Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Addis Bill Layouted

Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

Proceedings of the Third International Congress on Construction History, Cottbus, May 2009

A Brief History of Design Methods for Building Acoustics

Bill Addis
Buro Happold, London, UK

ABSTRACT: Acoustics has been important to the designers of buildings, especially theatres and concert halls,
for at least 2500 years. For most of this time, designers used empirical guidance which, while reliable within its
limitations, could not be applied successfully to rooms and auditoria which had no close precedent. The sci-
ence of acoustics developed in the 18th and 19th centuries, mainly in connection with musical instruments.
However, design methods for room acoustics in buildings become scientific and quantitative only in the 20th
century. The physicist Wallace Sabine discovered what affects the reverberation time of a room around 1895
and used this, first, to improve the acoustics of a lecture theatre and, later, in the design of new rooms – most
famously, the new Boston Music Hall in 1900. The measurement of sound and waveforms became practical
with the development of microphones, amplifiers, oscillosopes and early sound recording machines in the
1920s. From the 1930s physical scale models have been used to measure the acoustic response of auditoria.
They are still used today, together with increasingly sophisticated analysis of waveforms, to build up more and
more reliable predictions of acoustic performance. Architectural acoustic design methods are typical of other
engineering design disciplines in how they developed from being purely qualitative, then using empirical
data, then making use of physical model testing, and finally using comprehensive mathematical models.

INTRODUCTION
The historical development of architectural acoustics is similar to other fields of building design, in comprising
two parallel strands of ideas – the science and mathematics of the subject on the one hand, leading to im-
proved understanding of the phenomena, and the methods used by designers when faced with the chal-
lenge of a new building on the other, especially when the task differs markedly from precedent. The two nine-
teenth-century classic works on the physics of acoustics (Helmholz 1863; Strutt 1877-78) hardly mentioned the
acoustics of theatres or other rooms, and the science they contained only began to be used by architectural
acousticians in the mid-twentieth century. While these two branches of knowledge are closely related, it was
not a case of theory leading to practice, or vice versa: the two were symbiotic.

ACOUSTICS IN THE ANCIENT WORLD

Vitruvius on acoustics and theatre design


The Roman engineer Vitruvius devoted several chapters of his book on building design and construction to the
location and design of theatres (Vitruvius, Book V). He advised that they should be located away from winds
and from “marshy districts and other unwholesome quarters” and also on their orientation with respect to the
sun and the surrounding terrain. He addressed key geometric issues such as the plan and section, sight lines,
numbers and locations of entrances and exits, and finally considered the subject of acoustics. This highly theo-
retical section was not his own; he was repeating what he found in various Greek treatises on acoustics from
two or three centuries earlier which, in turn, probably had their origins in Pythagoras who first developed the
subject around 530 BC. Vitruvius dealt with acoustics from several points of view. First he introduced harmonics
– “ an obscure and difficult branch of musical science, especially for those who do not know Greek”. This sci-
ence explained the pitch of notes and the intervals between them in the Greek musical scale, as well as why
some combinations of notes are concordant and others discordant. Next Vitruvius discussed sound in the audi-
torium – in particular the need for sound of all pitches to travel from the stage to the ears of every member of
the audience by a direct route, in the manner of waves created by a pebble thrown into water. This led logi-
cally to both raked seating and the semi-circular plan. He advised against vertical reflective surfaces that
Proceedings of the Third International Congress on Construction History, May 2009

would prevent sound reaching the upper tiers of seats since this particularly impairs the intelligibility of word
endings which, in Greek and Latin, are vital to comprehension. Such reflected waves, he wrote, can also inter-
fere with the direct waves and distort sounds for the listener. These explanations differ remarkably little from
how we would put it today. Thirdly, Vitruvius explained that the site of a theatre itself must be carefully selected
taking account of acoustics: it must not have an echo, nor give reflections that can lead to direct (incident)
and reflected sounds interfering.
Vitruvius also discusses the use of sounding vessels – nowadays called Helmholz resonators, after the nine-
teenth-century German physicist who explained how they function – which, he says, reinforce certain fre-
quencies of the human voice and can increase intelligibility. These open-ended vessels were made of bronze
and tuned to six notes of the chromatic scale. Two sets of six were arranged beneath a tier of seats symmetri-
cally either side of the centre line of the theatre. If the theatre were particularly large, two additional sets of
vessels should be installed in higher rows, each a few semi tones lower in pitch – a total of thirty six different
notes. Vitruvius admits he knows of no theatres that had actually been built in Rome with sounding vessels. The
reason, he explains, is that “the many theatres that are constructed in Rome every year contain a good deal
of wood which does not lead to the same problems with reflections as stone”. Also, he says, the timber panels
themselves can resonate in a manner similar to the air in a sounding vessel and so improve intelligibility.
As to the effectiveness of sounding vessels, they are known today not to improve intelligibility and that is
probably why they were not used in Rome. Whether the Roman theatres were as good as the Greek ones, we
do not know, but there is no doubt that both were designed with great understanding of acoustics and exper-
tise in using this understanding to achieve demonstrably better results.
One final recommendation from Vitruvius on acoustics was for a senate house. The height of a senate house
should be half the width of the building, he says, and coronae, or cornices, made of woodwork or stucco,
should be fixed half way up the inside faces of the walls around the entire room. Without these, he says, the
voices of men engaged in discourse are lost in the high roof. With coronae, the sound of the voices is ‘de-
tained before rising’ and so is more intelligible to the ear.

