RICERCA ITALIANA     

Acoustics of worship spaces

Politecnico di Bari

Abstract

The present research program aims at contributing to the study of worship buildings acoustics which, according to the proposers, is not studied enough both on a national and international level. In fact, there is a wide scientific literature about theatres and concert halls, but the same is not true for churches and other worship spaces. In addition, such buildings are much more complex both because of the complicated volumes (made of transepts, naves, domes and so on), and because of the different acoustic characteristics required by the most frequent kind of sound messages (speech and music) that propagate inside. The proposed research follows the work began by two of the three sub-units during the PRIN 2003 and is focussed on several aspects of church acoustics: the definition of a specific measurement protocol; the propagation of sound in such places; the characterization of the materials; the relationship between geometric and acoustic parameters; the study of the optimal listening conditions for both speech and music; the validation of simulation programs in complex spaces.
The definition of a specific measurement protocol and the study of sound propagation in churches will be developed in strict cooperation between the proposing research units. In particular, the study sound propagation will be investigated in two ways: the research units of the Politecnico di Bari and of the University of Pavia will study the problem using, respectively, scale models and virtual acoustic models. The reserach unit of the IUAV of Venice will support the other units by measuring absorption and scattering coefficients of the elements found inside the churches. Since each research unit has a different acoustical simulation program, they will carry out jointly a round-robin test in order to validate the different softwares. Finally the other arguments will be investigated individually by each research unit, collaborating with the others when necessary.
The analysis of the statistical correlations between acoustical parameters, architectural parameters, and subjective judgments will, hopefully, provide useful information about the relationships between acoustics and architecture in churches. In addition, this research will be a useful instrument of knowledge of the past and of the present, but will also provide instruments to better "design" the future worship buildings. <<<

Principal Investigator

Ettore CIRILLO Politecnico di BARI

Research Objectives

The modern science of architectural acoustics provides powerful instruments to investigate the delicate relations between a musical composition, its interpretation, and the space in which it is played.
The best orchestra directors often use the acoustics of the places in which they perform as well as musicians use their instruments. So it can be reasonably supposed that also the best composers created their masterpieces taking into account the acoustics of the places in which the music had to be played. The relation between music and architecture is likely to have been clearer to musicians than to architects. In fact, the history of architecture shows that architects were, in most of the cases, unaware of the acoustical requirements of the buildings they created in order to achieve an optimal balance of form and function. However, Gregorian chants sound better in the large, resonant cathedrals in which they were originally sung. In fact, the melody of the Gregorian chants takes advantage of the long reverberation of the Romanesque and Gothic churches (mostly made of stone), to produce an harmonic effect given by the superposition of the slowly decaying sounds. Similarly, Willaert and Gabrieli took advantage of the architecture of St. Mark's Basilica in Venice by dividing the choir into two groups located at opposite positions in order to produce (together with the orchestra and the church echoes) a rich, colourful, and rhythmic polyphony. Similarly, J.S. Bach composed several works (such as St. Matthews Passion), having in mind the dry acoustics of the Tomaskirche in Leipzig.
All these examples suggest that music and architecture are certainly related and, consequently, it is likely that they also evolved together. Such relations need to be investigated because they play a role in the development of human knowledge in the musical field. In fact, since its beginnings the Catholic Church continued the Greek musical tradition and encouraged the development of the music which, consequently, consisted mostly of sacred works until the Renaissance period. During the Baroque period (that lasted from the XVII to the half of XVIII century) the Catholic Church began to loose its supremacy in the musical field. In that period most of the composed preferred secular music, in particular the new operatic form and instrumental works which favoured the spreading of opera houses and concert halls. At the end of the 17th century the best musicians preferred an independent and well-paid work rather than a fixed income work in a chapel. This fact determined the end of the aristocratic patronage and the diffusion of new management mostly held by middle-class entrepreneurs who encouraged public performances for payment.
The study of the acoustics of opera houses and concert halls has always raised a great interest (from the oldest examples to the contemporary halls), the same is not true for the churches which until the Baroque age were the only places where music (which was mostly sacred) was played and listened.
The analysis of the national and international literature about this issue, reported in the next section, shows that several studies about churches have been published till now, but in many cases they lack an adequate level of detail, which on the other hand, can be observed for theatres and concert halls. Only recently the proposers of this research have begun a systematic study of the acoustics of Italian churches, pointing out the need to further investigate, among the others, some aspects in particular. These aspects are described below and are the main objectives of the proposed research.
First of all, a specific protocol for acoustic measurements will be defined. The protocol will be based on current international standards (like ISO 3382:97 - Acoustics -- Measurement of the reverberation time of rooms with reference to other acoustical parameters), but will provide more detailed specifications in order to normalize measurements in such complex spaces.
Then the effect of the complex spatial layout of the churches on sound propagation will be investigated. In particular, the effect of domes, naves, transepts, and so on, will be studied by using both scale models and carefully validated virtual acoustic models.
The surfaces and the elements found in the churches will be characterized in terms of absorption and scattering coefficients because, as observed in the literature, both these aspects have been scarcely investigated.
A listening test will be carried out in order to investigate subjective preferences about the combined listening of speech and music as a function of different combinations of PA systems and natural acoustics.
The analysis of the statistical correlations between acoustical parameters, architectural parameters, and subjective judgments will, hopefully, provide useful information about the relationships between acoustics and architecture in churches. In addition, this research can be a useful instrument of knowledge of the past and of the present, but also a positive instrument to design the future. <<<

