Roman Theatres and Revival of Their Acoustics in the ERATO Project

Home , Amphitheatre, Odeon (building), Roman amphitheatre, Roman Theatre (Mérida)

ACTA ACUSTICA UNITED WITH ACUSTICA Vol. 99 (2013) 21 –29 DOI 10.3813/AAA.918584

Jens Holger Rindel Odeon A/S, Scion-DTU, Diplomvej 381, 2800 Kgs. Lyngby,Denmark. [email protected]

Summary The ERATO project wasathree-year research project financed by the European Commission under the Fifth Framework INCO–MED Program (2003–2006). The acronym refers to the project title: Identification, Evalua- tion and Revivalofthe Acoustical Heritage of Ancient Theatres and Odea. The ancient Greek and Roman theatres are famous for their excellent acoustics. However, it is not generally well known that different kinds of theatres were built, for different purposes and with different acoustical conditions. The development of the Roman theatre types (theatre, odeon and amphitheatre)particularly in the first Century BC is briefly outlined. One aim of the ERATO project has been to investigate the acoustics of the open air theatre and the odeon, using virtual reconstructions be means of computer models of the spaces, completed in accordance with available archaeological information. Musical instruments and short pieces of music have been reconstructed and recordings have been used for auralisation of some theatre scenarios. Ancient Greek and Roman theatres are often considered acoustically perfect. However, the semicircular shape of the audience area in theatres may cause acoustic problems, and there is also evidence that the ancient architects were aware of this. The Roman architect Vitruvius mentions in his famous books on architecture four different kinds of sound reflections in atheatre, one of them called ‘circumsonant’ which is probably the acoustical phenomenon that we today would name afocused echo. Computer simulations of some examples of ancient Greek and Roman theatres confirm that echo can occur at some places in the audience area. Apossible solution to these echo problems could be the introduction of sound absorption in the vertical, concave surfaces in away similar to that described by Vitruvius for the sounding vessels, i.e. in niches between the seats arranged in ahorizontal range halfway up (inthe diazoma). Thus it makes sense if the vessels were supposed to act as sound absorbing resonators, although the number of resonators is farfrom sufficient for having anyreal effect. The idea and principle of installing resonators in atheatre comes from Aristoxenus (4th century BC), who wasafamous Greek philosopher and scholar in music theory.Itisconcluded that the sounding vessels had no practical importance in the Roman theatres, and Vitruvius wasnot up-to-date in his writing on theatre design, butrelied heavily on older,Greek references.

PACS no. 43.55.Gx, 43.55.Ka

1. Introduction popular newer building types in Roman cities. As abackground it is important to have acloser look at the context It is impossible to discuss the acoustics of ancient theatres of Vitruvius’ books, and particularly the remarkable devel- without areference to the Roman architect Vitruvius and opment of theatre types during the ﬁrst century BC. his “Ten Books on Architecture” [1, 2], probably dating The EU project ERATO has dealt with virtual recon- from around ca. 30–27 BC. The ﬁfth book is devoted to structions and acoustical simulations of selected theatres theatres and their acoustics and while alot of information and odea. Acoustical measurements were also made, par- has been provided that is important for our understanding ticularly interesting in the Aspendos theatre because this of the acoustical design of the ancient masters, the book is one of the best preserved Roman theatres today.With also contains detailed descriptions about more doubtful the characteristic semicircular shape of the Roman the- “sounding vessels” that have led to manyquestions and atres it is no big surprise that focusing and echo problems speculations among modern acousticians. It is also impor- can occur.This wasalso known by the ancient architects. tant to note that only the older theatres of Greek or Roman By means of computer modelling an example of aRoman type are dealt with, whereas Vitruvius givesnoadvice con- theatre has been analysed applying an echo parameter.Fi- cerning the amphitheatre or odeon, which were both very nally,the sounding vessels described by Vitruvius are discussed. Apossible relation with echoes and focusing ef- fects is suggested because the recommended installation Received2May 2012, of the vessels could makesense, it least in principle, if accepted 2October 2012. theyshould work as resonant absorbers.

©S.Hirzel Verlag · EAA 21 ACTA ACUSTICA UNITED WITH ACUSTICA Rindel: Roman theatres in the ERATO project Vol. 99 (2013)