Acoustics in the mediaeval and renaissance eras


No significant writings on the acoustics of buildings survive from mediaeval or Renaissance times (Hunt 1978).
Vitruvius was published in the late 15th century and would have been known by most designers of large build-
ings. However, it is not possible to identify precise ways in which his guidance was followed, either in cathe-
drals or, from the late Renaissance, in theatres. The development of music from the 12th century provides evi-
dence of a good understanding of the acoustics of cathedrals, however, their legendary acoustic qualities
are more indebted to the skill of composers and musicians than to the buildings themselves or their designers.
They have long reverberation times because sound waves are reflected many times with little loss of intensity
which means that musical rhythms have to be slow to be intelligible, and percussive instruments must not be
used to avoid the inevitable machinegun effect of any echo. The acoustic of the space favours those instru-
ments with a gradual attack to each note, and which sustain their notes – for example the organ, flute, violin
and the human voice. For speech, however, the long reverberation time is a disaster. As the distance between
speaker and listener increases, so the sound reaching the ear directly is increasingly swamped by the reflected
sounds arriving by indirect, longer sound paths. Speech is thus generally unintelligible at any distance greater
than a few metres, which phenomenon has an interesting architectural effect. Since it is the longer wave-
lengths of lower notes that are more effectively reflected, people talking in cathedrals are naturally and un-
consciously inclined to whisper, irrespective of any reverence for the religious nature of the buildings they may
feel, because whispering removes the lower frequencies of the human voice.
The use of sounding vessels in a theatre with raked seating (as described by Vitruvius) is illustrated in a drawing
by Francesco di Giorgio di Martini in around 1485 (Maltese 1967, Vol.1, Plate 13) although he does not say if
they were used. However, such vessels made of pottery have been found in some mediaeval monasteries, in-
cluding Fountains Abbey in England (Arns; Crawford 1995).

EARLY MODERN DESIGN GUIDANCE – EIGHTEENTH CENTURY


As in the ancient world of building described by Vitruvius, it was the intelligibility of speech that drew the atten-
tion of 17th and 18th century building designers to the acoustic performance of building interiors, especially in
two types of building – theatres and the debating chambers used by politicians. During the eighteenth century
the importance of room acoustics was further heightened with the invention of a number of musical instru-
ments such as the harpsichord and fortepiano, and the growing popularity, in elite circles at least, of chamber
music. The new instruments used ingenious mechanisms and large sounding boards to produce plucked and
percussive notes with unprecedented speed and at much greater volumes than earlier instruments such as the
lute, harp and clavichord. When played in a room with a very live acoustic, the individual notes became indis-
tinguishable and the objectives of the instrument makers and musicians were ruined.
It was already well-understood that the size of a room and the reflectivity of the walls, floor and ceiling af-
fected the intelligibility of speech and music, and that two different effects were at work. One was the dimin-
ishing loudness or intensity of the sound, both with distance and according to the amount of sound absorbed
by the room’s surfaces and contents. The second was the increasing confusion of sounds caused by reflections
from the room’s surfaces.
Proceedings of the Third International Congress on Construction History, May 2009

This understanding of acoustics led building designers to use a number of pragmatic rules that helped them
achieve acceptable room acoustics. Three different types of surface could be used – stone or plaster would
enhance reflections, woven fabric such as tapestries and curtains would absorb sound, and timber panelling
was intermediate between the two. Based on the acoustic performance of existing rooms, a designer could
choose what he hoped would be a suitable combination of the three types of surface. The other design factor
was the distance between speaker or instrument, and the directness of the sound path. Theatre designers tried
to ensure that that the entire audience could see the actors, not only for theatrical effect, but so that at least
some of the sound could travel directly from speaker to listener. The other consideration was the distance be-
tween the stage and the listener. It was generally agreed that an actor speaking in a normal voice could be
understood clearly up to a distance of about 18 metres, and with difficulty up to about 25 metres. The result of
these basic rules was the development of the familiar raked seating and tiers of balconies in theatres and
dedicated concert halls.
The need for good acoustics was, however, not the only influence on the design of theatres. There was also a
tension between the need for actors to be understood, on the one hand, and the patron’s desire to increase
the size of audience and the income from ticket sales, on the other. Many theatres in the late eighteenth cen-
tury were built with distances between actor and listener more than thirty metres. This forced actors to change
their style of delivery and develop the skill of speaking more loudly, while sounding as if speaking in a normal
voice – a good actor could even fill an auditorium with what sounded like a whisper.
Designers of theatres intended for performances of both music and the spoken word faced a further, insuper-
able challenge since the acoustic characteristics needed for good intelligibility of speech make music sound
very dry and unpleasant to listen to. A hall well-suited to music, on the other hand, rendered the spoken word
almost unintelligible in all but the seats nearest the stage. The most practical solution to this dilemma was to
have different buildings dedicated to each form of entertainment. Theatre, however, often relied on music to
enhance the dramatic effect and orchestra pits were incorporated in most large theatres between the front
seats and the stage. The acoustics of such theatres was inevitably a matter of compromise, favouring the intel-
ligibility of the spoken word.
Design rules for theatre acoustics were not very reliable in the eighteenth century and there were many acous-
tic disasters. In his autobiography of 1740, the actor and playwright Colley Cibber wrote of Vanbrugh’s
Queen’s Theatre in London, built in 1704-05, that all its architectural elegance was of no avail:
. . . when scarce one word in ten could be distinctly heard . . . The extraordinary and superfluous space oc-
casioned such an undulation from the voice of every actor that generally what they said sounded like the
gabbling of so many people in the lofty aisles in a cathedral … [and] the articulate[d] sounds of a speaking
voice were drowned by the hollow reverberations of one word upon another. (Leacroft 1984, p.103)
Throughout Europe the second half of the eighteenth century saw a boom in theatre building in the major cit-
ies, and designers generally learned from the acoustic disasters of the early century. By the late eighteenth
century it was common practice to use the ceiling or soffit above the front of the stage as a ‘sounding board’
(actually a reflector) and the ceiling over the orchestra pit to ‘throw the voice forward’ from the stage to the
back of the stalls and to the galleries. The first design guides for theatres discussed acoustics alongside the
equally important issue of line-of-sight (Patte 1782, Saunders 1790, Rhode 1800, Langhans 1810). These and
others followed Patte’s example in showing ray diagrams to visualise sound paths (Fig. 1).

