefcient of the materials is carefully selected which are available from the software
library. Ten omni directional sound sources are used in the model and are located at the actual loca-
tions, i.e. two on the front wall, two on the side walls, four on the ceiling and two at the back pillars
facing the back wall. The receivers are located at 16 locations across the hall with 1.5 m height; as-
suming the listeners are in standing position. Fig. 2 shows the complete CATT model of the mosque
and examples of sections of the interior model.
Source 1
Source
at front wall
Mimbar
Source
at ceiling
Glass
window
Wooden
door
Womens
prayer hall
Receiver
Mihrab
Source
at side wall
Source
Behind pillar
(a) (b)
Figure 2. (a) 3D view of the CATT model of the Sayyidina Abu Bakar mosque and (b) interior view.
2.3 Measurement of reverberation time
Measurement of RT to validate the simulation model was conducted in the unoccupied prayer
hall using 01-dB Choralis Solo analyzer with dBBati software for the signal processing. The sound
source GDB-S 01dB Metravib was used as the noise generator. Measurements were taken at night to
reduce the effects from background noise.
Fig. 3 shows the comparison of RT from the CATT simulation and from experiment at the 16
measurement point which shows good agreement although discrepancy can be seen at several loca-
tions. High level of RT above 5 seconds can be seen at 500 Hz which causes poor speech intelligibility
in the mosque.
3. Acoustical parametric analysis
Having good agreement of RT from the measurement and simulation, it is therefore of interest
to evaluate other important acoustical parameters in order to have a complete analysis for the sound
quality inside the Sayyidina Abu Bakar mosque. The parameters other than the reverberation time
(RT) are clarity (C80), early decay time (EDT), denition (D50), lateral fraction (LF) and speech
transmission index (STI) . Figure 4 presents the results in terms of distribution maps of the parameters.
The mapping is made at the plane 1.5 m above the oor with 1 1 m surface divisions only at the
main prayer hall and is presented here only for the results at 500 Hz.
ICSV20, Bangkok, Thailand, July 7-11, 2013
3
20th International Congress on Sound and Vibration (ICSV20), Bangkok, Thailand, 7-11 July 2013
0 2 4 6 8 10 12 14 16
0
5
R
T

[
s
]
500 Hz
0 2 4 6 8 10 12 14 16
0
5
R
T

[
s
]
1 kHz
0 2 4 6 8 10 12 14 16
0
5
R
T

[
s
]
2 kHz
0 2 4 6 8 10 12 14 16
0
5
Frequency [Hz]
R
T

[
s
]
4 kHz
0 500 1000 1500 2000 2500 3000 3500 4000 4500
1
1.5
2
2.5
3
3.5
4
4.5
5
5.5
6
Frequency [Hz]
R
T

