Acoustical Behavior Of Churches: Mudejar-Gothic Churches

Acoustical Behavior Of Churches: Mudejar-Gothic Churches Juan J. Sendra*, T. Zamarreflot and J. Navarro* *Department of Architectural Constmctionr ‘Department of Applied Physics. Seville Universify. Avenida Reina Mercedes, 2, IiOiZ-Sevilla (Spain) Abstract: Christian churches have traditionally been considered to have very good acoustical behavior. This unfounded belief has led to important en-on in church rehabilitation works, especially when the church is to be used subsequently as an auditorium or theater; this is the case of many churches in Spain. Our research group has worked in acoustical analysis for more than ten years. We have participated in many cases of church rehabilitation and have had the opp&unity to achieve surprisingly good acwstical behavior in several. We deal with a very common type in southern Spain: Gothic-Mudejar churches, which have a small volume, a triple-nave layout, and wooden ceilings. INTRODUCTION The acoustical behavior of Roman Catholic churches has been the object of di@erent studies by several authors, including the authors of this paper (1,2). In reality, the special volumeaic and formal characteristics of these places of worship lend them a particular acoustical behavior which, with fay exceptions, can be rated as inadeqnate. The Audalusian MudejarGothic churches are among these exceptions. When Christians recovered southern Spain from the Arabs in the thirkenth century, they built churches over the mosques for symbolic reasons: they wanted to represent the new power. To do so, they brought masters from the north who knew the mlcs of Gothic architecture. However, a large nun&r of churches had to be built in a short time, and the available workers were mainly Mudejars (subjugated Moors), who were familiar with other constructive techniques, such as wooden roofs and brick walLs. This gave rise. to the Mudejar-Gothic styie, with churches whose most representative points (presbytery and portals) are built of stone, following the Gothic style, while the body of the church has brick walls and three naves with a coffered ceiling (figure 1). This is the style of most of the parish churches in Seville (Spain), and it is the style we will deal with in this study. We have chosen ten churches, a sampling which we consider sufficientJy representative. ACOUSTICAL BEHAVIOR We carried out a very detailed acoustical analysis on each building, including on-site tests to detemune the fohowingz a) Reverberation time at different frequencies on the octave band b) Distribution of sound pressure levels in a stationary state c) RASTl index at scvcral points, with and without electroacoustical support d) NR indexes to evahrate background noise. The sound distribution study allowed us to deduce a model similar to Bamm’s model (3) to describe the reverberated field, which we published recently (4). From the analysis of the reverberation times at difkrent frequencies in the ten churches (figure 1), we can conch& that their acoustical conditions are reasonably good, even when there is no audience. These conditions are close to optimum for music at medium and high fkquencies, ahhough the values arc rather low for bass because of the presence of a large amount of bass-absorbing mater&J (wood). When the church is tGll, reverberation times are near optimum for speech. only three of the ten churches studied differ fmm the above evaluation: the St. Marina, St. Julian, and St. Mark Churches. All three have recently undergone renovation and restoration work and have lost much of their ornamentation. Also, the St. Marina and St. Julian churches are characterized by their large size in relation with the other churches 2127 125 250 500 1000 Frequency (Hz) 2ooll FIGURE I. Inside of the M&jar-Gothic St Isidore church and reverberation times measured in lhe ten churches studied. St. Mark’s church is smaller, but its main nave is presently covered with ceramic pieces because the original wood was damaged hy termites. Finally, it is important to mention that none of these three churches have side chapels. The RASTI index measurements gave average values of between 0.35 and 0.48 without electzoacoustical support, which would classify their intelligibility as “poor: although it must he noted that the tests were run when the churches had no audience. The existence of a public-address system raised the average RASTT values to between 0.45 and 0.57. The background noise levels in the churches produce NX indexes of between 33 and 44 due to traflic noise hecause all of them, except one, are located in the historic center of the city, near narrow, heavily-traveled streets. The presence of a heavy wooden inner door at the main entrance in the St. Gil and St. Vicent Churches is the principal cause of the lower values shown for this index. REFERENCES I . Sendm, J.J. and Navarro, J., La Evolucidn de Las Condiciones Actisticas en las Iglesias, &villa: Instituto Universitario de Ciencias de la Construcci6n, 1997. 2 . Sendm, J. J., ZamarreiIo, T., Navarro, J. and Algaba, J., El Problema ok Ias Condiciones Ackticas en las Iglesias: Principios I propueStus para la Rehabiiitacidn, &villa: Ins&to Universitaxio de Ciencias de la Construcci6n, 1997. 3. Ekmm, M. and Lee, L. J., Journal ofthe Acoustical Society of America 84 (2), 618-628 (1988). 4. Sendra, J.J., Zamarreilo T. and Navarro, J. (Brebbia, C., Kenny, J. and Ciskowski R. D., ed.), Computational Acoustics and its Environmental Applications II. Southampon: Computational Mechanics Publications, 1997. pp. 139148. 2128