Figure 1: Ray diagrams for different theatre plans; (Patte 1782, Plate 1)

Of these authors it was the architect George Saunders (c.1762-1839) who was the first to publish results of tests
he had undertaken on theatre acoustics. He was familiar with Patte’s guide and also with the first modern sci-
entific work on acoustics by the French mathematician and music theorist Marin Mersenne (Mersenne 1638)
whose scientific approach he adopted:
Phonics or the doctrine of sounds is particularly distinguished from acoustics or the doctrine of hearing. In-
stead of dividing into direct, refracted and reflected, I shall, for the sake of brevity and perspicuity, divide
phonics into three distinct heads; namely formation of sound, combination of sound, and progression of
sound. . . . . But as sound is very much influenced or altered by the bodies it meets with, and the form of its
expansion depends much on the manner of its being transmitted, it is necessarily our business to enquire,
how it is affected by the different bodies it may meet with in its progress, and more particularly, of the man-
ner in which the voice expands. In examining what has already been written upon these subjects, very little
Proceedings of the Third International Congress on Construction History, May 2009

could be gained to our purpose; and this occasioned the following experiments, which may help to give
this part of science an additional degree of certainty. (Saunders 1790, pp.2-3)
Saunders goes on to describe how he tested seven “propositions” in real auditoria:
1 Of the extension of the voice on a plane;
2 Of the ascension, descension, and cubical form of sound;
3 Does the voice act upon a certain quantity of air of whatever form to which it may be confined?
4 Of screens partially placed before the voice, and opposing angles;
5 How sound operates in different airs, and how much it is affected by currents of air;
6 On the reflection of sound;
7 Of the property of different materials to alter and conduct sound. (Saunders 1790, pp.4-22)
And from the results of these experiments he concludes they lead to several “applications”:
First, that sound expands equally in every direction. Secondly, that to alter the form of its expansion, the in-
tervention of a body is necessary. Thirdly, that all bodies attract sound. Fourthly, that sound is absorbed, and
conducted by a body, more or less according to the nature of the material. Fifthly, that in proportion to the
conducting power of the material, will be the resonance it occasions. Which being admitted, it follows, that
nothing can be depended on, in a theatre, but the direct force of the voice. (Saunders 1790, p.23)
Saunders summarised his experience in a number of clear rules:
– the hearer should never be more than twenty two metres (seventy feet ) from the speaker;
– the ceiling should be used to convey sound to the upper seats of the auditorium
– the circular plan is best, and
– the entire audience should have sight of the stage.
Writing about the design of theatres Saunders uses the concept of ‘phonics’ (i.e. acoustics), for example, criti-
cising theatres that had deep boxes and galleries, as their low height ‘obstruct[ed] the sound … [and] the little
that enters is presently attracted and absorbed by the persons, clothes, etc. of the spectators in the foremost
rows’. Saunders conducted many experiments on the effect of the shape of theatre auditoria on intelligibility,
comparing the results to those obtained in the open air, and concluded that the oval or horseshoe form were
best, but that these did not allow good views of the stage. He concluded the semi-circular form was best, ‘with
its centre seventeen feet (five metres) in front of the speaker’ and the diameter of the circle should be no
greater than sixty feet (eighteen metres). Saunders was also aware that the voice did not carry well above an
angle of 45° to the horizontal, and thus the height of the auditorium should be no greater than three quarters
of the diameter of the circle. Concerning materials, he wrote that:
Wood is sonorous, conductive and produces a pleasing tone, and is therefore the very best material for lin-
ing a theatre; for not absorbing so much as some, and not conducting so much as others, this medium ren-
ders it peculiarly suitable to rooms for musical purposes; the little resonance it occasions being rather
agreeable than injurious. (Saunders 1790, p.201-2)
Saunders used his understanding of acoustics to produce designs for both an ‘Ideal Theatre’ and a similar op-
era house which also incorporated the latest ideas on protection against fire, including staircases constructed
entirely of stone and enclosed by walls.
The first modern comprehensive book on acoustics (Chladni 1802) was written by the German physicist Ernst
Chladni (1756 –1827) who studied how musical instruments create their characteristic sounds. His experiments
led him to study both the vibration of sheets of wood and metal and also how such vibrations were conveyed
through the air to the listener’s ear. He published the results of his experimental work on vibrating plates first in
“Entdeckungen über die Theorie des Klanges” (Discoveries concerning the theory of sound) in Leipzig in 1787,
and then in “Die Akustik” in 1802. Apart from the physics of sound, Chladni included some useful guidance on
designing auditoria, though drawn entirely from the practical guidance of both Saunders and Rhode (Saun-
ders 1790; Rhode 1800).
Chladni illustrates one intriguing design for a theatre in which the walls are made of panels that can be ro-
tated about their vertical axis to change their angle and, hence, the direction of reflected sound and also to
allow some sound to penetrate into the cavity behind thus reducing the intensity of reflections; he does not
say if the theatre was built.

