Variation of sound properties in the stage and orchestra pit of two European opera houses Lamberto Tronchin, Ryota Shimokura, Valerio Tarabusi DIENCA CIARM, Viale Risorgimento, 2 Bologna, Italy {tronchin, ryota}@ciarm.ing.unibo.it Opera houses differ considerably from normal theatres because of the orchestra pit and the stage, which represents two different areas with quite different sound characteristics for musical instruments and singers. The acoustics of the theatre is strongly depending on the position of sound sources in the area of stage and of orchestra pit, and not only depending on the position of the s. In this paper the blending of sound sources between stage and orchestra pit has been studied in two European theatres, namely the Teatro Nacional de São Carlos in Lisbon (P) and the Teatro Nuovo in Spoleto (I). Both the theatres have two wide stage and orchestra pit, and in both cases acoustical behaviour in the gallery and in the stall have been determined by means of binaural and b-format measurements. Nevertheless, due to space constrains, only binaural measurements will be here considered. The colouration of the orchestra pit in both cases revealed a special feature especially of spatiality, as pointed out from the calculation of binaural and spatial parameters (IACC, LE, LF). This behaviour was found also in the stage of both the theatres, where moving sound source from one side to the other caused variability in spatial parameters, especially in IACC, which at some frequencies ranged from 0.2 to 0.4. The directional path was also considered. Starting from 3D impulse responses measured in both cases, lateral sound propagation was considered, and it revealed to be strongly responsible of modification of the perception of spatiality. Finally, blending between orchestra pit and stage was studied, and related with sound path in the theatre. The balance between the orchestra pit and the stage was measured considering different sound propagation in the two theatres. 1 Introduction Nowadays, sound spatialisation is considered quite important during design of auditoria and virtual audio reproduction of sound quality in dedicated listening rooms for 3D reproduction purposes. Therefore, the definition and measurement of sound spatialisation have been strongly enhanced in last years. Even though international standards like ISO 3382 require measuring some spatial parameters (i.e. LE, LF, IACC), usually only binaural measurements are performed, by means of a dummy head, and rarely 3D impulse responses are measured and utilised for sound reproduction. In this paper, an innovative procedure of measuring spatial sound characteristics recently developed is utilized. The application of this new technique in virtual 3D sound reconstruction is presented. Furthermore, this innovative technique has been applied in two European Opera houses in order to study the blending of sound sources perceived in the stalls and boxes, by means of moving the position of sources in dozens of points in the stage and orchestra pit. Starting from the experimental data collected during the measurements, spatial parameters have been calculated. Moreover, a special care has been reserved for the variation of binaural acoustical parameters (Inter-Aural Cross Correlation) perceived in the two opera houses, analysing the variation of IACC and relating this variation with the geometry of the stage and orchestra pit. 2 The Teatro Nacional de São Carlos of Lisbon, Portugal The Teatro Nacional de São Carlos, Lisbon, was realised during the years 1792-1793. A Portuguese architect, José da Costa e Silva, designed the theatre. Costa e Silva studied in Italy, and was a member (socio) of the Academia Clementina of Bologna. Therefore Costa e Silva was partially influenced by the Bolognese architect Galli-Bibiena. He was in several towns in Italy, mainly in the Republic of Venice and Kingdom of Naples. However, the design of the Lisboetan theatre was mainly influenced by the Theatre alla Scala of Milan, built in 1776 by Piermarini. The theatre has an orchestra pit, which was realized few years before II World War. Below the Orchestra pit there is a special acoustic cavity, which is very to the shape of a boat, even though actually it is often used as a deposit of many different things. This special acoustic cavity (called acoustic well), could be found also in other Italian theatres, as in the Teatro Alighieri of Ravenna, and commonly it is likely that it could increase reverberation time at low frequency, and cause 2257
