Acoustical experimental analysis on worship buildings aimed to their use as auditorium

Acoustical experimental analysis on worship buildings aimed to their use as auditorium Anna Magrini, Paola Ricciardi Dipartimento di Ingegneria Idraulica e Ambientale, Università di Pavia, Via Ferrata 1, 27100 Pavia, Italy, magrini@unipv.it, paola.ricciardi@unipv.it The acoustic characterisation of buildings for music performances involves not only theatres, but also worship buildings, that with adequate acoustic quality, could held musical performances although temporarily. In order to complete previous analyses, focused on Churches with rectangular plans and two aisles, also central plan Churches were investigated. Considerations on experimental values of the main acoustic parameters obtained in a considerable number of Churches are here presented. Measurement results were compared with acoustic data available in literature regarding theatres and auditoriums. The examined Churches are classified in agreement with the optimal values of the acoustical parameters for the listening of music. Moreover the spatial distribution of the acoustical parameters, as a function of the source receiver distance, is evaluated. It is then correlated to the presence of architectural elements, such as lateral chapels, dome, vaults and columns. 1 Introduction The realisation of this work aims at the valorisation of the wide historical and artistic Italian heritage through the acoustic characterisation of buildings for music performances. Most part of the experimental analyses was focused on worship buildings that could be used as auditoriums. The authors carried out experimental investigations on Churches having different plan configurations: rectangular plan with two aisles [1, 2, 3] and circular plan [4]. Most of the considered worship buildings present articulated environments, such as lateral chapels, coupled rooms, diffusive elements or obstacles to the sound propagation, orientation of surfaces, convex surfaces than can create sound concentration. The internal geometric complexities certainly influence the acoustic field [5] even though do not lead to simple correlations among acoustical parameters and spatial geometric characteristics. With no doubts acoustical parameters such as Clarity and Definition Indexes are quite sensitive to early reflection components of sound and thus highly dependent on the building shape. Nevertheless this consideration could be extended to the Reverberation Time: recent studies on Churches with a central plan and dome [4] put in evidence the fact that, in some cases, reverberation time below the dome assumes lower values than in the other measurement points. Therefore comes the need of a deeper analysis on the way every single geometrical - architectural element, such as the presence and dimension of columns, arches, vaults and surfaces with elements in relief as altar or pulpit etc. could influence the sound paths and, consequently, the spatial distribution of the acoustical parameters. In literature, a great number of Church acoustical data is available, but the geometrical characteristics are often not indicated; therefore a wider examination of the correlation among various experimental parameter and geometric characteristics is not possible. In addition, this research is aimed to find out significant points where measurements should be carried out, on the base of the internal geometry and the peculiarity of the environment, in order to reduce the high number of measurement points and, therefore, the large amount of data to process. 2 State of art of the research In the research on the acoustic characterisation of worship buildings, a great amount of buildings was already subject of experimental investigations. In particular, formerly in the 50s Raes and Sacerdote [6] carried out acoustical measurements on Basilicas of S. Giovanni in Laterano and S. Paolo Fuori Le Mura in Rome. The same Roman Basilicas were subject of study by R.S. Shankland and H.K. Shankland [7], who in the 70s analysed also S. Peter s Church in Rome. In the same years Fearn [8] examined reverberation time in Spanish, English and French Churches. Successively Giulianini and Cocchi [9] experimentally studied 7 Italian Churches (St. Andrea in Mantova, St. Petronio in Bologna, St. Maria Maggiore, St. Agostino in the region of Marche, St. Elena, N.S. Moretta and St. Paolo in Alba, Piemonte). A few years later, another experimental analysis was developed by Recuero, Gil and González [10] in the 2387