EARLY MODERN DESIGN GUIDANCE – NINETEENTH CENTURY


Saunders’ and Chladni’s design rules were widely repeated and used for most of the 19th century, for example
when designing the growing number of music auditoria, theatres, lecture theatres in the many new polytech-
nic colleges being built throughout Europe, and also the chambers in which politicians and legislators congre-
gated, both of which also needed good acoustics. The job of ensuring good acoustics naturally fell to the de-
signers of the heating and ventilation systems since they all relied on the common medium of the air. It was
also often the case that acoustic performance could be adversely affected by noise born into the room from
outside through the ducts providing fresh air.
Proceedings of the Third International Congress on Construction History, May 2009

W.S Inman and David Reid


The destruction by fire, in 1834, of the House of Commons, where English members of Parliament congregated,
required the design and construction of a new chamber. As is often the case with new government buildings,
this provided an excellent opportunity for a state-of-the-art building. A study was commissioned to establish
the best form of room to ensure excellent intelligibility throughout. The resulting reports, papers and books, by
W. S. Inman and David Reid, collected not only contemporary experience and published guidance on theatre
design such as that written by Saunders and Chladni, but even referred to the acoustic science and building
designs from ancient Greece, as described in Vitruvius. In the book based on his studies Inman summarised
Chladni’s design guidance from a third of a century earlier:
. . . rooms will be favourable to the transmission of sound:
1. When arranged to facilitate its natural progress;
2. When its intensity is augmented by resonance or simultaneous reflection, so that the reaction is undistin-
guishable from the primitive sound;
3. When not too lofty or too vaulted;
4. When there is not a too extensive surface for the sound to strike against at once
5. When the seats are successively elevated.
He [also] observes that:
- when the enclosed space does not exceed 65 feet, any form may be adopted for a room;
- that elliptical, circular, and semi-circular plans produce prolonged reverberations;
- parabolic plans and ceilings are the best for distinct hearing, and that
- for concert rooms, square and polygonal plans should have pyramidal ceilings, and circular plans domed
ones, and the orchestra be placed on high, in the centre, to produce the best effect and avoid echo.
(Inman 1836, p.21)
David Reid described the work he undertook in his studies for the House of Commons in his book on ventilation
(Reid 1844; Addis 2007, pp.286-88). He approached his work in a manner little different from that of a modern
consulting engineer. Concerning the poor audibility of speakers in the House of Commons, he first established
the causes of the problems by interviewing members and by direct observation in the chamber itself. He iden-
tified ‘six principal causes of the defective communication of sound’. The ceiling was too high which meant
the voice had a large volume to fill, a particular problem when a speaker was competing with a general
background hum of conversation. The soft fabrics covering the walls of the chamber provided insufficient re-
flection of incident sounds; the noise entering the chamber through the windows, especially when opened for
ventilation, was ‘extreme’. Not only was there noise from the adjacent chamber, the House of Peers, but also
‘the noise of coaches, cabs and omnibuses, of the letter-carrier’s bell and every other noise produced in the
Old Palace Yard. He found that the floor covering in the chamber produced a considerable noise when
members walked upon it and there was also disturbing noise produced by the opening and closing of the
door to the room. Finally, he believed that the rising current of warm (and less-dense) air in the centre of the
chamber refracted the sound passing from one side of the room to the other, just as a glass prism refracts light.
In fact, while theoretically possible, it unlikely that Reid would have been able to observe such an effect; more
likely, the variable intelligibility depending on the listener’s position in the room was due to the presence or ab-
sence of reflections from walls and ceiling.
Based on his observations and explanations, Reid incorporated various acoustic devices in the new House of
Commons chamber built in the late 1830s. The ceiling was lowered and panels of wood and glass were incor-
porated in the upper parts of the room to reflect sound downwards and towards the centre of the room. Both
the floor and ceiling were porous to allow fresh air in and foul air to be extracted, and these surfaces also re-
duced the reflected sound that had previously led to the ‘offensive reverberation’ that reduced intelligibility.
The new heating system kept the air temperature the same throughout the chamber which prevented the re-
fraction of sound. The floor of the chamber was covered with a ‘soft, thick, porous and elastic hair-cloth car-
pet’ which prevented noise caused by people walking, which also absorbed unwanted reflections in the
room. Finally, the new and highly effective heating and ventilation system meant it was no longer necessary to
open windows and thus disturbance from noises outside the chamber was avoided.
After his work on the new House of Commons, Reid went on to collect information and conduct many experi-
ments on the acoustics of rooms and reported his results and guidance for designers in his book of 1844. He
noted particularly the importance of reverberation time and its effect on the intelligibility of speech. He meas-
ured an eight second reverberation time inside an iron boiler and mentioned several large rooms he had vis-
ited where a sound remained audible for between eight and ten seconds; he had even heard of a room in a
palace in St Petersburg where the noise made by stamping a foot was audible for a period of twelve seconds.
Reid realised that the key to intelligibility of speech depended on the ratio of the loudness of direct and re-
flected sound reaching the hearer. This was well exemplified by the benefit that the presence of an audience
would bring to a room by reducing reflected sounds, both by absorption of sound in their cloths and in pre-
venting sound being reflected from a hard floor. He was also aware that speakers could be disturbed by loud
reflections of their voice reaching them from the rear wall of a room. He mentioned many examples of how
the acoustics of rooms had been improved by applying the rationale he described to reduce disturbing re-
verberation. In one large lecture room the plaster ceiling was removed, leaving the bare beams and laths. In
other cases drapery, curtains, carpets, sofas and soft furniture were used to absorb sound. Reid also carried
Proceedings of the Third International Congress on Construction History, May 2009

out numerous experiments on the focusing effect of arched or domed ceilings, especially in Paris at the Halle
au blé and in the many rooms of the Louvre palace. In one room in Paris he found that the reflection from
such a ceiling had been so strong as to prevent it being used; the problem was remedied by suspending a silk
balloon in the centre of the ceiling.
Reid’s qualitative understanding of acoustics differed little from that of twentieth century acousticians. He was
prevented from taking acoustics into the realm of quantitative engineering only by the lack of equipment to
measure sound intensities and to analyse sound of different frequencies. This step awaited the invention of the
microphone, in the late nineteenth century, which converted sound waves into a varying electric current, and
the measuring of both relative and absolute sound intensities in the early twentieth century. Based on the work
of Inman and Reid, further useful design guidance began to appear (Lachéz 1848; Upham 1853; Smith 1861).

