by Trevor Cox
At an outdoor concert, an orchestra might play from a tented stage while the audience enjoys a picnic. The night often ends with bottles of champagne and fireworks exploding overhead. These concerts are fun, but the orchestra sounds thin and remote. In contrast, within a great venue like Symphony Hall the music appears to fill the room and envelop the audience from all sides. The reverberation inside amplifies the orchestra, allowing for more impressive loud playing. It also makes sounds linger a little, enabling musicians to make smoother transitions from note to note. Reverberation helps create a more blended and rich tone. As the twentieth-century conductor Sir Adrian Boult put it, “The ideal concert hall is obviously that into which you make a not very pleasant sound and the audience receives something that is quite beautiful.”8
The transformational effect of reverberation is not restricted to classical music; it is also used extensively in pop. The 1947 number one hit “Peg o’ My Heart” (a slow instrumental played on giant harmonicas), by Jerry Murad’s Harmonicats, was the first recording to use reverberation artistically.9 Since then, “reverb” has become a ubiquitous part of the music producer’s tool kit. It makes voices sound richer and more powerful, mimicking what would happen if the person were singing from the stage of a theater. On many television programs where people with lousy voices attempt to sing, as soon as the person hits the first note you can hear the audio engineers slathering on reverberation to rescue the sound.
Reverberation is not the only important feature of a good auditorium. The most infamous concert hall failure is probably the original Philharmonic Hall at Lincoln Center in New York, which opened in 1962 (and was later rebuilt as Avery Fisher Hall). Acoustician Mike Barron describes it as “the most publicized acoustic disaster of the twentieth century.”10 Influential music critic Harold C. Schonberg was particularly vocal, describing the hall as a “great big, yellow, $16,000,000 lemon.”11 Acoustic expert Chris Jaffe described how Schonberg “had a field day writing article after article on the acoustics of the hall as a sort of All My Children–type soap opera.”12 Ironically, the acoustic consultant for the hall was Leo Beranek, possibly the most influential architectural acoustician of the twentieth century, and also the only person famous enough to be pursued by groupies at acoustics conferences. I remember my first meeting with Leo over breakfast at a conference when I was a young academic. It was a brilliant chance to talk about my research into concert hall acoustics with this superstar. Unfortunately, he greeted me with a question about why I had been measuring echoes from duck calls (see Chapter 4).
According to Beranek, late alterations to the design doomed Philharmonic Hall. The original concept called for a simple shoe box shape similar to Boston’s Symphony Hall. But some thought there were not enough seats in the proposed auditorium. Several New York newspapers campaigned to have the capacity increased, and Beranek says the committee overseeing the building “caved in.”13 The new design changed the shapes of the balconies and sidewalls, and called for a raft of reflectors above the audience. When the hall opened, critics complained that there was too much treble and not enough bass, and musicians struggled to hear each other, making it difficult for the orchestra to blend its sound. Looking back with current scientific knowledge, Beranek now claims that without these alterations, “we would have been the toasts of New York.”14
Room shape plays an enormous role in the quality of concert halls. Sound reflections heard from the side are very important because the acoustic waves at our two ears are different. It takes longer for reflections from each side to reach the farthest ear; in addition, that ear is in an acoustic shadow and thus picks up fewer high frequencies because that sound does not easily bend around the head. These two cues signal to the brain that music is not just coming from the stage, but also from room reflections. Because of side reflections, we feel enveloped by the music rather than perceiving the sound as coming from the performers on a distant stage. These reflections also make the orchestra appear physically wider than it is—an effect called source broadening, which listeners tend to like.15 Boston’s Symphony Hall achieves this effect through the narrow shoe box shape, which offers plenty of side reflections. The scientific understanding of side reflections has inspired new designs and shapes for halls. Near my home in Manchester, England, the Hallé orchestra plays in the Bridgewater Hall, built in the 1990s. The back half of the audience is divided into blocks with walls between in a pattern called vineyard terracing. The partitions separating the audience areas have been carefully angled to create reflections arriving from the side.