a sense of warmness of sound quality. The musicians, however, consider the orchestra pit having a poor acoustics, and they complain about its dimensions, being considered too small. Figure 1: Acoustic cavity below the orchestra pit in São Carlos, Lisbon in a box as shown in Fig. (2). Ten positions of sound source were chosen coincident between stage and pit, i.e. the positions of the stage and pit were the same (once below and once above). For sake of convenience, the 3-dimensions in the theater are named here as depth (front rear for the audience in the stalls), width (left right for it) and height (down - up for it). The floors of stage were marked with a 1 m 1 m grid and the sound source was put one after another in the 11 positions (for the width direction) 3 positions (for the depth direction). In the pit the sound source was put one after another in the 11 position of the first raw, whereas alternated in the remaining 12 positions. The height of the sound source was 1.4 m in the stage and 1.2 m in the orchestra pit. The in the stalls and in the box was put as shown in figure 2, whereas in the box at the fourth floor, near the opening. The face of dummy head was turned to the center of the stage during the measurements. The stage did not contain any scenery, and there were no musical instruments or chairs in the orchestra pit. In the stage of the theatre, there is a cyclorama, a sort of orchestra shell (in plaster and concrete), which was realized during the refurbishments of 1940s. The cyclorama influences the acoustics of the stage, as well as the realization of operas in the theatre. 2.1 Measurement set up In the Teatro São Carlos the impulse responses (binaural and b-format) have been obtained generating a logarithmically sine-swept FM chirp by a PC. The starting and ending frequencies were 20 Hz and 20 khz, with duration of 30 s, followed by 10 s of silence. These sine-swept signals were emitted by a portable, omni-directional, frequency-equalized loudspeaker (Look Line) put in the stage or in the orchestra pit. The responses were picked up by a dummy head (Neumann: KU 100) and a Soundfield Microphone (ST-250). When the source signals were generated and the responses were saved in the PC, a Digigram VX V2 board with AD/DA converters of 24 bit resolution employed the cording with a sampling rate 48.1 khz and a 32-bit sampling size. These responses were convoluted with an inverse filter that was a reversed sine-swept signal in terms of time, and the linear impulse responses were therefore obtained. 2.2 Arrangements of sound source and dummy head The sound source was located one by one in 33 positions on the stage and in 23 positions in the pit, whereas the was located either in the stalls or Figure 2: Source positions in São Carlos, Lisbon. Above. stage; below: orchestra pit 3 The Teatro Nuovo of Spoleto, Italy The Teatro Nuovo of Spoleto was realized in 1853 by Ireneo Aleandri, a well-known architect for his works in theater s planning, like the Sferisterio in Macerata, the Theater of Ascoli Piceno and the Theater of San Severino in the central part of Italy. It is likely that, during his design work on Teatro Nuovo, Aleandri 2258
asked to an architect, Luigi Poletti, for some suggestions about planning problems regarding first of all architectural acoustics. However, he did not follow Poletti s theory about the shape of the ceiling of the main hall. Aleandri studied also some documents of another architect in Mantua (Italy) dating back of almost two centuries, Fabricio Motta, who wrote one of the earliest textbooks about plans for theaters with particular attention to acoustical aspects. The Teatro Nuovo in Spoleto opened in the 1864 in spite of some vicissitudes; the stalls plan are horse-shoe shaped in the style of the classical Italian opera house, and the frontage of the four box rows or orders faces directly to the stalls. The last box order was crowned by a loggia or a balcony, and the ceiling was connected to it by a kind of coupling called Vanvitelli style or umbrella, typical in that period. Boxes and loggia are located within a wooden structure made by beams and pillars starting from the lowest balcony s floor and rising until to the ceiling; the central sides of the beams are linked together by a curved wood structure supporting the ceiling frame. Stalls, boxes and loggia can contain a maximum of 800 persons. 