Page 26 - Winter 2011

almost invariably had curtains in the wings, at the back of the stage, and above the proscenium. Sound from the performers that did not travel directly to the audience was absorbed, increasing the relative strength of the direct sound. Excess loudness was also controlled—and the drama of the per- formance was maximized. We need not use curtains or com- pletely cover the stage walls to get the benefit.
The bottom line: we need to adjust the shape of a hall to match the size of the hall. Large halls can use a shoebox shape successfully if they absorb or back scatter the first order reflections that would otherwise travel from the orchestra to the rear of the hall. This suggestion may be anathema to those acousticians who believe these reflections are essential to give support and loudness to the listeners in the rear. But these reflections contribute very little to loudness—remem- ber that individually they are weaker than the direct sound, and the total reverberant energy is often more than ten times the strength of the direct sound. Loudness comes from the reverberation. Sitting in the first row of the first balcony in BSH proves the point.
As halls become smaller the design goal should be to choose a shape that brings the audience closer to the musi- cians, and to obtain the needed reverberation by increasing the room volume overhead. Such halls need not have the 1.9 second reverberation time of BSH. BSH is large enough pro- vide late reverberation without excessive energy in the first 100 milliseconds of decay. Efforts to reduce the total absorp- tion of a small hall to the point where it can achieve the same reverberation time as BSH will result in massive amounts of early reverberant energy. This will prevent a clear sound, and prevent the formation of a background sound stream. The hall will sound less reverberant than if the sound was clear and the reverberation time was only 1.3 or 1.4 seconds. Jordan Hall at New England Conservatory demonstrates this effect beautifully.
There are examples of very small halls (~300 seats) that manage to combine both good clarity and reverberation through a combination of absorbing stage elements and a large internal volume. Both features must be present. Internal volume is expensive, and adding absorption to the stage can be politically difficult, so many small halls lack these features. But if there is enough stage and audience absorption to give good clarity throughout the hall, it can be relatively simple and inexpensive to increase the late reverberation through a modern electro-acoustic system. Properly designed—and this is not always the case—these systems increase the late reverberation time without reducing clarity, and transparent- ly add substantial beauty to the sound. In at least three major opera houses and spaces of all shapes and sizes these systems have been operating for more than twenty years with excel- lent reviews from the critics and the public.
Postlude
This paper is not as controversial as it might seem. The model of hearing we present is similar to the latest work on
the subject, particularly the model proposed by Torsten Dau at the Danish Technical University. The sections on stream formation and its effects on sound are found in standard lit- erature such as “Auditory Scene Analysis” by Bregman. Comb filters are also not new, having been proposed by Peter Cariani. What may be new is our proposal that separation of sound elements by pitch can precede their analysis for timbre and direction, and that the information necessary for this separation lies in modulations induced by the phases of upper harmonics. We have found the physics needed to make this idea work, made a model of the process in the C lan- guage, and shown that when the model operates on binaural recordings of live music it predicts the point in a hall where localization disappears.
Perhaps the most controversial proposition is that pop- ular acoustical thinking is incorrect in believing that more early lateral reflections are always good, that clarity can be measured by standardized measures such as C80 and C50, and that the strength of reflections and reverberation should be independent of frequency. In our view when the direct sound is weak early reflections from any direction, but especially medial reflections (those from the front, rear, and ceiling), are detrimental to the sound. If clarity is defined by the ability to distinctly hear the notes in a per- formance, a high value of C80 or C50 often predicts the opposite. The best of the standard measures, IACC80, is insensitive to medial reflections. But at the recent International Conference on Acoustics in Sidney a keynote speech by Leo Beranek and several other papers called into question the reliability and even the relevance of these measures. The field of room acoustics seems open for change. Hopefully the neurology of hearing will play a prominent role in this process. AT
Reference
1 With the permission of the Pacifica String Quartet we can hear two examples from a concert in a 1300 seat shoebox hall. The sound in row F is quite different from the sound in row K. The recordings are from the author’s eardrums, and are equalized for playback over loudspeakers or headphones equalized to sound identical to loudspeakers. (Most headphones have too bright a sound to reproduce them correctly. Pink noise played though the headphones should sound identical in timbre to the same noise played through a frontal loudspeaker.) Instructions for downloading the audio clip examples are given in the side- bar.
“Binaural Recording of the Pacifica String Quartet in Concert row F”; (http://www.davidgriesinger.com/Acoustics_Today /row_f_excerpt.mp3)
“Binaural Recording of the Pacifica String Quartet in Concert row K”; (http://www.davidgriesinger.com/Acoustics_Today/ row_k_excerpt.mp3)
22 Acoustics Today, January 2011