Reverberation is all about striking a balance between too little (like being outdoors) and too much. Composer and musician Brian Eno explained the consequences of excess reverberation in the Royal Albert Hall before it was modified:
It was awful, any piece of music which had rhythm or any kind of speed to it would be completely lost there because every event carried on for long after it was supposed to have ended. It reminds me of when we were at art school, we had a model who was very very fat. We used to say it took her 20 minutes to settle, she was impossible to draw. Well playing fast music in a lot of reverberation is a bit like that.16
The desirable amount of reverberation depends on the music that’s being listened to. Intricate chamber music by Haydn or Mozart was composed to be heard in courts and palaces, so it works best in smaller spaces with shorter reverberation times—say, 1.5 seconds. The French romantic composer Hector Berlioz wrote about hearing Haydn and Mozart played “in a building far too large and acoustically unsuitable,” complaining that they might as well have been played in an open field: “They sounded small, frigid and incoherent.”17
For romantic music by Berlioz, Tchaikovsky, or Beethoven, more reverberance is needed than for chamber music—say, a reverberation time of about 2 seconds. Organ and choral music demands even more. As renowned American concert organist E. Power Biggs said, “An organist will take all the reverberation time he is given, and then ask for a bit more . . . Many of Bach’s organ works are designed . . . to explore reverberation. Consider the pause that follows the ornamented proclamation that opens the famous Toccata in D minor. Obviously this is for the enjoyment of the notes as they remain suspended in the air.”18
The Royal Festival Hall in London was built as part of the Festival of Britain in 1951, which was meant to help cheer up the nation after years of rationing and austerity during and after the Second World War.19 While critics adored the building, reviews of the acoustics in the concert hall were mixed, with an eventual consensus forming that the reverberation time was too short, being only 1½ seconds. In 1999, the conductor Sir Simon Rattle said, “The RFH is the worst major concert arena in Europe. The will to live slips away in the first half hour of rehearsal.”20 Hope Bagenal was the original senior acoustic consultant for the hall. Surprisingly, he was not a scientist by training. Acoustic engineer David Trevor-Jones wrote that Bagenal’s “broad, liberal education” was important because it gave him “the inquisitiveness and . . . competence to take up as much of the physics of acoustics as he needed.”21 Sabine’s equation would have shown Bagenal that there were two solutions to the hall’s dry acoustics. The first was to increase the size of the room, allowing more space for the sound to bounce around. Raising the roof would have been effective but impossibly expensive. The second solution was to reduce the acoustic absorption in the room. In a concert hall, the audience is responsible for most of the absorption. Bagenal recommended removing 500 seats to increase the reverberation time, but this was not done.22 Instead, a revolutionary solution was sought: to use electronics to artificially enhance the acoustics.
In the ceiling of the hall, microphones were mounted inside pots to pick up sound at particular frequencies. The electronic signals from the microphones were then amplified and fed to loudspeakers elsewhere in the ceiling. The sound circulated around a loop, going from microphone to loudspeaker through the electronics, and then from loudspeaker to microphone through the air. This s
etup sustained the sound for longer in the hall, creating artificial reverberation. This was a remarkable feat of engineering, considering the crude electronics available in the 1960s. The mastermind behind this assisted resonance system was Peter Parkin, who began working in acoustics in World War II, helping to defeat underwater acoustic mines. For his work on the Royal Festival Hall, Parkin had a dedicated telephone line from the hall to his house so that he could listen in and check whether the system was working correctly.23 He was monitoring for faults in the system that might cause the sound circling between the microphones and loudspeakers to grow louder and louder, resulting in feedback, the howling and screeching we associate with heavy metal.
Peter Parkin’s electronic system raised the reverberation time from about 1.4 seconds to over 2 seconds for low frequencies, vastly improving the warmth of the sound. But Parkin kept it secret. The use of electronic enhancement with classical music is so controversial that when the assisted-resonance system was first installed, it was brought in gradually without telling the orchestra, audience, or conductors. Once the full system had been covertly used for eight concerts, the engineers then dared to reveal its presence. The system was used until December 1998, when a nonelectronic solution was sought.