The late nineteenth century


The late nineteenth century saw the construction of new concert halls to meet the growing popularity of con-
certs featuring symphony orchestras. These halls differed markedly from the three traditional venues for musical
performance – churches, the salons of palaces in which chamber music was played, and theatres and opera
houses. The size and construction material of churches were guaranteed to give them long reverberations
times (up to six or seven seconds) which constrained the type of music that could be enjoyed. Multiple reflec-
tions from the hard surfaces of walls and floor forced composers and musicians to avoid rapid rhythms and
staccato or percussive sounds which would degenerate into confusion as new notes interfered with the reflec-
tions of earlier ones, both in pitch and rhythm. Chamber music, usually featuring fewer than a score of musi-
cians, developed during the eighteenth and early nineteenth centuries in the houses of the rich and was sel-
dom intended for large audiences. The small sizes of the rooms and the plastered interiors gave a much
shorter reverberation time (perhaps three or four seconds) and a lively room acoustic that so well suits the Ba-
roque and early classical music of Handel and Haydn. Such rooms were ideal for the relatively quiet instru-
ments such as the violin family, woodwind and early keyboard instruments such as the harpsichord.
Theatres and opera houses, on the other hand, were dedicated mainly to the human voice. Their design was
driven by the need to ensure intelligibility and to avoid the need for actors to have to speak unduly loudly. The
maximum distance between the speaker or singer and the listener should be no more than about twenty me-
tres and the interior of the auditorium was covered in fabric and soft furnishings that ensured a low reverbera-
tion time (between two and three seconds). The symphony orchestra, by contrast, playing the music of Bee-
thoven or Mahler, could be very loud indeed; and the promoters of public symphony concerts wanted to seat
as many listeners as possible. Both features called for much larger halls. In many places the large assembly
rooms in town and city halls often served, and still do serve, as acceptable music venues, although their multi-
purpose function often meant they were seldom ideally suited to any one purpose. The acoustic design of
such halls was entirely empirical and based on the observed characteristics of previous halls that had proved
successful, though it would not have been understood precisely what had lead to past successes. A number of
halls have attracted widespread acclaim among musicians and concert-going audiences – for example, Vi-
enna’s Grosser Musikvereinssaal (1870), the Leipzig Gewandhaus (1885), and the Conzertgebouw in Amster-
dam (1895). Yet, until recent times, when a new hall was designed there was little agreement as to precisely
which features of an excellent hall should be copied in a new one. Acoustic disasters were not uncommon,
and were no better exemplified than at London’s Albert Hall, opened in 1871, whose prominent echo gained it
the reputation for giving audiences good value for money as they heard every concert twice.

WALLACE CLEMENT SABINE – “FATHER OF ARCHITECTURAL ACOUSTICS”


The man who has no rival in being called “the father of architectural acoustics” was Wallace Clement Sabine
(1868-1919) (Sabine 1922; Beyer 1999, pp.186-191; Thompson 1992, 2002). Sabine was a lecturer in physics in the
department of natural philosophy at Harvard University and was approached in 1895 to advise on how to im-
prove the poor acoustics of a new lecture theatre in the University’s Fogg Art Museum. This lecture theatre had
been designed to emulate a classical Greek theatre and followed the same principles of acoustic design that
Vitruvius had written down. These addressed the need of intelligibility by focusing mainly on maintaining the
volume of the direct sound that reached the listener’s ear. The speaker was placed above the level of the
front row of the audience; the seating was raked upwards towards back of the auditorium; and a wall was
placed behind the speaker to reflect sound into the auditorium. However, such principles were intended for
open-air theatres and took no account of sound reflected from walls or the roof. In an enclosed room these re-
flected sounds also reach the listener’s ear and, since there will be many sounds, arriving at different times, the
result is confusion with direct sound from a speaker competing with reflections of earlier sounds.
Sabine realised this was how intelligibility was lost, like many before him. Being a physicist, however, his ap-
proach was to conduct experiments to measure how the loudness of the reflections was influenced by the re-
flecting surfaces in the lecture theatre. His aim was to discover the relationship between the dimensions of the
room and the rate at which a sound became quieter and eventually became inaudible. He called this rate of
decay the reverberation time and defined it as the time, in seconds, for a sound to decay to one millionth of
its original loudness (a fall of 60dB).
Sabine had to work at night to ensure all extraneous sounds were avoided. He used a single organ pipe with a
frequency of 512 Herz (an octave above middle C). In 1895 there were no microphones or audio-electronics
Proceedings of the Third International Congress on Construction History, May 2009

and the judgement as to when the sound was inaudible was made by the experimenter himself. An electric
chronograph recorded the times to one-hundredth of a second. By covering more and more of the audito-
rium’s wooden seats with soft cushions, he showed that the reverberation time was inversely proportional to
the number of seats covered with cushions. He repeated the experiments in eleven other rooms in the univer-
sity, with volumes ranging from a lecture theatre of 9300 cubic metres down to an office of just 35 cubic me-
tres. From the results he derived the equation for which his name is well-known giving the relationship between
the reverberation time (RT), of a room, in seconds, its volume (V), in cubic metres, and the area (A), in square
metres, of sound-absorbing surfaces in the room.

