Michael Drolet -- 2005
|Home||SOUND DESIGN FOR MUSICAL THEATRE|
Sound design for musical theatre is a system design process involving a lot of common sense. This article will describe the design of the sound system for a semi-professional musical, with consideration given to the compromises and trade-offs required meeting the design goals within the constraints of time and budget.
In September 1987, I was asked to design, install and operate the sound system for a production of the musical comedy, "Sweet Charity". The show was to be presented in Concordia's D. B. Clarke Theatre and run for a week in January 1988. Proceeds would be donated to the Children's Wish Foundation. A production meeting with the director, set designer and producers provided me with enough information to begin the design process.
The sound system would be required to provide the following features:
- Sound reinforcement for the performers and orchestra.
- Feeds of onstage sound to the orchestra, dressing rooms and stage manager.
- A feed of the orchestra for the singers and dancers.
An important goal for me was to design as transparent a system as possible. The audience should be unaware of its operation.
The design of the system would be constrained by a budget limit of $1500 for equipment rental. A further constraint would be the availability of suitable equipment on the rental market. Finally, as the D. B. Clarke Theatre is used during the day for classes, set-up time would be limited to twenty hours, spread over several evenings. Good planning would be required to take maximum advantage of the funds, equipment and time available.
A sound reinforcement system is an electro-acoustic system designed to enhance the transmission of speech in a public space. It is a fairly common experience that it is difficult to understand someone speaking across a large room.
Outdoors, sound intensity decreases with the square of the distance from the sound source. In terms of decibels (dB), this means a 6 dB decrease in intensity for every doubling of the distance from the source. A listener positioned twenty feet from the sound source will hear 6 dB less than a listener ten feet from the source. Outdoors, this decrease continues at the same rate to infinity.
In an enclosed space, near to the source, the inverse square law holds. Some of the direct sound from the source is reflected by the walls, floor and ceiling of the space before reaching the listener. The sound may be reflected more than once. The different paths followed cause each reflected sound to arrive at the listener later than the direct sound. These multiple reflections build up a reverberant sound field. This reverberant field exists everywhere in the space, but close to the source, the direct sound field predominates in intensity.
At a certain distance, dependent on the directionality of the source, the direct sound will be equal to the reverberant sound in intensity. This distance is referred to as the critical distance (Dc).
Dc= 0.141Ö Qsa (Equation 1)
Q is the directivity factor of the source;
S is the surface area of the space;
A is the average coefficient of sound absorption. (1)
Fig. 1 shows how critical distance is affected by varying the directivity factor (Q). The data plotted is for a room volume of 500,000 cubic feet, a reverb time of 1.5 seconds and a directivity factor ranging from 3 to 27. Critical distance can be seen to vary from 20 feet to 100 feet over this range of directivity factors. Clearly, critical distance increases with the directivity of the source.
Figure 1 -- Effective of Increasing Q
Beyond critical distance, the reverberant field will tend to mask the higher frequency components of the direct sound, which are responsible for the intelligibility of speech. This reverberant field is the main cause of a taler or singer's difficulty in being understood in a large room. Another source of difficulty can be interference from ambient noise, such as traffic rumble or air conditioning. In musical theatre, the orchestra, which is meant to accompany a singer, can just as well interfere with the intelligibility of the lyrics, especially in the case of an untrained voice singing against a loud orchestra. The sound reinforcement system for "Sweet Charity"; was expected to improve intelligibility and increase the signal-to-noise (lyric-to-orchestra) ratio.
B. Elements of a Sound Reinforcement System
The simplest sound reinforcement system consists of a microphone to convert the sound generated by the performers into electrical signals, an amplifier to provide electrical gain, and a loudspeaker to convert the electrical signals back into sound. A practical system would provide for more microphones to cover various areas on stage. It might also include more than one loudspeaker to cover more audience area.
C. Loudspeaker Selection and Placement
The loss of intelligibility (%Al) is inversely proportional to the directivity factor of the source. A practical limit for %Al s considered to be 15%.
%Al = 656D22 RT602
D2 is the distance from source to listener;
RT60 is the reverberation time of the room;
V is the room volume;
Q is the directivity factor of the source. (2)
Increasing the directivity factor of the source will minimize the loss of intelligibility. This is effectively what happens when you cup your hands in front of your mouth to shout to someone. The directivity factor of the unaided human vocal system is about 2.5. (3) That is, the sound energy of the voice is funneled into a cone covering only 40% (1/2.5) of the spherical volume surrounding the head. Critical distance increases with directivity factor. Using a loudspeaker with a directivity factor greater than 2.5 to artificially reproduce the voice will improve intelligibility. Solving Equation 2 for Q, it is possible to calculate the minimum Q necessary to limit the loss of intelligibility to 15%. For the D. B. Clarke Theatre, the minimum acceptable Q is 7.0.