Timescale

24 months

National and international background

The study of the acoustic characteristics of churches begun in the fifties, when Raes and Sacerdote [1] analysed the Roman Basilcas of St. John in the Lateran and St. Paul outside Walls. In the next years few papers were written about other Roman churches [2], about a group of Spanish, English and French churches [3], about Byzantine churches in Thessalonika [4,5], and about St. Paul Cathedral in London [6]. Later on, Giulianini and Cocchi [7] studied the churches of S. Andrea in Mantova and St. Petronius in Bologna. Another experimental survey was carried out in the monastery of Santo Domingo de Silos in Burgos (Spain) by Recuero, Gil and Gonzales [8]. In 1994 Lannie and Soukchov [9] published experimental results of five Russian tent-shaped temples.
All these works were essentially based on reverberation time measurements. Only recently other acoustic aspects have been taken into account. Among them, the speech intelligibility was measured in a large group of Swiss churches [10,11], and in an even larger group of Portuguese churches [12-20]. The latter research introduced the measurement of several acoustic parameters, and in particular, those used to characterize concert halls acoustics. In addition subjective listening tests were performed by a group of selected listeners. The most important results of Carvalho's research refer to the existence of correlations between acoustic and architectural parameters [12,13], the dependence of acoustics on architectural style [14,18], the correlations between subjective impression and objective parameters [15], the influence of occupancy on listening conditions [20]. Further researches were carried out on Serbian-Orthodox churches [21] and on Gothic-Mudejar churches in Seville [22-25]. They proposed an empirical model to predict relative sound pressure level and clarity parameters, however, predictions are reliable only for this specific type of churches surveyed and only for sound level.
An attempt to organize this complex matter was performed by Desarnaulds [11], who grouped the data of about 200 Swiss churches and other 200 churches built in other countries, statistically analysing the results. An aspect of great interest is the analysis of the optimal values of the acoustic parameters (particularly of the reverberation time) to ensure good listening conditions. He correlated objective measurements with the results of listening tests carried out during the celebration of the services. He proposed different optimal values according to the liturgical style and to the volume of the church.
In the last years the proposing research units started a study of the acoustical characteristics of Italian churches.
Magrini and Ricciardi surveyed ten churches in Genova [26-27] having rectangular plan with three naves, with or without side chapels and domes, and volume varying between 2000 and 40000 m3. A further survey was made in 14 churches in Genova with central plan and volume varying between 1500 and 45000 m3 [28], and in three Cistercian abbeys in the Northern Italy [29], with volume varying between 3000 and 15000 m3. These surveys showed that acoustical parameters like clarity and definition are strongly dependent on early reflections and, consequently, on the architecture. This observation can be extended to reverberation time as well, because as shown in [28] below domes the reverberation is shorter than in other locations.
Cirillo and Martellotta began a study of the acoustics of churches starting from Apulian-Romanesque churches [30-35], which was later extended to a wider sample of churches distributed over the national territory and particularly significant in terms of cultural heritage, but also representative of the different architectural styles, localizations, and architectural typology [36-39]. In this group of churches the most important acoustic parameters (monaural, binaural, and spatial) were measured, emphasizing the effects of architectural differences on acoustic behaviour, showing the existence of correlations between different acoustical parameters. The study allowed to define a method to predict energy-based acoustical parameters originally defined and validated on Romanesque churches [34], and later refined and generalized to churches of different style and typology [39]. Recently, they started to investigate correlations between acoustics, architecture, and subjective judgments with reference to the optimal conditions to listen to music. These researches are based on the statistical analysis of the results of a listening test [38], carried out in an acoustically controlled room where different music motifs are presented to a panel of listener by using auralization. According to the early-stage results of the subjective investigation it seems that in acoustically critical places like churches, listeners tend to adapt to that specific place. In fact, listeners show a preference for listening conditions where the lack in some parameters (e.g. too much reverberation) is compensated by a better performance of the others (e.g. the spatial parameters). However, the complexity of this matter (involving subjective preferences) needs further investigations.
For historical opera houses a specific measurement protocol has been proposed in order to allow comparisons between different experimental results [40], but for churches similar guidelines are not available.
A further lack shown by the bibliographic analysis is referred to the acoustic characterization of materials and elements typically found in churches. The lack of such information is mostly due to the difficulty to move or reproduce such structures in a reverberant room. However, alternative in-situ techniques are now available [41-44], allowing a significant improvement of the knowledge in this field. A further characterization of the materials is required in terms of scattering coefficients, that were defined and standardised only recently [45-46] but they play a fundamental role in acoustic simulation and in order to understand the way in which apparently simple structures interact with the sound field. <<<