V
RT = 0.163 (1)
A
Sabine used this equation to give an objective means of comparing different auditoria and, in particular, to
compare the proposed design for the new Boston Music Hall with the Leipzig Gewandhaus, on which its overall
shape was based, and the old Music Hall in Boston. He was able to specify, for the first time, the precise de-
gree of sound absorption in the interior of the new Boston hall needed to achieve the same reverberation time
as the Leipzig Gewandhaus whose seating capacity it exceeded by 70%, and volume by 40%. Sabine’s pre-
dictions were accurate and the acoustic of the new hall was widely praised. He had fulfilled his goal of over-
coming the “unwarranted mysticism” that then surrounded the subject of architectural acoustics and, most
importantly, achieved “the calculation of reverberation in advance of construction”.
Sabine was soon being approached by the owners of various types of room to advise on how to rectify their
acoustic problems. Often this followed the failed attempts by others to deal with the problems. Sabine noted
the persistent use of a traditional but wholly ineffective remedy which involved stretching a grid of steel wires
in the top of a church, theatre or court room which suffered too much reverberation on the mistaken believe
that the wires would resonate and absorb sound. In New York and Boston he had seen theatres and churches
with just four or five wires stretched across the room while in other auditoria several miles of wire had been
used, all without the slightest effect.
As part of his diagnosis of acoustic problems he would sometimes plot a contour map showing the distribution
of the sound intensity. This helped him identify the source of the worst sound reflections from the walls and ceil-
ing and hence reduce them by using sound-absorbing panels or adding decorations that would break up
strong reflections from large plane surfaces.
Sabine also turned his attention to the design of new theatres and how best to create a near-uniform acoustic
experience for every member of the audience. To help him in these studies he used the newly-perfected
schlieren method of photography to show sound waves passing through air in two-dimensional models of au-
ditoria (Fig.2). He was thus able to show in plan and section, how sound waves were reflected and broken up
as they emanated from the stage into the auditorium. Outside the field of building structures this was probably
the first use of a scale model to investigate the engineering behaviour of a building.

Figure 2: Photographs showing the progress of sound waves through a model of a theatre; (Sabine 1922)

THE PERSISTENCE OF THE TRADITIONAL DESIGN METHODS

Design methods in the 1920s


When designing recording studios, auditoria and concert halls in the 1920s and 1930s, the reverberation time
could be predicted using Sabine’s method, but otherwise the acoustics was addressed largely as an empirical
process based on experience. Good common-sense design guidance available since the 1830s served de-
signers moderately well and was usually adequate as long as concert halls were built in traditional ways. In the
1920s, however, architects began experimenting with new shapes of auditoria to which the familiar acoustic
design rules did not apply. It became necessary to think about their acoustics from first principles, considering
both the different paths by which sound might reach the listener in the auditorium, and the correct balance
between direct sound and sound reflected off the walls or ceiling. The French acoustician Gustave Lyon was
one of the pioneers of this approach at the Salle Pleyel in Paris, completed in 1927 who made much use of
ray-tracing, the method first illustrated by Patte in 1782 (Fig.1). The architect Victor Horta also used ray-tracing
when designing the Salle Henry-le-Boeuf in Brussels in 1928 (Fig.3). Despite his client’s wishes, however, Horta re-
fused to consult Lyon, then considered to be the leader of the French school of acoustic design, because he
Proceedings of the Third International Congress on Construction History, May 2009

considered Lyon’s purely scientific approach paid insufficient attention to how music actually sounded in
concert halls. Horta visited a great many halls in Europe and the USA.

Figure 3: Ray diagrams for Salle Henri-le-Boeuf, Brussels, by Victor Horta, 1928; (Archives of Horta Museum)

In the early 1930s some German acousticians used rays of light in three-dimensional models of theatres to
study the path taken by sound waves, but such models gave no help in investigating different frequencies of
sound, the reverberation time or the different times that sound would take to reach a listener by different
sound paths. Acoustic design was still largely a subjective art at this time.

Design methods in the 1950s


The science of acoustics received an unexpected boost during the Second World War when sound was the
primary means for locating the sites of enemy guns and, before the development of radar, in detecting the
approach of enemy aircraft. Sound is transmitted even further in water than in air, and underwater micro-
phones were used to detect the engines of ships and submarines. There was also the development of sonar,
an underwater version of radar which used the reflections of sound to locate underwater objects such as
mines and vessels and also to build up a 3-D map of the sea bed. After the War, many of these acoustics ex-
perts became consultants advising building designers on acoustics. Nevertheless, the war-time developments
in acoustics did not transform building design overnight. When the Royal Festival Hall in London was designed
in 1949-50 the main acoustic consultant was ‘one of the old school’ who based his advice on his 30 or so years
of experience in building design and discussions with musicians, conductors, music critics, music lovers and
other acoustics experts, rather than on the latest acoustic science. Some fifty years after Sabine’s work, still the
only measurable and (relatively) predictable design parameter was the reverberation time. Nevertheless con-
sultants still differed in their recommendations for the most desirable value for the Festival Hall – one said 2.2
seconds, the other 1.7 seconds, a very large difference. 1.7 seconds was chosen but, as it turned out, the re-
verberation time of the finished hall was found to be just 1.5 seconds, which most people agreed was too
short. Acoustics was still far from being a precise science, and assessment was subjective - some musicians and
conductors liked the acoustic, while others did not. In its early days a great many studies were undertaken in
the Festival Hall to find ways of lengthening the reverberation, but no solution was found. Following further
studies in the early 1960s (it was said that the Royal Festival Hall was the most studied hall in the history of
acoustics) when the hall was being refurbished, a number of microphones and loudspeakers were installed to
provide ‘assisted resonance’ which increased the reverberation time to the desired value of 1.7 seconds, and
the acoustic of the refurbished hall met with widespread acclaim. Taste in the world of music can change,
however, and, after many discussions and several trials, it was finally agreed in 1995 to stop using the assisted
resonance, since when the acoustics of the hall is still acclaimed.

THE USE OF PHYSICAL MODELS IN ACOUSTIC DESIGN


After Sabine’s pioneering work, the most far-reaching progress in acoustic design methods came with the use
of scale models. Models have been used to aid the designers of buildings for many millennia, however, their
use entered the modern scientific age only in the 1860s. The English engineer William Froude studied the per-
formance of scale models of ships and used non-dimensional constants to adapt the results and predict the
performance of full-size ships, even when the parameters varied non-linearly with scale (Addis 2007 p.479-80).
Scale models were first used to measure and analyse the acoustic performance of auditoria by the German
physicist Friedrich Spandök in the early 1930s (Spandök 1934) (Baron 1983). Using simple dimensional analysis to
find suitable dimensionless constants, he showed that the acoustic behaviour of a room varies inversely with
the model scale – for a one-fifth scale model, the sound frequencies used for testing need to be five times
higher than normal frequencies. He also recognised the need to ensure that the temperature, pressure and
density of the air in the model room were identical to a real room, since these affect the speed of different
frequencies of sound differently. Spandök’s first concern was to demonstrate that tests on a model were in-
deed a reliable representation of the acoustic behaviour of a full-size room. He then studied the effect on
both the decay of sound and its distribution throughout an auditorium with raked seating (better) and with a
Proceedings of the Third International Congress on Construction History, May 2009

semi-circular wall behind the stage (worse). Using a microphone he displayed the decay of sound in the
model auditorium on the screen of an oscilloscope, and recorded the results in a photograph. Although similar
studies were undertaken in the late 1930s in a number of other university physics departments, which all dem-
onstrated the possibility of using models during the design of an auditorium, the technique did not gain wide-
spread use by building designers until after the war.
The Danish acoustician Vilhelm Jordan had been one of the pioneers in using scale models in acoustic design
In the 1930s using a wire recorder to store the results and allow them to be analysed. He was selected in the
mid-1960s to advise the designers of the Sydney Opera House on how to achieve a satisfactory acoustic for
the many different uses – including concerts, opera and speech – that were proposed for the auditoria. Since
these different uses require different acoustic characteristics, especially reverberation time, one early plan was
to provide a large moveable ceiling that could be adjusted to create a reverberation time to suit each need.
So different from established practice were the proposals for the auditoria that Jordan advised undertaking
acoustic tests using scale models. Over five or six years models of several different proposals for the two audi-
toria were made at 1-to-10 scale. Made of wood, they included models of the audience with bodies made
from neoprene and heads of cardboard. The tests soon established that it would not be possible to create an
acoustic that could serve the original idea of multi-purpose auditoria. The curved form of the walls also caused
problems in achieving a satisfactory distribution of reflected sound in all parts of the auditoria.

Figure 4: 1/10 scale wooden models for studying the acoustic performance of (a) early, (b) intermediate, and
(c) late interior designs for the Sydney Opera House. The figure in (c) is V.L. Jordan; (V.L. Jordan)

The model tests demonstrated the benefits of introducing flat side walls to the auditoria to provide lateral re-
flections and reflective interiors to the boxes. They also showed that the original ceiling of the auditorium, con-
sisting of large catenaries hung longitudinally from the external shell, left the orchestra area suffering from a
deficiency of early reflected sound. Adding suspended reflectors over the stage did not substantially change
this situation and it was decided to reduce the distance between the side walls and raise the auditorium ceil-
ing substantially. Many different shapes and positions of reflectors suspended above the orchestra were stud-
ied and the final arrangement used small toroidal discs (slightly convex on the underside) suspended from the
ceiling by adjustable cables.
The acoustics of the finished auditoria were finally assessed in a number of test performances with real orches-
tras and audiences. Magnetic-tape recorders were used to record the decay of sound both after firing a gun
on stage (with a blank cartridge), and by getting the conductor to halt the orchestra abruptly while playing
Beethoven at full power. The results of these tests were used to recalibrate the model tests and to identify a
number of adjustments that could be made to the various moveable reflectors to achieve the best acoustic.
Jordan’s work on the Sydney Opera House was a milestone in the development of acoustics (Jordan 1973,
1980). His particular goal was to seek suitable criteria or parameters that could be used to define the acoustic
of a room. Since Wallace Sabine’s work in the early 1900s, it had become accepted that the key parameter
was the reverberation time. However, more and more auditorium designers and musicians were coming to re-
alise that this was not the only factor that made some halls good for concerts, and others bad. The question
was, what other parameters should be considered and, if they were to help designers, it had to be possible to
measure them. From the 1950s, developments in electronics and tape records gradually opened up the possi-
bility of measuring the acoustic behaviour of auditoria. Gradually qualitative terms often used by musicians,
such as ‘warmth’, ‘intimacy’, resonance’ and ‘fullness of tone’ came to be replaced by terms with precise
scientific definitions. One criterion for speech, proposed by a German acoustician in 1953, was named ‘Deut-
lichkeit’ (intelligibility) translated into the English of acoustics as ‘definition’. This was the proportion of the total
sound energy that had arrived at the listener within the first 50 milliseconds. For music one criterion was the ‘rise
time’ – the time it takes for the sound intensity to reach half its steady-state value. Another was ‘steepness’ –
the rate at which the energy arrived at the listener. For the Sydney Opera House, Jordan was one of the first
acousticians to use the ‘Early Decay Time’ (EDT) – the rate at which the sound intensity decays during the first
few milliseconds, rather than later when the sound has reached every part of the auditorium – as a measure
that gave a better correspondence between measurable quantities and listeners’ subjective judgements of
quality. During the final decades of the twentieth century many other acoustic parameters, often with curious
names, have been proposed as scientific measure of acoustic quality – terms such as the ‘inversion index’, the
Proceedings of the Third International Congress on Construction History, May 2009

ratio of the rise time measured in the auditorium to the rise time measured on the stage, which reflects the
need for the performers on stage to hear sounds before the audience, and others such as ‘hall-mass’, ‘point of
gravity time’, ‘clarity’ and ‘index of room impression’. Even today, however, different acousticians prefer their
own approach to defining acoustic performance and hence to designing auditoria to achieve the right con-
ditions for different uses (Barron 1997). Perhaps the very fact that acoustic quality is finally judged by people
rather than measuring instruments means that some subjectivity is likely to remain.