[
s
]
Average RT
Figure 3. Comparison of the reverberation time (RT) from measurement () and simulation () and
average RT for 16 receiver locations ().
3.1 Reverberation time (RT)
Reverberation time (RT) is the time required for a sound energy to decay by 60 dB after the
sound source is switched off. RT is usually the basic parameter in the assessment of room acoustics.
For speech, RT of less than 1 second in average is preferred at frequency down to 100 Hz. However
this also depends on the size of the room.
7
For the Sayyidina Abu Bakar mosque having volume of
19,500 m
3
, the required reverberation time should be no more than 1.8 seconds.
1
From Fig. 3, high RT above 5 second at 500 Hz has been shown. It is too high for a room
categorized as house of worship emphasizing speech.
7
The RT mapping in Fig. 4(a) also shows
distributions of RT from 5.3 s to 5.6 s at 500 Hz. Domination of high reective materials such glass
windows and concretes usually results in late sound reection. The only absorptive materials in the
mosque is a thick carpet covering the whole surface of the oor. Long sound decay at low frequencies
can reduce the speech intelligibility. Moreover, as the sermon and prayers are led by man, situation
becomes more critical as the male voice obviously contains more low-frequency components than the
female voice.
8
3.2 Early decay time (EDT)
The dened reverberation time as above corresponds to the terminal reverberation time where
we have to wait for sometimes for the sound energy to decay by 60 dB. However, in real situation
where the speech is continuous, we feel only the early portion of the energy, i.e. the rst 10 dB decay.
This is called the early decay time (EDT) which is a better indicator for the subjective feeling of the
reverberation time.
6, 8
In a highly diffuse room, the ratio of EDT/RT is almost equal to 1.0 which indicates diffuseness
and directness of sound energy. For the Sayyidina Abu Bakar mosque as seen in Fig. 4(b), the EDT at
500 Hz from simulation shows range from 5.3 to 5.7 s throughout the hall. The EDT/RT is therefore
around 0.91.0 implying good sound distribution throughout the mosque. This could be due to the
pyramidal ceiling which is discussed in a different section.
3.3 Clarity (C50)
Clarity (also called early-to-late sound index) is the ratio of direct sound energy arriving before
50 ms over reverberant energy arriving after 50 ms. C50 is applied for speech clarity while C80
is more suitable for music.
8
According to Cavanugh et al.,
9
good C50 should have clarity above
-1 dB. However, it is not about the higher C50 the better, but this should also be balanced with the
ICSV20, Bangkok, Thailand, July 7-11, 2013
4
20th International Congress on Sound and Vibration (ICSV20), Bangkok, Thailand, 7-11 July 2013
(a) (b)
(c) (d)
(e) (f)
Figure 4. Distribution maps of acoustical parameters from Sayyidina Abu Bakar mosque: (a)
reverberation time (RT), (b) early decay time (EDT), (c) clarity (C50), (d) denition (D50), (e) lateral fraction
(LF) and (f) speech transmission index (STI).
reverberation time. For 1 kHz, good C50 is considered to be +5 dB with RT = 0.5 s. Poor C50 is
-3 dB with RT = 2 s.
10
This is not much different for 500 Hz.
In Fig. 4(c), good clarity of -1 to 3 dB can be seen at the area behind the back pillars and at the
side walls; all of these are area close to the sound sources therefore the receivers will have dominant
direct sound. However, at the area in front of the mimbar and the mihrab up to the middle of the
hall below the dome, poor clarity of -8 to -5 dB can be observed which is due to the dominant late
reections from the ceiling and the walls. Although at the front walls there are two main loudspeakers,
the dominant late reection energy is masking the direct sound energy. The loudspeakers at the ceiling
are also suspected to create late reections rather than contributing direct eld of sound energy.
3.4 Denition (D50)
The subjective speech intelligibility is often described by Denition or Deutlichkeit or early-to-
total sound energy ratio dened as the ratio of the early received sound energy up to 50 ms after the
arrival of the direct sound to the total received energy. This parameter should be greater than 20% to
ICSV20, Bangkok, Thailand, July 7-11, 2013
5
20th International Congress on Sound and Vibration (ICSV20), Bangkok, Thailand, 7-11 July 2013
satisfy both music and speech performances.
6
Again from Fig. 4(d), at the front wall area in front of the mihrab and mimbar to the middle of
the hall, D50 can be seen to only reach 15% in average, while only small area at the back pillars and
at the side walls has D50 above 40%.
3.5 Lateral Fraction (LF)
One indicator of good acoustic quality is where the audience will feel enveloped or surrounded
by the sound. In a mosque, this is important to add the sacred sense and devotion when listening
the recitation of the Holy Quran and the sermon delivered by the Imam. This requires strong lateral
reections with a signicant fraction of the energy arriving from the side of the listeners. The magni-
tude of this effect is found to be related to the proportion of sound which arrives from the side within
the rst 80 ms after the direct sound.
8
For music, the parameter should be between 0.10.35 and
greater than 0.35 for all other purposes including speech.
From Fig. 4(e), in contrary from the previous results, good LF can be seen around the area in
front of the mihrab straight up to the back wall with LF is between 0.40.5 due to the strong reections
from the side walls. The rest of the area has LF below 0.3.
3.6 Speech Transmission Index (STI)
The nal acoustical parameter evaluated for the mosque is the speech transmission index (STI)
which is often regarded as the direct measure of the speech intelligibility. STI accommodates both