2. Theatres in the Roman Empire Thus, aconnection between gladiatorial combats and mil- itary training wasestablished [6, p. 79]. 2.1. Types of Roman theatres 2.1.1. Theatre 2.2. The development of theatredesign during the The ancient Greek theatre had Archaic forerunners from first century BC the 5th century BC and developed as an architectural struc- As abackground for evaluation of theatre design as de- ture during the 4th century BC [3]. The theatres were scribed in the fifth book of Vitruvius, it is useful to takea spread overthe Greek cultural area during the following closer look at the development of theatre design in the first centuries. Theywere open-air performance spaces related century BC. to the religion (Dionysos)and designed for drama perfor- In Rome it wasnot allowed to build permanent theatres mance with emphasis on good speech intelligibility.The for political reasons; instead smaller,temporary theatres theatre and particularly the skene building wasfurther de- were built of wood. The first permanent masonry theatre in veloped and improvedinthe Hellenistic period, and from Rome wasbuilt 61–55 BC by Pompeyonhis private prop- around the middle of the first Century BC continued by erty and he claimed it to be atemple, not atheatre -hence the Romans with further development including an en- the small Venus temple in the top rear of the audience area. larged skene building connected to acollonade behind the It had room for about 17 000 people. The Marcellus the- audience, together forming asemicircular building struc- atre wasprobably the second permanent theatre in Rome, ture that could be covered with textile likesails as asun- inaugurated in 13 or 11 BC, butithad already been used shield (velum). Behind the skene building a portico (nearly for afeast in 17 BC. It seated about 15 000 people. Athird square place surrounded with acolonnade)was often an theatre in Rome wasthe Balbus theatre, which might have integral part of the Roman theatre complex, as mentioned been earlier than the Marcellus theatre, butlittle is known by Vitruvius [1, Book V.ix.1]. While Greek theatres were about this [6, p. 121]. built into hillsides, the Roman theatres were generally built In Pompeii alot of building activity started after 80 BC, on flat ground with access for the audience via vormitoria when it wasmade acolonyfor Sulla’sveteran soldiers. (i.e. entrances for the audience from the outside through Pompeii had already an older Hellenistic theatre from 3rd– narrowpassages under the seats of the theatre). 2nd century BC, butinaddition to this twocompletely newtypes of Roman theatres were erected in Pompeii 2.1.2. Odeon shortly after 80 BC: the odeon for song and music around The odeon (gr. ωδεoν,lat. odeum)was also originally a 75 BC, and the amphitheatre for gladiatorial combats and Greek building type used for song and music. It wasa entertainment with wild animals (lat. ‘venatio’) probably roofed, columned hall, butthe columns disturbed sight- around 70 BC [4, p. 174, 177]. The odeon wasrelatively lines, as the odeon of Pericles at Athens, built ca. 446–442 small and surrounded by thick walls to carry the roof cov- BC [4, p. 59]. From the first century BC the Romans de- ering both auditorium and stage. Both newbuildings were veloped the odeon into aroofed theatre without columns erected under the local reign of Quinctius Valgus and Mar- and with the audience seated in asemicircular arrange- cus Porcius, and both building types became very popular ment similar to that of the theatre [5]. It is characteristic in the following centuries as asupplement to the theatre: for the odeon that the building has very thick walls that the odeon in manyRoman cities all overthe empire and could carry the roof, and the shape is either rectangular or the amphitheatre only in the western part of the empire. semicircular,the latter presenting aconsiderable challenge The amphitheatre wasnever adopted by the Greek culture to structural engineering. in the east. “While the Romans built amphitheatres side by side with the theatres, the Greeks built music halls or au- 2.1.3. Amphitheatre ditoria even in remote parts of the Greek world during the The amphitheatre is aspace connected to gladiatorial com- imperial period.” [4, p. 222] bats as an entertainment, which had its origin in Italy,pos- In Rome the gladiatorial combats originally took place sibly from the Etruscian or Samnite culture. Traditionally in the Forum (Agora)following ancient traditions in Italy the fights took place in the Agora as mentioned by Vitru- [1, Book V.i.1], butin30BCRome’sfirst stone amphithe- vius [1, Book V.i.1], butduring the first Century BC aspe- atre waserected by Statilius Taurus as apart of Augustus’ cificplace for the games wasdeveloped, first as awooden great building program [6, p. 108–127]. Taurus’ amphithe- structure, and later as amasonry building [6]. The early atre may have played an important role as amodel for versions were called Spectacula,aplace for watching, and the design on manyamphitheatres in the following years. later the name ‘Amphitheatre’ wasused. The main char- However, the amphitheatre wasdestroyed in the great fire acteristics are the oval shaped ‘Arena’ in the middle and AD 64. Soon after,the Flavian Amphitheatre (Colosseum) the audience seated on all sides around the arena. Likethe wasbuilt AD 75–80, amuch bigger and structurally ad- theatre it wasanopen-air space with the possibility of a vanced masonry building. Only one odeon is known to velum as asun shield. It seems that the amphitheatres sud- have been built in Rome, namely by Domitian in AD 86. denly appeared in Roman towns and cities, and according So, the Taurus stone amphitheatre existed in Rome when to Welch [6] the reason for that wasamilitary crisis in Vitruvius wrote his books, and at least 16 other amphithe- 105 BC when it wasdecided to use gladiators for train- atres in Italy date to the late republican period [6, p. 191], ing of newsoldiers in the efficient use of various weapons. i.e. theyexisted at the time of the writings of Vitruvius.