3.1 Measurement set up To obtain binaural and b-format impulse responses in Teatro Nuovo of Spoleto, a logarithmically sineswept FM chirp was generated by a PC. The starting and ending frequencies were respectively 40 Hz and 20 khz, and the duration of each sequence was 18 s. These sine-swept signals were emitted by an omnidirectional, pre-equalized loudspeaker (Look line dodecahedral sound source), and put in the stage or in the orchestra pit. The responses were picked up by a dummy head (Neumann: KU 100) and a Soundfield MK-V probe. s16-s20 s11-s15 s6-s10 s1-s5 p1-p5 in the box on the stalls (a) in the box (b) in the stalls Figure 3: Arrangement of 20 sound source positions either in the stage or in the pit and 1 position either in the stalls or in the box. (a) plan and (b) section of Teatro Nuovo in Spoleto (Italy). When the source signals were generated and the responses were saved in the PC, a Layla24 board with AD/DA converters of 24 bit resolution employed the recording with a sampling rate of 96 khz and a 32-bit sampling size. These responses were convoluted with their inverse filter that was a reversed sine-swept signal in terms of time, and the linear impulse responses were obtained immediately. 3.2 Arrangements of sound source and dummy head The sound source was located one by one in 20 positions on the stage and in 20 positions in the pit, arranged, while the was located either in the stalls or in a box as shown in Fig. (3). For sake of convenience, the 3-dimensions in the theater are named here as depth (front rear for the audience in the stalls), width (left right for it) and height (down - up for it). The floors of stage and orchestra pit were marked with a 2 m 2m grid and the sound source was put one after another in the 5 2259
positions (for the width direction) 4 positions (for the depth direction). The height of the sound source was 1.4 m in the stage and 1.2 m in the orchestra pit. The in the stalls was put in the middle of the stalls (for the depth direction) and the right (for the width direction). The box keeping the was located in the rear (for the depth direction) and the right (for the width direction) at the third floor. In the box, the dummy head was brought near to the opening and chairs were moved close to the door. In each position, the height of the dummy head above the floor level was 1.1 m in base of the ear position when a listener sits down. The face of dummy head was turned to the center of the stage during the measurements. The stage did not contain any scenery, and there were no musical instruments or chairs in the orchestra pit. 4 Comparison between the two theatres The overall values of IACC measured in the two theaters are reported in figures 4 and 5, respectively. Figure 4 IACC. Teatro São Carlos, Lisbon. Left: mic in box; sources in the stage (above) and pit (below). Right: mic in stalls; sources in the stage (above) and pit (below). Figure 5 IACC. Teatro Nuovo, Spoleto. Left: mic in box; sources in the stage (above) and pit (below). Right: mic in stalls; sources in the stage (above) and pit (below). 2260
From an architectural point of view, although both the theaters have a horse-shoe shape, with similar overall dimensions, they considerably differ in relation with the dimension and shape of the orchestra pit. In Lisbon the orchestra pit is little, whereas in Teatro Nuovo it is very large. Moreover, also the stages differ, since in São Carlos the cyclorama influences early reflections coming from the rear walls. Nevertheless, as a general tendency, in both theatres sound quality in the orchestra pit and stage are different, as reported also by musicians in Lisbon. 5 Analysis of results Among Teatro São Carlos (Lisbon, Portugal) and Teatro Nuovo (Spoleto, Italy), it is not immediate to compare the distributions of IACC measured in the box s, because the positions of boxes in which these two measurements carried out are slightly different: in Lisbon the was located at fourth order box, whereas in Spoleto at third order box. For the measurements using stalls s, on the other hand, the s positions are similar as shown in Figure (2) and Figure (3), so that the common tendencies can be observed easily in these distributions of IACC for the stalls s. In the two measurements, values of IACC measured in the pits are smaller than those measured in the stages. The direct sounds from pit sources are obstructed by the pit rails (fence between pit and stalls) in the pathways to the stalls s. The fence acts as an important support for singers in the stage, and in the same time enhances the diffusion of sound perceived mainly by the listeners in the stalls. In the stalls, the weaken direct sounds make IACC lower and audience does not perceive the distinct direction from which the sound is coming. Also in the boxes this tendency could be found, even though the position of in the box itself (close to the fence or in the middle of the box area) influences the results The effect of rear walls of the stage rooms and the pit rooms has influence to the values of IACC. In Teatro São Carlos this effect is emphasized, since the cyclorama focalizes reflections coming from behind in the stage. When the sound sources in the stage and pit are close to the rear walls, the reflections from the rear walls are coming to the stalls s immediately after the direct sounds and the values of IACC become larger. 6 Discussion and conclusions The variation of IACC from stage and orchestra pit, with s located in fixed positions in stalls and boxes in both Italian-style Opera houses, resulted very noticeable. It is very interesting to compare in both cases also other spatial parameter, like LE and LF, which showed a similar tendency, even though not so marked as in the case of IACC. It should be underlined that also other, stable monoaural parameters, like center time, clarity and reverberation time, differ considerably in relation with sound source position in the stage or orchestra pit, and with frequency. All these effect could be finally virtually reproduced in a dedicated, listening room (Ambisonics + Stereo Dipole), as Arlecchino room in the University of Bologna. References [1] L. Tronchin, A. Farina - "The acoustics of the former Teatro "La Fenice", Venice", JAES Vol. 45, Number 12 p. 1051 (1997) [2] M. Gerzon - "Recording Concert Hall Acoustics for Posterity", JAES Vol. 23, Number 7 p. 569 (1975) [3] A. Farina, L. Tronchin Measurement and reproduction of spatial sound characteristics of auditoria, Acoustical Science and Technology, 26(2), pag 193-199, 2005 [4] A. Farina, L. Tronchin Misurazioni acustiche nei teatri e nelle sale, Acustica musicale ed architettonica, Cap. 19, UTET, Torino, 2005 [5] A. Farina, R. Ayalon - "Recording Concert Hall Acoustics for Posterity" - 24th AES Conference on Surround Sound, Techniques, Technology and Perception Banff, Canada 26-28 June 2003. [6] A. Karamustafaoglu, U. Horbach, R. Pellegrini P. Mackensen, G. Theile - "Design and Applications of a Data-based Auralisation System for Surround Sound, 106th AES Convention, pre-print n. 4976 (1999). [7] M. A. Poletti - "A Unified Theory of Horizontal Holographic Sound Systems", JAES Vol. 48, Number 12 p. 1049 (2000). [8] E. Hulsebos, D.de Vries, and E. Bourdillat - "Improved Microphone Array Configurations for Auralization of Sound Fields by Wave-Field Synthesis", JAES Vol. 50, Number 10 p. 779 (2002) [9] A. Farina, R. Glasgal, E. Armelloni, A. Torger - "Ambiophonic Principles for the Recording and Reproduction of Surround Sound for Music" - 19th AES Conference on Surround Sound, Techniques, Technology and Perception - Schloss Elmau, Germany, 21-24 June 2001. [10] A. Farina Simultaneous measurement of impulse response and distortion with a swept-sine technique, 110th AES Convention, Paris 18-22 February 2000. [11] L. Tronchin, V. Tarabusi Auditoria and dismissed structures: the acoustical design, 2261
WSEAS Transactions on Mathematics, 3(3), pag. 671-674, 2004. [12] G. Theile Multichannel Natural Music Recording Based on Psychoacoustic Principles - AES 19 th International Conference, May 2001. [13] L. Tronchin, L. Lanciotti, V. Tarabusi La variazione della qualità acustica in relazione alla localizzazione delle sorgenti sonore, Proc. XXXI AIA, Venezia, Italy, 2004 [14] M. Williams, G. Le Du Multichannel Microphone Array Design, 108th AES Convention, 2000, Preprint 5157. [15] A. Farina, L. Tronchin Advanced techniques for measuring and reproducing spatial sound properties of auditoria, Invited Paper. Proc. of RADS, Room Acoustics: Design and Science, Awaji, Japan, 2004 [16] L. Tronchin, V. Tarabusi, A. Giusto: The realization of Ambisonics and Ambiophonics listening room Arlecchino for car sound systems evaluation, Proc. 21st AES Conference, St. Petersburg, Russia, 2002 2262