In choosing a loudspeaker, the angular audience area to be covered must also be considered. A single device with a high directivity factor may have too narrow a coverage pattern to include all of the audience. It may be necessary to use multiple speakers fanned out to cover the entire area. I was able to obtain accurate drawings of the D. B. Clarke Theatre from the Concordia Department of Theatre. From these I constructed a less detailed scale drawing of the theatre in plan and elevation. (Fig. 2). To obtain as natural effect as possible, it is best to use a single loudspeaker location above the stage and along the centre line of the room. (4) The ear cannot easily identify the direction of a sound source in the vertical plane. Positioning the speaker above the stage will give the impression that the sound is coming from the performers on stage. Fortunately, there was an unused lighting pipe at exactly the right position in the D. B. Clarke. I decided to mount the speakers there.
Figure 2 -- D. B. Clarke Theatre
I have developed a computer program to calculate the necessary loudspeaker coverage angles in the vertical and horizontal planes, given the room dimensions. It is based on a set of formulas by Uzzle. (5) A listing of the program is included in Appendix C. The data listed in lines 200 through 260 are for the D. B. Clarke. The last three lines on the page are the results of a sample run of the program using the D. B. Clarke data. It can be seen that to include the first and last rows of seats, a loudspeaker with a vertical coverage angle (theta) of 58 degrees is required. A horizontal coverage of 65 degrees (alpha) and 70 degrees (phi) is required for the first and the last rows of seats, respectively.
The next step was to find a loudspeaker that I could rent, to meet these specifications. The rental company I usually use had JBL model 2385 speakers shich have a nominal 60 by 40 degree pattern, and a Q of 19. Two of these rotated horizontally, and splayed-out 40 degrees, would suit the job perfectly. It turned out that the model 2385 devices were not available for the dates I needed them. I had to settle for JBL model 2380 withcoverage of 90 by 40 degrees and a Q of 10.7. The wider coverage would mean a potential decrease in system gain of 2.5 dB which I would have to accept. (6)
D. Microphone Selection and Placement
My previous experience in sound of rmusical theatre had given me a preference for boundary effect microphone. (7) Placing a microphone diaphragm in the plane of a large boundary, such as the floor or wall, has two advantages. First, there is a doubling of microphone sensitivity caused by the pressure build-up at the boundary. This provides an increase of 6dB in gain. Secondly, interference effects between direct sound and reflections from the boundary are eliminated. This provides a smoother frequency response from the microphone. (8)
Fig. 3 shows the version of the microphone I have found most useful. It was first suggested by Becker and Wahrenbrock. (9) A secondary boundary has been added perpendicular to the floor. This boundary is formed to an angle of 135 degrees to further restrict the microphone's polar reponse. This restriction rejects reverberant sound from behind the microphone. The decrease in reverberant pickup[ permits the microphone to be used further away from the source (performer) without sounding hollow or distant. (10)
Figure 3 -- Pressure Zone Microphone
I felt that three of these microphones placed across the front of the stage would be sufficent to cover the entire stage to a depth of about twenty feet. Working from the set designer's sketches, I found that there would be several areas beyond this limit where dialogue and musical numbers would take place. I planned to hang two boundary microphones mounted on three-foot square plexiglass sheets from the lighting grid to give general coverage of the upstage areas.
Figure 4 -- PZM on Set
Once we had started rehearsing in the theatre, it became apparent that there were a few other places where I would need additional pickup. I was able to conceal two more microphones in the set. In one number, 'The Rhythm of Life', the performed moved over large areas of the stage. For this number, I used a wireless microphone concealed in his costume. A radio transmitter relayed the microphone signal back to the console. I prefer not to use wireless mics. because of their limited dynamic range and notorious unreliability.
I have already discussed how I selected and positioned the loudspeakers. I should mention that the JBL Model 2380 is a high frequency device. It maintains its Q or directivity only above 1.2 kHz. It would be necessary to find a low frequency loudspeaker to reproduce the rest of the audio spectrum. The Bose model 802 is a full-range speaker which is small enough to hang on a lighting rail. (11) I used an electronic filter (crossover) to split the audio into two bands which fed separate power amplifiers for each type of speaker. This technique, known as biamplification serves to reduce intermodulation distortion caused in the power amplifier. (12)
Figure 5 -- Sound Operations
In order to maximize the gain of the sound system before onset of acoustic feedback, I used a technique known as equalization. A graphic equalizer, with a separate control for gain in eachof twenty-seven one-third-octave bands, was used to remove peaks in the frequency response of the frequency response of the speaker array. After this process, no one frequency band had more gain than any other so the system had maximum stability. (13)
I stated earlier that the central overhead position of the loudspeaker helped to give the impression that the reinforced sound was coming from th eperformers rather than from an artificial source. There is a way to further enhance this impression. The brain cannot differentiate between direct sound and an echo arriving at the listener within a window of 5 to 35 milliseconds later. (14) I electronically delayed the reinforced sound by 20 milliseconds ro make use of this so-called precedeence or Haas effect. The listener's ear-brain system mistook the delayed sound for an echo, and perceived the total direct and delayed sound energy as a single source onstage.