22 Rindel: Roman theatres in the ERATO project ACTA ACUSTICA UNITED WITH ACUSTICA Vol. 99 (2013)

2.3. Theatres in the city plan so the ancients deserve no ordinary,but unending The different purpose and use of the various theatres is thanks, because theydid not pass on in envious si- reflected in the city plan. Avery good example is Pom- lence,but took carethat their ideasofevery kind peii, where the theatre and the odeon are in the city centre should be transmitted to the futureintheir writings.” near the temples and other public buildings, whereas the Alittle later Vitruvius continues in the same manner [1, amphitheatre is located in aremote corner near the city Book VII.introduction.14]: wall. ManyRoman towns were founded as colonies where “. ..Fromtheir commentaries Ihave gatheredwhat I army veterans could settle. This wasdone already by Sulla saw was useful for the present subject, and formed it (Pompeii, 80 BC)and Julius Caesar (e.g. Corinth, 44 BC) into one complete treatise,and this principally,be- and on agreater scale by Augustus. In all these cases the cause Isaw that manybooks in this field hadbeen amphitheatre wasafundamental part of the city plan. Au- published by the Greeks, butvery fewindeed by our gusta Praetoria (Aosta)inNorthern Italy wasone of Au- countrymen.” gustus’ first veteran colonies founded 24 BC, and both an Concerning his ownexperience as an architect he men- odeon and an amphitheatre were among the public build- tions, that he superintended the basilica in Fana, Italy [1, ings, being important parts of the city plan, the whole city Book V.i.6]. This is the only building that we knowVit- being surrounded by ahigh city wall [7]. Another exam- ruvius had actually worked on. It is noted that the Book ple is Augusta Emerita (Mérida)inSpain founded 25 BC Vontheatre design provides details of the Roman and as the capital of anew colony. Again the theatre and am- the Greek theatre. However, the latter were no longer be- phitheatre constituted afundamental part of the overall ur- ing built, whereas the newRoman amphitheatre and the ban design. Already after ten years the theatre wasinaugu- odeon had already become significant public buildings in rated and alittle later the amphitheatre, which wasfinished the cities, particularly obvious in the newcities founded as in 8BC. aresult of Augustus’ great building program. But Vitru- Atypical city plan is also found in Verona with the fa- vius is silent about these buildings, probably because they mous amphitheatre (early 1st century AD)erected in the were not included in the Greek treatises, that he wasre- outskirts of the ancient city,whereas remains of an older lying on, and apparently he had no experience of his own theatre and aneighbouring odeon are found further east at concerning theatre design. the riverside, near the original city centre. 4. The ERATO project 3. Vitruvius and his books on architecture The ERATO project wasathree year project under the EU Marcus Vitruvius Pollio wasaRoman architect and en- 5th Framework “Preserving and using cultural heritage”, gineer,known today as the author of “The ten books on Project ICA3-CT-2002-10031 running from 1February architecture” [1, 2]. The exact years of his birth and death 2003 to 31 January 2006 [8]. The acronym refers to the are unknown. However, he mentions in the preface of the project title: Identification, Evaluation and Revivalofthe books that he had served under Julius Caesar during wars. Acoustical Heritage of Ancient Theatres and Odea. The main objectivesofthis research were identification, If he means the Gallic wars (58-51 BC), Vitruvius may virtual restoration and revivalofthe acoustical heritage have been born before or around 80 BC. He mentions that in afew,selected examples of the Roman theatre and the he served to supply and repair ballistae, scorpiones,and roofed odeon in a3Dvirtual environment. The amphithe- other artillery [1, Book I.preface.2]. atres with their clearly different purpose were not included The dedication in the introduction to “Caesar Impera- in this project. The virtual restitution integrates the visual tor”, i.e. Octavian, suggests that the books were written be- and acoustical simulations, and is based on the most recent fore 27 BC, because from that year the emperor wasgiven results of research in archaeology,theatre history,clothing, the title “Augustus”. This dating coincides with the great theatre performance and early music. building program initiated by Octavian (Augustus)around The Project group had sevenpartners: 30 BC, including the establishing of manycolonies and • The Technical University of Denmark (DTU), Acoustic newtowns all overthe empire. So, presumably there was Technology,Kgs. Lyngby,Denmark, aneed for architectural guidelines in Latin, and Vitruvius • Yildiz Technical University,Physics Department, Istan- wasgiven the task to write such guidelines, since existing bul, Turkey, guidelines were mostly in Greek language. Thus, Vitruvius • The Hashemite University,Institute of Tourism and draws extensively on older,mainly Greek, writings on ar- Heritage, Amman, Jordan, chitecture and building technology,none of which exists • The University of Ferrara, Department of Engineering, today [1, Book VII.introduction.1]: Ferrara, Italy, “It was awise and useful provision of the ancients • AEDIFICE, Laboratory of Psycho-acoustics, Lyon, to transmit their thoughts to posterity by recording France, them in treatises, so that theyshould not be lost, but, • Swiss Federal Institute of Technology (EPFL), Lau- being developed in succeeding generations through sanne, Switzerland, publication in books, should gradually attain in later • University of Geneva,MIRALAB, Geneva,Switzer- times, to the highest refinement of learning.And land.

23 ACTA ACUSTICA UNITED WITH ACUSTICA Rindel: Roman theatres in the ERATO project Vol. 99 (2013)

Table I. Location and main dimensions of the selected theatres and odea as reconstructed in the computer models. Caveadiameter is up to the last rowofseats. The slope is givenseparately for the lower part (ima cavea) and the upper part (summa cavea).

CaveaOrchestra Ima caveaSumma cavea Location diameter [m] diameter [m] slope [degree] slope [degree]

Aspendos theatre Turkey 84.4 24.0 35 35 Jerash South theatre Jordan 60.8 18.6 41 41 Syracuse theatre, Roman Sicily 135.5 30.6 24 20 Aosta odeon Italy 49.1 12.6 30 30 Aphrodisia odeon Turkey 42.5 10.0 36 31

Figure 1. Above:Photos from the three selected theatres, below: View from computer models, reconstructed for the Roman period. Left: Jerash (South theatre), Middle: Aspendos, Right: Syracuse.

The following summary will focus on the acoustical re- cal excavations, and manyofthe interior details, likestat- sults. The acoustical simulations were made with the room ues and marble floor in the orchestra, still exist. However, acoustics software ODEON ver. 7.0, developed at the Aphrodisias also has awell preserved large theatre. Technical University of Denmark. Acoustical measurements were made in the theatres in Aspendos and the South theatre in Jerash. In the theatre of 4.1. The selected theatres and odea Syracuse and the twoodea the state of preservation wasnot Five spaces have been selected for virtual reconstruction in sufficient to makeacoustical measurements meaningful. the ERATO project: three theatres, see Figure 1, and two odea, see Figure 2. Information about name, location and 4.2. Acoustical results from virtual reconstructions main dimensions are collected in Table I. In Aspendos the theatre is one of the best preserved examples of aRoman 4.2.1. The theatres in Aspendos, Jerash and Syracuse theatre, butinthe nearby archaeological site the remains These three theatres differ in shape and size as well as in of an odeon are also found. In Jerash there are remains the slope of the Cavea (the auditorium of the theatre). In of twotheatres, the newly excavated North theatre, which Roman times the Syracuse theatre had acolonnade behind wasprobably an odeon, and the larger South theatre, which the last rows of the cavea likethe one found at the the- wasselected for this project. The theatre in Syracuse has atre in Aspendos. The Aspendos theatre had a Velum (sun- along history of evolution from Greek to Hellenistic and screen overthe audience area)made of wool likeasail and finally to aRoman style theatre, and it wasselected for that carried by poles or masts, the mounting system of which reason. is still clearly visible. Probably the Jerash South theatre The odeon in Aosta wasparticularly interesting for this also had acolonnade, butthere are no remains left from project because this is the only known example where that and thus no colonnade wasincluded in the reconstruc- some of the outer walls still exist in full height. The odeon tion of this theatre. However, it is very unlikely that the in Aphrodisias wasselected because there wasvery good theatre should have been built without the colonnade; it is and detailed information available from the archaeologi- mentioned by Vitruvius as an important part of the Roman

24 Rindel: Roman theatres in the ERATO project ACTA ACUSTICA UNITED WITH ACUSTICA Vol. 99 (2013)

Table II. Calculated parameters in diﬀerent conﬁgurations (empty and fully occupied)ofthe theatres. The parameter values are averaged overall source-receiverpositions and overmid-frequencies 500–1000 Hz in 1/1-octave bands.

T30 [s] G [dB] C80 [dB] STI Empty Full Empty Full Empty Full Empty Full

Aspendos Roman 1.95 1.59 -2.34 -4.36 1.17 4.08 0.53 0.61 Aspendos present 1.89 1.53 -4.37 -6.09 2.68 6.53 0.60 0.70 Aspendos present (new stage)1.77 1.43 -4.49 -6.05 4.42 8.27 0.63 0.71 Jerash South Roman 1.54 1.06 -0.72 -3.05 3.46 6.88 0.62 0.70 Jerash South present 1.21 0.86 -1.18 -3.29 5.98 9.85 0.67 0.75 Syracuse Roman 1.81 1.67 -6.69 -8.24 4.07 8.25 0.62 0.70 Syracuse present 1.25 0.97 -10.60 -11.61 12.88 18.12 0.88 0.93

Figure 2. Above:Photos from the twose- lected odea, below: View from computer models, reconstructed for the Roman period. Left: Aosta, Right: Aphrodisias.

theatre, and it also had astructural function as support for remarkably high in theory if we neglect the background the masts that should carry the Velum. noise. In Table II are shown the acoustical parameters for the Forthe Aspendos theatre in present state, the modern three theatres in their different configurations. stage provides adecrease of reverberation time at mid- frequencies, and in the case of the Roman reconstruction The Table II shows that the difference in reverberation (with added velum and stage canopy) there is only minor time between empty and full is about 0.3–0.4 sgenerally difference in reverberation. in all theatres, Jerash South having aslightly bigger dif- From Table II it is seen that in Jerash South theatre ference. The reverberation time when fully occupied is the reconstruction of mainly the Frons Scaenae (the richly around 1.6 sinthe Roman reconstructions with colonnade, decorated front of the skene building)results in an increase butonly around 1.0 sinthe reconstruction of Jerash South of the reverberation time of around 0.3 s. It is also seen theatre without colonnade. that the Syracuse theatre in the Roman era had around The overall strength is the highest in Jerash and low- 0.6 slonger reverberation than in the present stage. This est in Syracuse, partly due to the different slopes and the is mainly due to the Frons Scaenae,but also the colonnade great difference in cavea diameter,see Table I. In general provided some reverberation. the strength is very lowcompared to amodern enclosed theatre. The sound field in an open-air theatre is approxi- 4.2.2. The odea in Aosta and Aphrodisias mately atwo-dimensional field, and for that reason the re- These odea were closed spaces with wooden roof struc- lation between strength and reverberation time is different tures; theywere used for more intimate music and theatre from that normally found in closed spaces with athree- plays, often only for an exclusive audience. Apart from the dimensional sound field. roof, theywere constructed of hard materials such as stone The clarity is exceptionally high in all the theatres de- or marble. Theyare presumed to have had open windows spite the levels of reverberation, and this is due to the lack to let in daylight and provide ventilation. The open win- of roof that makethe field more likeafree field than adif- dows and audience seated on the caveawould have pro- fuse field. As aconsequence of this, the STI values are also vided the main acoustical absorption. From Table III one

25 ACTA ACUSTICA UNITED WITH ACUSTICA Rindel: Roman theatres in the ERATO project Vol. 99 (2013)

Table III. Calculated parameters in diﬀerent conﬁgurations (empty and fully occupied)ofeach odeon. The parameter values are averaged overall source-receiverpositions and overmid-frequencies 500–1000 Hz in 1/1-octave bands.

T30 [s] G [dB] C80 [dB] STI Empty Full Empty Full Empty Full Empty Full

Aosta Roman 5.97 3.49 7.14 4.26 -5.28 -2.41 0.36 0.43 Aphrodisias Roman 4.02 1.62 10.45 5.45 -4.21 1.86 0.38 0.55 Aphrodisias present 0.37 0.24 6.42 3.05 15.19 23.21 0.85 0.90 sees that these rooms would have been over-reverberant by time waslonger in the Roman theatre than in the previous present standards. Hellenistic theatre. This work is documented in twoMSc The Aosta and Aphrodisias odea differ both in their vol- projects and one PhD project [9, 10, 11]. ume and in their shape. The outer walls of the Aosta odeon followarectangular shape, whereas in Aphrodisias odeon 4.3. Reconstruction of musical instruments, music, the shape is semicircular following the seating area. The song, clothes and style of performance Aosta odeon had avolume that wasalmost twice that of The musical instruments that have been reconstructed are Aphrodisias odeon as reconstructed, mainly due to the dif- Kithara (a plugged string instrument), Aulos (a double ference in ceiling height. wind instrument), Tympanon (a flat drum)and Scabellum By comparing the reverberation times of the twoodea (a percussion instrument operated by the foot). The instru- as reconstructed to Roman time it is seen, that the Aosta ments represent Roman times around 1st Century AD, al- odeon had alonger reverberation time, mainly because of though similar instruments were used also by the Greeks in the greater volume. earlier centuries. Three pieces of music in the antic style The present ruins of the Aphrodisias odeon have too have been composed for the instruments and recorded in fewsurfaces to provide areverberant field for satisfactory an anechoic environment. acoustics as it is seen from the table. The Aosta odeon has Adatabase on virtual people has been created, based on not been modelled in its present state, since there is only historical descriptions of clothes, shoes and hairstyle. The one wall standing. database consists of twelvevirtual actors from the Roman In the reconstructed models of the Roman era both odea period, three actors from the Greek period and aset of vir- are over-reverberant when theyare empty.The Aphro- tual musicians. Forthe virtual audience acrowd rendering disias odeon has areverberation time T30 when full, which and animation machine has been developed. The method is comparable to the optimum for modern concert halls allows real-time rendering of thousands of animated vir- of similar volume. The Aosta odeon seems to be over- tual humans. reverberant even when full. Visual VR reconstructions in computer models have The strength G of both odea is seen to have optimum been made for the Aspendos theatre and the Aphrodisias values both when empty and full. The clarity C80 of Aosta odeon. The models include details likecolumns and stat- odeon is too low, mainly because of its high reverbera- ues, texture to represent the building materials, the cof- tion, whereas Aphrodisias has an adequate clarity.The STI fered ceiling in the odeon and the velarium for sun shield- values showthat the Aphrodisias odeon is satisfactory for ing overthe open air theatre. Over 150 different textures speech when full, whereas the Aosta odeon is just bear- have been created to represent different surfaces in the able. It has to be mentioned that the background noise buildings. The light conditions can be controlled using a levelofthe audience is not known and probably resulted virtual light probe to simulate anyposition of the sun and in lower speech intelligibility. condition of the sky. Overall, the calculation results of the reconstructed The visual simulations have been integrated with the model of Aphrodisias odeon have shown an excellent acoustical simulations and anumber of animated scenarios acoustic ambience comparable to modern halls. It must have been produced, including sounds from the audience. have been ahall that wasoptimal for music butalso acceptable for plays and chorus events. The Aosta odeon 4.4. Other results from the ERATO project would have been less suitable for spoken events butstill acceptable for music. In relation to restoration works on ancient theatres, newin- formation has been provided concerning the acoustical im- 4.2.3. Acoustical scale model investigations portance of different parts of the ancient theatres and odea. At the University of Ferrara adetailed study wasmade In the Roman theatre the height of the skene building and of the acoustics in different stages of development of the the decoration with columns and statues are particularly large theatre of Syracuse. From scale model investiga- important, butalso the colonnade behind the audience is tions it wasfound that the different architectural layout of acoustically important. In the odeon the roof and the open the theatre in the Greek and Roman periods has an influ- windows are acoustically most important; however, it is ence on the acoustical parameters, and the reverberation not very likely that anyodeon will be fully restored with a newroof.