I sympathize with the view that classical music should not be electronically enhanced, especially after hearing a demonstration of a different electronic system about twenty years ago in a theater near London. As the engineers switched between different settings, I heard strange mechanical and unnatural distortions, and the sound sometimes even appeared to come from behind me rather than from the stage. Amazingly, this was a demonstration designed to encourage people to buy the technology. Nowadays, however, modern digital systems, used in many contemporary theaters, can be impressively effective. I heard one demonstrated last year at an acoustics conference, and with a flick of the switch the lecture hall was transformed into a lyric theater or a grand concert hall with natural-sounding acoustics.
A list of very reverberant spaces would include many mausoleums: the Taj Mahal and Gol Gumbaz in India, Hamilton Mausoleum in Scotland and Tomba Emmanuelle in Oslo, Norway.24 The large rooms and hard stone walls make these places very lively.
The artist Emanuel Vigeland built Tomba Emmanuelle in 1926 as a museum for his works, but later he decided to use it as his last resting place. Norwegian acoustician and composer Tor Halmrast, who is physically and aurally larger than life, described stooping to enter Tomba Emmanuelle—actually bowing beneath the ashes of the artist, which are in an urn over the door. Halmrast entered a tall, barrel-vaulted room covered with frescoes. He explained, “When entering, you see almost nothing, as the walls are very dark. After some time you see the paintings all over the walls and curved ceiling: everything from life (even copulation before life) until death.”25 One fresco shows a plume of smoke and children rising from a pair of skeletons reclining in the missionary position. The midfrequency reverberation time is 8 seconds, a value you might expect in a very large church, which Halmrast thinks is remarkably long, considering that the room is relatively small.26
The explicitly sexual frescoes in Tomba Emmanuelle are in marked contrast to the dour interior of the Hamilton Mausoleum, but which is more reverberant? The world record was based on slamming the bronze doors of the mausoleum’s chapel—a very unscientific test. To properly compare the reverberation, one needs initial sounds of equal quality and strength.27 If Rebecca Offendort from Hilaire Belloc’s cautionary tale “Rebecca Who Slammed Doors for Fun and Perished Miserably” were doing a measurement, she would “slam the door like billy-o!” and it would take a long time for the sound to dissipate.28 A less vigorous experimenter would record a shorter time.
For my visit to the Hamilton Mausoleum, acoustician Bill McTaggart brought along proper measurement gear. Across the room he placed on a tripod a strange-looking loudspeaker that sends noise out in all directions (Figure 1.1). It was an icosidodecahedron about the size of a beach ball. A microphone on another tripod stood some yards away. These were all linked to analyzers, whose screens showed graphs of jagged lines sloping from top left to bottom right, characteristic of decaying sound. Normally, acoustic engineers use this equipment to examine whether walls permit too much sound to leak between neighboring houses, or whether the reverberance of a classroom will be too great, undermining the teacher’s lessons.
Figure 1.1 Loudspeaker used to test Hamilton Mausoleum, with cupola above.
Bill gave a signal, and I quickly stuck my fingers in my ears to protect my hearing. The loudspeakers roared into action, blasting out a growling noise, which even with my ear canals blocked sounded immensely powerful. After 10 seconds, Bill suddenly cut the loudspeaker and measured the noise decay while I quickly unblocked my ears so that I could enjoy the swirl of reverberation. The massive solid walls reflect sound very efficiently, and it took a long time before the noise completely disappeared. The initial overwhelming hissing roar morphed into a rumble that moved above my head, disappearing and dying away up near the cupola. There was a brief moment of silence before the assembled acoustic experts broke into fevered discussions.
What was the reverberation time in the Hamilton Mausoleum? Since it is a large space constructed mainly of stone, the reverberation time is very different at low and high frequencies. At low frequency—say, 125 hertz, which is an octave below middle C (a frequency typical of a bass guitar)—the reverberation time was 18.7 seconds. For midfrequencies it was just over 9 seconds.29 Impressive, but if it is the longest reverberation in the world, I would be very surprised.