the signal-to-noise and impulse response aspects which affect intelligibility where the calculation
involves modulation of the measured signal.
8
The category division can be seen in Fig. 4(e).
For the Sayyidina Abu Bakar mosque, again at the back pillars and side walls the STI is around
0.4 which is categorized as Fair. The remaining area with STI less than 0.4, the speech index is
considered as Poor and Bad.
4. Ray tracing and echogram analysis
In this analysis, a large number of sound rays are traced from the sound source point up to high
order reections following the optical law of reection. The main result is the distribution of ray
directing to the receiver point. This is analysed in appropriate intervals of the time delay providing a
qualitative presentation of the sound distribution in space and time.
Fig. 5(a) shows the ray tracing simulation to observe the ray distribution to the audience around
the middle of the hall below the dome (receiver no. 6). The diagram only shows the second order ray
which is sound reected two times on the surface before hitting the receiver.
The sound can be seen to reect from surfaces especially from the pyramidal ceiling before
reaching the receiver. However, the second order ray can be seen in the echogram in Fig 5(b) to arrive
mostly at 25 ms. Other higher order rays arrive at above 50 ms, i.e. late reections which is why
the C50 and D50 are poor around this area. This is reasonable proving large volume of the mosque.
Absorptive materials and resonators can be installed to reduce these reected energies.
Roof with pyramidal shape in Melaka mosques may have advantages of distributing the sound
better than the dome shape. Fig. 6 shows the ray tracing simulation between mosque having the
pyramidal and dome roofs. In order to have fair comparison, the roof is made such that the total
volume of each room is almost equal.
The second order rays in Fig. 6(a) can be seen to have multiple reections from the roof before
arriving to the receiver. Meanwhile less rays of the second order with dome roof can be observed to
hit the dome. This indicates that the same rays seen in the room with the pyramidal roof are reected
again to the other surfaces in the room with the dome roof before arriving at the receiver locations
creating high order of reections. The sphere shape of the dome tends to trap the sound longer.
11
ICSV20, Bangkok, Thailand, July 7-11, 2013
6
20th International Congress on Sound and Vibration (ICSV20), Bangkok, Thailand, 7-11 July 2013
(a)
(b)
Figure 5. (a) Ray tracing to receiver no. 6 and (b) its specular echogram.
The echogramin Fig. 6(d) can also be seen to have less component of the second order compared
to that in Fig. 6(b). Especially between 50 and 75 ms, components of the rst and second orders can
be observed in the former and not in the latter. However, these rays come in late reections due to
the large volume of the mosque. In both cases, the sound energy due to these late reections can be
reduced by installing acoustic materials on the room surfaces to increse the acoustic clarity. Further
investigation on the effect of the pyramidal surface angle on the sound distribution is also of interest.
(a) (b)
(c) (d)
Figure 6. (a) Ray tracing to receiver 8 and (b) its specular echogram.
ICSV20, Bangkok, Thailand, July 7-11, 2013
7
20th International Congress on Sound and Vibration (ICSV20), Bangkok, Thailand, 7-11 July 2013
5. Conclusion
Assessment of acoustical performance for Sayyidina Abu Bakar mosque has been conducted.
Simulation results of several important acoustical parameters show that the mosque has poor acous-
tical performance particularly at 500 Hz. The reverberation time at this frequency is more than 5 s.
The acoustic clarity is also poor mainly at the middle area of the hall. Sound distribution (diffuseness
and directness) is however acceptable which could be due to the help of the pyramidal ceiling. With
the large volume of the mosque, acoustical performance can be improved by introducing acoustic
materials and resonators to reduce the reected sound energy especially for the late reections.
6. Acknowledgment
This project is funded by the Short Term Research Grant from Universiti Teknikal Malaysia
Melaka (UTeM) No. PJP/2012/FKM(50C)/S01069.
REFERENCES
1
Orfali, W. Room Acoustic and Modern Electro-Acoustic Sound System Design during Constructing
and Reconstructing Mosques, Ph.D Thesis, Technischen Universitat Berlin (2007).
2
Abdou, A. A. Measurement of acoustical characteristics of mosques in Saudi Arabia, Journal of
the Acoustical Society of America, 113 (3), 15051517, (2003).
3
Hammad, R. N. S. RASTI measurement in mosques in Amman, Jordan, Applied Acoustics, 30 (4),
335345, (1990).
4
Abdelazeez, M. K., Hammad, R. N and Mustafa, A. A. Acoustics of King Abdullah Mosque,
Journal of the Acoustical Society of America, 90 (3), 14411445, (1991).
5
Abdou, A. A. Comparison of the acoustical performance of mosque geometry using computer
model studies. Proceedings of the Eight International IBPSA Conference, Eindhoven, Netherlands,
1114 August, (2003).
6
S u, Z and Yilmazer, S. The acoustical characteristics of the Kocatepe Mosque in Ankara, Turkey,
Architectural Science Review, 51 (1), 2130, (2007).
7
Long, M. Architectural Acoustics, Academic Press, California, (2006).
8
Barron, M. Auditorium Acoustics and Architectural Design, 2nd Ed., Spon Press, New York,
(2009).
9
Cavanaugh, W. J., Tocci, G. C. and Wilkes, J. A., Architectural Acoustics: Principles and Practice,
John Wiley & Sons, New Jersey, (2010).
10
Bradley, J. S., Reich. R. and Norcross, S. G. A just noticeable difference in C50 for speech, Applied
Acoustics, 58 (2), 99108, (1999).
11
Setiyowati, E. Strategies to increase the acoustical quality of the mosques without reinforcement
system, Journal of Islamic Architecture, 1 (1), 2731, (2010).
ICSV20, Bangkok, Thailand, July 7-11, 2013
8