26 Rindel: Roman theatres in the ERATO project ACTA ACUSTICA UNITED WITH ACUSTICA Vol. 99 (2013)

Figure 3. Computer model of the reconstructed Aspendos theatre.

Figure 4. Early reflections from the stage position to areceiverin 5. Sound reflections in the theatre the orchestra. 5.1. The knowledge of sound reflections in ancient times Vitruvius describes four different kinds of sound reflections in atheatre [1, Book V.viii.1]; the Greek terms are quoted, and in translation theyare named dissonant, circumsonant,resonant,and consonant.Hefurther explains [1, Book V.viii.2]: “The circumsonant arethose in which the voice spreads all round, and then is forced into the middle,whereitdissolves, the case-endings arenot heard, and it dies awaythereinsounds of indistinct meaning.The resonant arethose in whichitcomes into contact with some solid substance and recoils, thus producing an echo, and making the terminations of cases sound double.” The reflection called circumsonant seems to be the focusing effect from concave surfaces. Although the descrip- Figure 5. Early reflections from the stage position to areceiverin tion is not very clear,itisobvious that the focusing effect rear side position. increases the risk of echo problems.

5.2. Echo in the theatre In order to study the possibility of echoes in atypical Ro- man theatre, the room acoustics software ODEON version 11 wasused. The theatre chosen for this study is the reconstruction of the Roman theatre in Aspendos, see Figure 3. With the source near the centre of the stage at aheight of 1.5 mabove the stage floor,the early reflections are shown for areceiverposition in the orchestra in Figure 4, and for areceiverinarear position to one side in Figure 5. The focusing of reflections from the concave seating arrangement is clearly seen in Figure 4, butthere are also re- Figure 6. Calculated impulse response in the receiverposition flections from the concave ‘Diazoma’ (horizontal passage from Figure 5. between several rows of seats)via the canopyabove the stage, and late reflections from the colonnade surrounding more than 50% would evaluate this as aclearly audible the theatre. The calculated impulse response in the latter echo when listening to speech. position shows aclear echo with adelay of ca. 125 ms, see Figure 6. 6. The ‘sounding vessels’ In order to makeitpossible to locate positions with echo problems, the echo criterion suggested by Dietsch 6.1. The purpose and Kraak [12] has been implemented in ODEON version The so-called sounding vessels described by Vitruvius 11. As an example Figure 7shows the echo-curveinthe have givenrise to much speculation, because it is un- same position as the impulse response in Figure 6. The clear what acoustical function theycould serveinathe- echo parameter has avalue above 1.0, which means that atre. Theywere obviously acoustic resonators, buttoday