CONCLUSIONS
The development of design methods for the acoustics of auditoria has followed the same pattern observed in
other branches of building engineering design. Initially designers used their own experience to observe and
improve their art and collected their experience in the form of simple design rules which could be passed on
to other designers. In acoustics this approach was known in ancient times and has continued even into the
twentieth century. The technical difficulty of measuring acoustic phenomena delayed a truly scientific ap-
proach to understanding acoustics until the late eighteenth century (over a century later than for structural
engineering). The first scientific concept in acoustics, defined in quantitative terms by Sabine in the 1890s, was
the reverberation time whose relationship to the dimensions of a room was expressed as an empirical quantity
known as the absorptivity of the surfaces of the room. This approach remains the most important in acoustic
design today. The testing of scale models together with the use of non-dimensional constants was developed
in acoustics simultaneously with their use in the design of building structures, first in the 1930s and more widely
in the 1960s. Their use consolidated the understanding of acoustic phenomena and laid the foundation for
creating mathematical models using computers.

REFERENCES
Addis, Bill, 2007: Building: 3000 years of Design Engineering and Construction. London, New York: Phaidon.
Arns, Robert G.; Crawford, Bret E., 1995: Resonant Cavities in the History of Architectural Acoustics. Technology
and Culture 36, pp.104-35.
Barron, M., 1983: Auditorium acoustic modelling now. Applied Acoustics 16, pp.279-290.
Barron, M., 1997: Acoustic scale model testing over 21 years. Acoustics Bulletin May/June, pp.5-12.
Beyer, Robert T., 1999: Sounds of our times: two hundred years of acoustics. New York: Springer Verlag,.
Chladni, Ernst, 1787: Entdeckungen über die Theorie des Klanges. Leipzig.
Chladni, Ernst, 1802: Die Akustik. Leipzig.
Helmholtz, H. von, 1863: Die Lehre von den Tonempfindungen als physiologische Grundlage für die Theorie der
Musik (On the Sensations of Tone as a Physiological Basis for the Theory of Music). Berlin
Hunt, Frederick Vinton, 1978: Origins in Acoustics: the Science of Sound from Antiquity to the Age of Newton.
Woodbury, N.Y. (Reprinted 1992, Acoustical Society of America).
Inman, W.S., 1836: Principles of ventilation, warming and the transmission of sound. London.
Jordan, V.L., 1973: Acoustical design considerations of the Sydney Opera House. Jnl and Proceedings, Royal
Society of New South Wales, 106, pp.33-53.
Jordan, V.L., 1980: Acoustical design of concert halls and theatres. London: Applied Science Publishers.
Lachéz, Théodore, 1848: Acoustique et Optique des Salles de Réunions Publiques. Paris.
Langhans, C.F., 1810: Über Theater; oder Bermerkungen über Katakustik in Beziehung auf Theater. Berlin: Hayn.
Leacroft, R. and H., 1984: Theatre and Playhouse: an illustrated survey of theatre building from ancient Greece
to the present day. London: Methuen.
Maltese, Corrado (ed.), 1967: Francesco di Giorgio Martini “Trattati di architettura, ingegneria e arte militare”
(c.1485). Milan: Polifilio.
Mersenne, Marin, 1636: Harmonie Univesellle, contenant la Théorie et la Pratique de la Musique. Paris.
Patte, Pierre, 1782: Essai sur l’architecture théâtrale. Ou de l’ordonnance la plus avantageuse á une sale de
spectacles, relativement aux principes de l’optique et de l’acoustique. Paris: Chex Moutard.
Reid, D.B., 1844: Illustrations of the theory and practice of ventilation, with remarks on warming, exclusive light-
ing and the communication of sound. London.
Rhode, J.G., 1800: Theorie der Verbreitung des Schalles für Baukunstler. Berlin.
Sabine, Wallace C., 1922: Collected Papers on Acoustics. Cambridge, US: Harvard University Press.
Saunders, George, 1790: A Treatise on Theatres. London.
Smith, Thomas Roger, 1861: Acoustics of Public Buildings; or the principles of the science of sound applied to
the purposes of the architect and builder. London: Weale.
Spandök, F., 1934: Akustische Modelversuche. Annalen der Physik 20, pp.345-60.
Strutt, John William, (3rd Baron Rayleigh), 1877-78: The theory of sound. London: Macmillan.
Thompson, E., 1992: Mysteries of acoustics: Architectural acoustics in America, 1800-1932. Thesis. Princeton Uni-
versity.
Thompson, E., 2002: The Soundscape of Modernity: Architectural Acoustics and the culture of Listening in
America, 1900-1933. Cambridge, Mass: MIT Press.
Upham J.B. 1853, A consideration of Some of the Phenomena and Laws of Sound, and their Application in the
Construction of Buildings Designed especially for Musical Effects. Amer. Jour. of Science and Art, 65 pp.215-
226; 348-363; 66, pp.21-33.
Vitruvius, c.25BC: de architectura, Rome. English version reprint: Dover Publications, 1960.

You might also like