monastery of Santo Domingo of Silos in Burgos, in Spain. In 1994 Lannie and Soukchov [11] published experimental results taken in 5 Russian temples with tent roof. In 1998 Hamed and Hassan presented reverberation time values of the State Mosque of Kuwait [12]. In the same period, another Italian Church, the Gesù Church of Ancona, was analysed by Iannotti e Mattei [13], while Galindo and others [14, 15] carried out measurements in 8 Churches of the Gothic - Mudejar period, in Spain. Carvalho [16] analysed 41 Portuguese Churches and, in collaboration with Desarnaulds, Swiss Churches [17]. In the late years, worship buildings were widely studied both in the north and in the south of Italy: Martellotta [18] analysed Romanic Churches in Puglia, Magrini and Ricciardi carried out experimental campaigns in the city of Genoa on 10 Churches [1, 2], with rectangular plans and two aisles, on 14 Churches [4], with central plans. Moreover Magrini A. and Magrini U. [3, 19] investigated 6 Medieval Abbeys of the Cistercian order, 3 in the North of Italy and 3 in the South of France. In the first measurement campaign, ten Churches [1, 2] of the city of Genoa were selected, with rectangular plans and two aisles with and without lateral chapels and/or cupolas, with volume variable from 2183 to 43540 m 3. The typical Church plan and section considered in this measurement session are shown in Figure 1. The second experimental analysis was developed in 14 Genovese Churches [4] with central plan, with volume variable from 1409 to 43756 m 3. Plan and section of one of them are reported in Figure 2. 3 Experimental analysis Subject of the investigation were numerous Churches, with various geometrical configurations, various historical periods and, consequently, various typologies of furnishing. Figure 2 - Plan and Section of St. Zita Church Most recent measurements were developed in Medieval Cistercian Abbeys with volume variable from 3390 to 14970 m 3 [17]. Plan and section of one of them are shown in Figure 3 and 4. The method, common to all the measurements, was based on the backward integration of Schroeder [20]. Churches were unoccupied during measurement. The source was positioned near the altar and the microphone positions were chosen as function of the various configurations of the plan, variable from a minimum number of 6 to a maximum number of 28 for each Church. Figure 1 - Plan and Section of St. Donato Church 2388

cupolas and it was built in Baroque style (started in 1592 by Marcello Pallavicino and completed in 1637). Figure 3 - Plan of Morimondo Abbey and Cloister Figure 5 - Reverberation Time RT at 500 Hz as function of the volume V of the Churches and the optimum trend of RT for various types of music Moreover, other 3 rectangular plan Churches could be employed for the listening of classical or organ music: the cathedral of Genoa, St. Lorenzo, St. Vallisa of Bari and St. Nicola of Bari. Some other Churches, mainly with rectangular plans and two aisles could be used for organ music, such as the small St. Cosma and Damiano, the Romanic St. Donato and St. Maria di Castello. It can be noticed that almost all the central plan worship buildings are not appropriate to be used as auditorium, unless some acoustic corrections are realised. Figure 4 - Section of Morimondo Abbey 4 Discussion In Figure 5 experimental values of the Reverberation Time at 500 Hz are plotted as function of the Church volumes. Data referred to the Church investigated by the authors [1, 2, 3, 4] and by other scientists [14, 15, 18] are reported. In order to classify their potential musical fruition, RT values are compared with the optimal parameters for the listening of music and specifically for organ music, classical music and modern (pop) music. The Church that better agrees with the listening of modern music is St. Agostino [2], which was, in effect, quite recently acoustically restored. Only one central plan Church seems to be suitable either for modern (pop) or classical music: Gesù Church. It is also the most articulated, since it has 7 Figure 6- Mean Reverberation Time as function of the V/S ratio for various worship buildings 2389

