CIPA 2003 XIX th International Symposium, 30 September - 04 October, 2003, Antalya, Turkey 
230 
It is also to be underlined that, even if the qualitative course of 
the measured curve could be estimated, a theoretical model of 
this system is still lacking. 
Moreover, the measured data showed discrepancies with the 
data formerly available and this confirmed the need of such 
measurement campaigns in order to better understand the 
acoustics of mosques. 
5. THE TRANSITION OF ACOUSTICS FROM THE 
BYZANTINE CHURCH TO THE MOSQUE 
With the available data it was possible also to investigate the 
transition of the acoustical characteristics from byzantine 
churches to mosques. This happened for example when St. 
Sergius and St. Bacchus (SB) was converted into a mosque. The 
SB church is quite similar to the Basilica of S.Vitale in 
Ravenna, Italy (SV) since both churches are byzantine-style and 
date of the first half of the VI century. They have a central plan 
and are characterized by a principal volume covered by a dome 
and surrounded by an ambulacre, which is surmounted by a 
balcony. But, while SV is conserved in its original state with 
mostly a sound-reflecting floor, SB is today used as a mosque 
and includes typical sound-absorbing carpets covering the floor 
structure. 
... - * 
—*- SV measured 
* *. 
- -o- • SB' predicted 
'*» 
—*— SB measured 
I 
T"n^. 
f ~ + 
125 250 500 1000 2000 4000 
Hz 
Figure 9. Experimental reverberation time curves for SV and SV and 
thearetical predictions for SB’. 
On the other hand in the higher frequency range, where the 
effect of the carpets is paramount, the insertion of a sound 
absorbing floor completely alters the sound field and this can be 
predicted with good accuracy. 
6. CONCLUDING REMARKS 
Church 
V [m 3 ] s. 
r [m 2 ] S K 
[m 2 ] 
V/Sr 
V/Sp 
SV 
25800 
11400 
980 
226 
2633 
SB 
14900 
6700 
750 
222 
19.87 
SV/SB 
ratio 
1.73 
1.70 
1.31 
1.02 
1.33 
Table 2. Basic geometrical data of the rooms. 
Table 2 reports the main geometrical data regarding the two 
churches: volume (V), total surface (S-p), floor surface (Sp), 
and the ratios V/Sj and V/Sp. Finally, also the ratios between 
respective quantities in the two rooms are included (SV/SB). 
While V and Sy have the same ratios (with SV being the largest 
in absolute extensive terms), the floor surface is proportionally 
more extense in SB. 
The effect of the floor was studied by calculating the theoretical 
curve for reverberation time and comparing it with the 
experimental results. The theoretical curve was obtained with 
the geometrical data of SB after introducing the modification 
according to the ratios in Tab. 1. A room equivalent to SV was 
obtained (called SB') but, differently from SB, the floor in SB' 
was considered sound reflective. The consistency of the 
theoretical predictions with the data collected in SB and SV was 
tested. 
Fig. 9 shows the reverberation time data measured in SB and 
SV and compares them with theoretical predictions for SB'. It 
can be noted that, for RT, the values of the prediction for SB' in 
the lower frequency ranges are overestimated. This evidence is 
in line with former findings (Marsilio, Prodi, Pompoli, 2001) 
where the difficulty in predicting RT correctly for mosques in 
the lower frequency range had been pointed out. This 
inconsistency is not due to the carpets but to other architectural 
details. In particular the articulation of the spaces into many 
niches, half domes and finally the presence of huge domes 
could have an impact on the reverberation time (Prodi, Marsilio 
2003). 
The systematic collection of acoustical data of ancient byzantine 
churches and mosques gave the possibility of a scientific 
qualification and comparison of the acoustics of those two types 
of buildings for worship. The data, taken in the unoccupied 
rooms, show that the values of reverberation are very high, 
providing the room with a unique feeling of majesty. The rooms 
are generally poorly suited for musical performance and the 
delivering of speech messages suffers from the excess of 
reverberation, even though inside mosques the situation seems 
somehow mitigated by the sound absorbing effect of the floor 
surface. The result is that better speech communication can be 
established, at least in the short range. Finally, most of the 
spaces tested are inserted in a noisy city environment. Since no 
specific sound insulation was projected, they all suffer from 
noise intrusion which reduces even more the signal to noise 
ratio. 
7. REFERENCES 
ISO 3382, 1997, Acoustics - Measurement of reverberation 
time with reference to other acosutical parameters. 
Karabiber, Z., 2000, A new approach to an ancient subject: 
CAHRISMA project, Proceedings of the 7 th ICSV Conference. 
Marsilio, M., Prodi, N., Pompoli, R., 2001, On the prediction of 
reverberation time and strength in mosques, Proceedings of the 
Conference ICA2001 
Prodi, N., Marsilio, M., Pompoli, R., 2001, On the effect of 
floor inside mosques, Proceedings of the Conference ICA200I. 
Prodi, N., Marsilio, M., 2003, On the effect of domed ceiling in 
worship spaces: a scale model study of a mosque. Building 
Acoustics, 10(2), in print.