27 ACTA ACUSTICA UNITED WITH ACUSTICA Rindel: Roman theatres in the ERATO project Vol. 99 (2013)

theatres that were built every year in Rome, because the boarding itself is resonant –“But when theatres arebuild of solid materials likemasonry, stone or marble,which cannot be resonant, then the principle of the ‘echea’must be applied.”Asweknowtoday,the lowfrequencyabsorp- tion can be obtained with panel absorbers, which would be more efficient than the rather small number of Helmholtz resonators recommended by Vitruvius. It is interesting to note that Vitruvius uses the word echea (ηχεια)which in modern Greek means aloud- speaker system. The sounding vessels were discussed in some detail by Izenour [14, p. 34] and with reference to Figure 7. Echo criterion curves for the same position as in Fig- Vitruvius [1, Book V.iii.8] he suggests an explanation: ures 5and 6. The curves are overlapping for the fiveoctave bands “The ancients, using commonsense reasoning,considered from 125 to 2000 Hz. The peak is at 125 ms. it perfectly obvious that if resonatorsorsounding boards of bronze or horn wereused to strengthen the sound produced by astring or wind instrument, the samewould we knowthat the effect can either be sound absorption or hold true for the arrangement of atheater in accordance sound radiation, depending on the internal losses of the with the science of harmony,because in this way (Vitru- resonator.The vessels in the theatre are mentioned sev- vius says) the Greeks increased (amplified) the power of eral times by Vitruvius, first in book I[1, Book I.i.9]: “In the speaking voice.”Atthe end of this discussion Izenour theatres, likewise,thereare the bronze vessels (in Greek concludes [14, p. 39]: “What charmingnonsense!” ηχεια)whichare placedinniches under the seatsinac- cordance with the musical intervals on mathematicalprin- 6.2. The origin of the sounding vessels –Aristoxenus ciples. These vessels arearranged with aviewtomusical concords or harmony,and apportioned in the compass of Concerning the principle behind the sounding vessels and the fourth, the fifth, and the octave,and so on up to the the musical scales used for the tuning, Vitruvius refers double octave,insuchaway that when thevoice of an ac- to the ancient Greek music theoretician and philosopher tor falls in unison with any of them its power is increased, Aristoxenus [1, Book V.v.6]. He livedinAthens around and it reachesthe earsofthe audience with greater clear- 350 BC, wasapupil of Aristoteles and wrote alarge num- ness and sweetness.” ber of treatises on topics within music, ethics and philos- The explanation of the purpose [2, Book V.v.3] reads ophy. His theory on musical scales wasinopposition to in the translation by Bill Thayer: “By the adoption of this the one by Pythagoras based on mathematical principles. plan, thevoice whichissues from the scene,expanding as Aristoxenus claimed that the superior evaluation of musi- from acentre, and striking against thecavity of eachvase, cal intervals should be made by the human ear.Hesaid will soundwith increased clearness and harmony,fromits that we evaluate the size of the intervals by the ear,and the unison with one or other of them.” properties by the brain. In fact he suggested the equally In both quotations it is said that the purpose is to in- tempered scale more than twothousand years before this crease the clearness. Thus this is the opposite of creating scale became generally accepted in the 18th century. reflections that might increase the reverberance. The de- 6.3. Wheredid they exist? scription could makesome sense if the vessels are meant for sound absorption. The position of the vessels should be The evidence of the existence of sounding vessels in the- under the seats, i.e. in the concave semicircle that can cre- atres is very sparse. Obviously Vitruvius has neverseen ate focusing reflections, and thus the sound absorption by them himself, butherefers to [1, Book V.v.8] “...the dis- the vessels will attenuate the reflections that might cause tricts of Italy and in agood many Greek states. We have an echo problem. also the evidence of Lucius Mummius, who, after destroy- Today it is well known that sound absorption is possible ing the theatreinCorinth, brought its bronze vessels to with Helmholtz resonators. However, theyare only effi- Rome.”Itisafact that Lucius Mummius wasaRoman cient in narrowfrequencybands. So it also makes sense general who conquered Corinth in 146 BC, demolished to apply different sizes with resonance frequencies dis- the city and brought lots of treasures to Rome. The theatre tributed overtwo octaves, as described in detail by Vit- wasaHellenistic theatre from the 3rd century BC [3]. The ruvius. This may correspond to the frequencyrange 220– time and location of this theatre fit quite well with those of 880 Hz as suggested by Barba Sevillano et al. [13]; this Aristoxenus, see above. paper contains athorough analysis of the frequencies of Izenour has described the existence of nine equally the vessels in relation to the ancient Greek musical system spaced cavities located behind the diazoma in the ruins of developed by Aristoxenus. aRoman theatre in Beth Shean, Israel [14, p. 39–40]. Al- Another remark from Vitruvius, that is interesting in though the number should have been 13 according to Vit- this connection, is in [1, Book V.v.7], where he says that ruvius, and not only nine, Izenour has shown in sketches there is no need for the sounding vessels in the wooden howthe vessels might have been installed in the cavities.