In Figure 6 experimental values of the Reverberation Time (125-4 khz average values) are plotted as function of V/S ratio [m] (volume of the Church/pavement surface) of the Churches subject of investigation [1, 2, 3, 4] and of other ones analysed by Italian and Spanish scientists [14, 15, 18]. For a deeper comparison, additional data referred to theatres are presented on the graphic [21]. Considering only the values referred to worship buildings, a regression line was traced, defined by the equation y = 0.25 x, with R 2 = 0.53. It identifies two zones with distinguished geometrical characteristics, one above and one below the regression line. The RT values of Cistercian Abbeys [3] and of Spanish Churches [14, 15] are distributed above the line, as well as the values detected in almost all the Churches with rectangular plan and two aisles [1, 2]. RT values referred to an analysis conducted in Southern Italy [18] are all located below the regression line. Values of RT correlated to central plan Churches are more dispersed [4]. Moreover the unique Churches with rectangular plan and two aisles [1, 2] that are below the regression line are without cupola (St. Agostino, St. Donato e St. Siro). Among the ones with central plan, more dispersed values from the regression line are the ones related to Churches having higher volume of cupolas (St. Zita, and Gesù Church), respectively 7657 m 3 e 7310 m 3 [4]. Figure 7 shows experimental values of Clarity Index C80 as function of the volume of the Church pavement surface ratio (V/S), of the Churches subject of investigation [1, 2, 3, 4], and of other ones [14, 15, 18]. The regression line has the equation y = - 0.2 x - 2.33 y and R 2 = 0.5. For the Clarity Index there is not a clear subdivision among the values lying below and above the regression line. The more dispersed data from the regression line are the ones related to Churches with rectangular plans and two aisles and precisely, N.S. Consolazione, St. Annunziata and St. M. di Castello, having volumes respectively of 18842, 26603 and 21100 m 3. In addition an analysis on the spatial distribution of the acoustical parameters was carried out. The RT average values (125 Hz 4 khz), as function of the S-R (source-receiver) distance were examined. The S-R distance was referred to the central nave, for longitudinal plans [1, 2, 3], and to the direction linking source (altar) and centre of the plan, up to the end of the Church, for central plan Churches [4]. In details, Figure 8 shows that for values lower than 3.5 s, the RT remains quite constant towards the back. For values higher than 3.5 s the variation of RT is more evident. More precisely, the highest variation of 1.46 s is reached in St. Maria Assunta in Carignano, which is the central plan Church with the biggest volume (43756 m 3 ). Figure 7- Clarity Index as function of V/S of various worship buildings Figure 8 Mean Reverberation Time as function of the source receiver distance 2390

In some of the analysed Churches there is a difference of behaviour as a function of the major and minor distance to the altar: in Chiaravalle for a distance minor than 27-30 meters the RT varies of 0.6 s. Over this limit, this effect seems to be less evident. In Morimondo the same trend is shown in the first 15 m [22]. than 35 m, C80 values tend to be more constant, apart from Morimondo Abbey [22]. This behaviour confirms previous analyses regarding C80 distribution, specifically for longitudinal plan Churches [1, 2]. The plot of the Definition Index versus the sourcereceiver distance is shown in Figure 10. The same decreasing tendency of C80 is here only for sourcereceiver distance shorter than 20 m. Beyond this limit, D50 values are more constant. 5 Conclusions Figure 9 Clarity Index as function of the source receiver distance Experimental measurements taken in 10 Churches with two aisles, 14 with central plans and 3 Medieval Abbeys, were examined, aiming to classify them in accordance with the optimal values of the acoustical parameters for the listening of music. The results put into evidence that worship buildings with more articulated volumes, for example Gesù Church, with 7 cupolas, are more adequate for music performances. Almost all the central plan Churches are not appropriate to be used as auditorium, unless some acoustic corrections are realised. The most evident characteristic of the spatial distribution of the acoustical parameters, as a function of the source receiver distance, was associated with the presence and the volume of cupolas. It