The midfrequencies are where speech is the most powerful, where our ears are the most sensitive, and hence where reverberation times are the most important for clear hearing. No wonder any ideas of holding ceremonies in the chapel were abandoned. Someone talking in a normal way says about three syllables every second. At that rate of speaking in the mausoleum, you utter several words before the first one dies away 9 seconds later. Inevitably, the sounds from many different words mingle and merge. Speech in the chapel is not too bad if you have a conversation with someone close by, because the direct sound from a nearby voice is much louder, making it easier to ignore the reflections. Speaking more slowly also helps. But once you get too far away from someone you want to talk with, the direct sound is less than the reflections, and the reverberation fills in the silences between the syllables, blurring the peaks and troughs in the sound wave and making the speech unintelligible.
Some cathedrals are ten times bigger than the chapel at the Hamilton Mausoleum, and a larger size should mean a longer reverberation time, according to Sabine’s equation. The vast, visually imposing cathedrals, built to glorify God, naturally have awe-inspiring acoustics. The sonic qualities are associated with spirituality. The excessive reverberation forces the congregation into silence or hushed whispers, because otherwise speech is rapidly amplified by reflections and creates an ungodly cacophony. During services, the music and words appear to wrap around you like the omnipresent God being worshipped. The acoustic has also influenced services, as the use of chanting and the slow liturgical voice counters the muddy speech in such reverberant spaces.30
Many centuries ago, the priest would stand in the chancel, almost cut off from the congregation in the nave. Typically, only a small opening above the chancel screen and below the tympanum would allow sound to reach the churchgoers. The priest would chant facing the altar, his back to the worshippers, so any speech reaching the congregation would be a mush of reflections, with all sound arriving indirectly via reflections off the walls and ceiling. Mind you, with the service mostly in Latin it could be argued it was not the acoustics that caused the speech to be unintelligible.
The Reformation of the sixteenth century changed all that: Anglican priests were instructed by the Book of Common Prayer to speak from a place where they could be heard more clearly.31 Services in English meant that speech would have to be understood. Innovations such as the pulpit in the nave enabled liste
ners to hear more clearly. There were still reflections, but because the direct speech reached the congregants’ ears quickly and some strong reflections arrived shortly after, the setup tended to aid communication. Late reflections, however, make matters worse.
Why are some reflections useful and others detrimental? It comes down to how our hearing has evolved to cope with a complex soundscape. In a cathedral, like most places, the ear is bombarded with reflections from all around—from the floor, walls, ceiling, pews, members of the congregation, and so on. In a large cathedral there are many thousands of reflections per second.32 Perceiving each individual reflection would quickly overwhelm our hearing. Consequently, the inner ear and brain combine the reflections into a single perceived sound event. Thus, when we clap our hands in a room, we usually hear only one “clap,” even though the ear actually receives many thousands of slightly different reflections of the sound in close proximity. A room does not turn a single hand clap into applause.
The ear is a little bit sluggish, rather like a heavyweight boxer. When the ear receives a very short sound, like a hand clap, or when a boxer is hit by a fast punch, it takes a little time for the system to respond to the stimulus. Both the ear and the boxer also continue to respond after the initial stimulus has gone away: the heavyweight boxer reels and rocks for some time after the punch has landed, and similarly, the hair cells in the inner ear continue to send signals up to the brain for some time after the clap has stopped. On top of this physical sluggishness in the ear, the brain is also constantly trying to make sense of the electrical signals coming up the auditory nerves. The brain employs several tactics to separate the priest’s direct speech from the morass of late reflections reverberating around the cathedral.33
If the priest is to one side, the ear nearer the priest receives louder sound waves because the farther ear receives only speech that has bent around the head. The brain thus attends more to the nearer ear, where the speech is louder and easier to pick out among the reflections. Attention focused in this way becomes less effective if there are lots of reflections from many directions, because both ears become overloaded with a wash of unwanted reverberance.