28 Rindel: Roman theatres in the ERATO project ACTA ACUSTICA UNITED WITH ACUSTICA Vol. 99 (2013)

In 1958 some clay vessels were found when atheatre possibly have made anyimprovement to the acoustics in wasexcavated in Nora, Sardinia. The site wasvisited by practice. It seems that Vitruvius’ efforts to promote these Dr.Brüel, who made aphoto of the vessels together with ancient ideas for Roman theatres have been more or less other observations [15, p. 18]. However, he concludes that ignored by those, who actually were responsible for de- from the findings there is nothing making it likely,that the signing and building the theatres. vessels have improvedthe acoustics of the theatre in any way. Acknowledgement So, the sounding vessels did actually exist in some the- The ERATO project (Contract Number ICA3-CT-2002- atres, butonly in very fewrare cases. In the vast majority 10031), wasfinanced by the European Commission under of Greek and Roman theatres the sounding vessels were the Fifth Framework INCO–MED Program. not used. References 6.4. The sounding vessels, asummary [1] M. Vitruvius: The ten books on architecture. 1st century The idea of installing resonators in atheatre dates to Aris- BC, translated by Morris HickyMorgan. Harvard Univer- toxenos, i.e. to the early days of the Greek theatre. Al- sity Press, 1914, reprint by DoverPublications, NewYork, though it theoretically could makesome sense as amea- 1960. sure against the focusing of sound from the concave shape [2] M. Vitruvius: The ten books on architecture. 1st century of the stone seats in atheatre, the very small amount of ab- BC, original text in Latin and English translation by Bill Thayer.http://penelope.uchicago.edu/ Thayer/ E/ Roman/ sorption that can be obtained in practice makes it clear,that Texts/ Vitruvius/ home.html. the sounding vessels had no audible effect when installed [3] R. Frederiksen: Typology of the Greek theatre building in in atheatre. It is clear that Vitruvius had no experience late classical and Hellenistic times. –In: Proceedings of the with theatre design, butinhis work dealing with transla- Danish Institute at Athens -III. S. Isager,I.Nielsen (eds.). tion of old Greek treatises on building design he probably Athens, 2000, 135–175. found the idea of the sounding vessels so fascinating, that [4] M. Bieber: The history of the Greek and Roman theater. he decided to promote it. However, as concluded in sec- Second edition. Princeton University Press, 1961, 1971. tion 3, Vitruvius wasnot up-to-date in his book on theatre [5] G. C. Izenour: Roofed theaters of classical antiquity.Yale design, and therefore it is most likely that the architects University Press, NewHaven and London, 1992. responsible for the building of Roman theatres, odea and [6] K. E. Welch: The roman amphitheatre, from its origins to amphitheatres have simply ignored Vitruvius’ advice con- the Colosseum. Cambridge University Press, NewYork, cerning the sounding vessels. 2007. [7] F. Corni: Aosta Antica –Aoste Antique, la città romana – La cite romaine (inItalian, French translation by A. Perrin). 7. Conclusion TipigrafiaValdostana, Aosta, Italy,2004. [8] ERATO,Final report. INCO-MED project ICA3-CT-2002- The results from the ERATO project of the acoustical sim- 10031, 2006. ulations in the reconstructed theatres and odea confirm [9] A. Farnetani: Investigations on the acoustics of ancient the- the assumption, that theywere dedicated for different pur- atres by means of modern technologies. PhD Thesis, Fac- ulty of Engineering, University of Ferrara, Italy,2006. poses. The theatres with very high clarity of sound were [10] G. Govoni: Sviluppo di strumentazione per misure acus- excellent for speech, whereas the odea with ahigher sound tiche su modelli in scala. Master Thesis. Faculty of Engi- strength and more reverberant sound were excellent for neering, University of Ferrara, Italy,2004. song and music from weaker instruments likethe lyre or [11] F. Bettarello: Progetto, realizzazione emisure acustiche su kithara. More information can be found in the final project modello in scala del teatro antico di Siracusa, nell’ambito report [8] and in the proceedings of the ERATO project del progetto di ricerca ERATO.Master Thesis, Faculty of symposium [16]. Detailed studies on the acoustics of an- Engineering, University of Ferrara, Italy,2004. cient theatres are described in aPhD thesis by Farnetani [12] L. Dietsch, W. Kraak: Ein objektivesKriterium zur Erfas- [9]. sung vonEchostörungen bei Musik- und Sprachdarbietun- gen. Acustica 60 (1986)205–216. Echo problems can occur in certain places in the ancient theatres, particularly in the orchestra area due to the focus- [13] A. Barba Sevillano, R. Lacatis, A. Giménez, J. Romero: Acoustics vases in ancient theatres: disposition, analysis ing effect of the concave shaped steps of seats. In this re- from the ancient tetracordal musical system. Proc. Acous- spect the diazoma is particularly important because of the tics’08, Paris, 2008, 4155–4160. higher wall, and this is precisely where the sounding ves- [14] G. C. Izenour: Theater design. McGraw-Hill, NewYork, sels described by Vitruvius should be installed. It appears 1977. that the idea and the guidelines for the sounding vessels go [15] P. V. Brüel: Episodes and achievements within Acous- back to Aristoxinus in the 4th century BC, in those days tics before 1954 (inDanish). Danish Acoustical Society, afamous scholar in music theory.From the description Copenhagen, 2005. givenbyVitruvius, and the collection of information from [16] Audio visual conservation of the architectural spaces in vir- excavated theatres where the vessels might have been in- tual environment: Proceedings of the ERATO project sym- stalled, it is concluded that the sounding vessels could not posium. Yildiz Technical University,Istanbul, 2006.