was observed that reverberation time as function of the V/S ratio follows a regression line common to Churches available from literature. The RT distribution for Churches with rectangular plan and two aisles without cupolas was significantly different from the ones with central plans, and, among these last ones, bigger volume Churches presented more scattered RT values. The Clarity Index distribution showed decreasing values when the source receiver distance was increasing up to 35 m. The same effect was evident also for the Definition Index, but only for sourcereceiver distance lower than 20 m. References Figure 10 Definition Index as function of the source receiver distance In Figure 9 the Clarity Index C80 is plotted versus the S-R (source-receiver) distance. It can be noticed that there is a decreasing tendency of the Clarity Index C80 with increasing values of the source receiver distance. When the source-receiver distance is more [1] A. Magrini, P. Ricciardi, An Experimental Study of Acoustical Parameters in Churches, International J. of Acoustics and Vibration, 7, N. 3, (2002). [2] A. Magrini, P. Ricciardi, "Churches as auditorium: analysis of acoustical parameters for a better knowledge of sound quality", J. Build. Acoust, 10, No. 2 (2003). 2391

[3] A. Magrini, U. Magrini, Measurement of acoustic properties in medieval Abbeys, Proceedings of 18th ICSV, St. Petersburg, Russia, 5-8 July (2004). [4] A. Magrini, P. Ricciardi, Experimental analysis on acoustical parameters in central plan Christian Churches, Proceedings of 18th ICSV, St. Petersburg, Russia, 5-8 July (2004). [5] A. Magrini, L. Magnani, Models of the influence of coupled spaces in Christian Churches, Building Acoustics Vol.12 N.2 (2005). [6] C. Raes, G. G. Sacerdote, Measurements of the Acoustical Properties of two Roman Basilicas", J. Acoustic. Soc. Am., 25, 954-961 (1953). [7] R. S Shankland., H. K. Shankland, Acoustics of St. Peter's and Patriarchal Basilicas in Rome, J. Acoustic. Soc. Am., 50, 389-396 (1971). [8] R. W. Fearn, Reverberation in Spanish, English and French Churches, J. Sound Vibrat., 43, 562-567, (1975). [9] A. Giulianini, A. Cocchi, A contribution to the acknowledge of the acoustic characteristics in close environments: Churches (in Italian), Rivista Italiana di Acustica, Vol. IX, N.1, pp. 3-28, (1985). [10] L. Recuero, M. Gil, C. González, "Experimental study of the acoustics in the Church of the monastery of S.to Domingo de Silos", Acustica, 62, (1987). [11] M. Y Lannie, V. N Soukchov, Acoustics of the russian tent-shaped temples, AES Preprint 3854, Amsterdam, (1994). [12] A. Hamed, M. H Hassan., Assessment of Speech Intelligibility in Large Auditoria - Case Study: Kuwait State Mosque, Applied Acoustics, Vol. 54, N. 4, pp. 273-289, (1998). [13] A. Iannotti, E. Mattei, Analisi delle caratteristiche acustiche ed ipotesi di bonifica della Chiesa del Gesù ad Ancona da adibire ad auditorium, Proceedings of 27 Nat. Congr. of Acoustics, 26-28 May, 60-63, (1999). [14] M. Galindo et al. Clarity and Definition in Mudejar-Gothic Churches, J.Build. Acoust., 6, 1-16, (1999). [15] M. Galindo et al. Speech intelligibility in Mudejar-Gothic Churches, Acustica, 86, 381-384, (2000). [16] A. P. O. Carvalho, Acoustical measures in Churches: Porto s Clerigos Church, a comprehensive example, Proceedings of 7th Int. Congr. Sound and Vibr., Garmisch- Partenkirchen, Germany 1644-1652, (2000). [17] V. Desarnaulds, A. P. O. Carvalho and G. Monay, Church Acoustics and the Influence of Occupancy, Building Acoustics, 9, N. 1, 29 47 (2002). [18] F. Martellotta, Caratteristiche acustiche delle chiese romane in Puglia (Acoustic features of Roman Churches in Puglia, Italy), PhD Thesis, University of Ancona and Politecnico of Bari, (2001). [19] C. Dufour Bozzo, U. Magrini Acoustic properties of a few Cistercian Abbeys. Historical aspects and measurements, Proceedings of Int. Congress on Acoustics Rome, Sept. 2-7 (2001). [20] M. R. Schroeder, New Method of Measuring Reverberation Time, J. Acoust. Soc. Am., 37,n.3, 409 (1965). [21] L. L Beranek., Music, Acoustics and Architecture. (Robert E. Krieger Publishing Company, Huntington, New York, 1979). [22] A. Magrini, U. Magrini Acoustic field in two Medieval Abbeys: relationships between acoustical parameters and architecture in Morimondo and Chiaravalle Abbeys, Proceedings of Forum Acusticum, 29 August 2 September, Budapest, Hungary (2005). 2392