Figure 6 -- Speakers during Installation
I also found it necessary to delay (15 milliseconds) the sound from the upstage microphones, to avoid an upleasant 'phasing'; effect when both upstage and downstage microphones were in use.
In order to keep the production running on cue, it was considered necessary to feed the sound of what was happening onstage to several offstage areas. The orchestra needed to hear the performers to keep in tempo with their performance. The stage manager needed to hear clearly cue lines delivered anywhere onstage. The performers in the dressing and 'Green' rooms had to hear so they wouldn't miss an entrance. I devosed a system to feed the pickup of the centre-stage microphone continuously to these areas. I augmented this with other microphones as as I followed the action onstage. Because there is is a great difference in the sound intensity of one performer delivering intimate dialogue and the intensity of the entire cast singing, it was necessary to compress the dynamic range of the the offstage feed to keep the level reasonably constant. Because of the long cable runs to the dressing rooms, I used a constant (70 Volt rms) voltage transmission system.
The original set design called for the orchestra to be onstage, as part of the visual experience. Due to a miscalculation, it turned out there was not enough space within the set for the seven musicians and their instruments.
This problem was overcome by moving the orchestra into the theatre's orchestra pit which had been completely closed and used as storage space for several years. The musicians could only just be heard throught the stage floor. It became necessary to mike each instrument and feed the orchestra sound to the audience, through speakers to the left and right of the proscenium. Some of the orchestra sound had to be fed onstage as well, for the benefit of the performers. Loudspeakers were hidden in the wings to reproduce the orchestra onstage. I had to adjust the sound levels and speaker placement carefully to prevent the orchestra sound from leaking into the stage microphones and causing acoustic feedback.
Figure 7 -- Orchestra Installation
Since the conductor and musicians didn't have a view of the stage, it was necessary to give them a sound feed from the stage. This was done with a combination of speakers and headphones fed from the dressing room system.
The final version of the sound system design (Figs 4-8) was successful from a technical standpoint. Performers could be heard clearly from all seats in the house. There were no apparent 'dead' spots onstage and the orchestra didn't drown out the performers. Best of all, the system was inobtrusive. There were no hums, buzzes or squeals to distract attention form the performance.
Figure 8 -- Entire Set Speakers and Mics.
The sound design was also successful in not exceeding the constraints imposed on it. Rental of the equipment came to only $925, 40% under budget. I had been able to achieve a working system with equipment available on the rental market, and the system was up and running for the cue-to cue run through right on schedule.
- Eargle, John, 'An Overview of Sound Reinforcement', db magazine (1982 February), pg. 28.
- Davis, Don and Carolyn, 'Sound System Engineering -- 2nd ed.' (Indianapolis, IN: Howard Sams and Co.), pg. 237.
ISBN 0 672 21857 7
- Eargle, John, 'Sound Recording -- 2nd ed.' (New York:Van Nostrand Reinhold Co., 1980), pg. 121.
ISBN 0 442 22587 1
- Davis, pg. 321.
- Uzzle, Ted, 'Room Geometry for Acoustics', Synergetic Audio Concepts Tech Topics. Vol. 6, No. 1, 1978, pg. 1.
- 'Models 2380 and 2385 Flat-Front Bi-Radial Horns', Data Sheet (Northridge, CA: JBL Inc., 1983), pg. 4.
- Drolet, Michael, 'Pressure Zone Microphones', Perforations, publ. National Film Board of Canada, 1981 May/June, pp 35-37.
- Herrold, Robert, 'The Realities of Surface Mounted Microphones', Sound and Video Contractor, 1988 February, pg. 78.
- Wharenbrock, Ken, 'The Development of the Multiboundary Pressure Zone Microphone', Synergetic Audio Concepts Tech Topics. Vol. 11, No. 4 (1984), pg. 2.
- Eargle, 'Sound Recording', pg. 126.
- Engineers' and Architects' Design Guide, Bose Model 802 Loudspeaker System. 2nd ed. (Boston, MA: Bose Corporation), pg. 6.
- Davis, pg. 405.
- Engebretson, Mark E., 'One Third Octave Equalization Techniques and Recommended Practices', Altec Engineering Notes Technical Letter No. 232 (1974 March), pg. 4.
- Haas , Helmut, 'The Influence of a Single Echo on the Audibility of Speech', Journal of the Audio Engineering Society Vol. 20, No. 2 (1972 March), pg. 150.
(Back to Document)
7. Appendix C: