By Ralph Heinz
Communication and performance are important functions in many types of building spaces, from hotel ballrooms to classrooms, transit terminals to sports venues, auditoria to places of worship.
Key factors that must be controlled for intelligible speech and enjoyable music include multiple sound arrivals and the ratio of direct-to-reverberant sound. These can be controlled physically through the design of the space and the choice of surface materials and seating, acoustically through the design and placement of loudspeaker systems, electronically through the use of microphones and amplification, and digitally through the use of digital signal processing. Each of these technologies has certain advantages and disadvantages. Recent advances in digital signal processing (DSP) have made it cost-effective to use this technology. When integrated with proper electronic and acoustical design, DSP-driven arrays can control acoustical energy while minimizing visual impact on the architecture.
Many public spaces are designed to enable the users to hear the space and the level of activity and excitement within it. But excessive amounts of uncontrolled acoustic energy (sound) will interfere with both the understanding of speech and the enjoyment of music.
Sound is a pressure wave that travels through the air. Like light, it comes in both direct and indirect forms. But where indirect lighting is often preferred, secondary or indirect acoustical energy is of little practical use. In fact, it is inimical to both spoken communication and musical performance, except in very moderate doses.
The speed of sound is orders of magnitude less than the speed of light. This means that when the same initial sound reaches a listener's head through different pathways, similar signals arrive at different times. A full explanation of psychoacoustics and the ear/brain mechanism for perceiving language, melody, rhythm, and harmony would be too long for this article, but we can state that multiple arrivals rapidly turn the "signal," whether it is a spoken word or a musical phrase, into "noise."
Consequences of multiple arrival include "lost consonants" (listeners can't hear the difference between "deer" and "beer" so communication suffers) and "temporal smear" that interferes with musical perception.
Uncontrolled direct sound is radiated spherically, in all directions at once. Secondary acoustic energy is reflected or absorbed and diffused (scattered) and diffracted (bent) by the various surfaces it encounters on its way to the audience.
Passive vs. Active Sound Control
Controlling acoustical energy means controlling the direct radiation, the secondary pathways, or both. The goal is to obtain an optimal direct-to-reverberant (D/R) ratio, which for our purposes is almost the same thing as a signal-to-noise ratio.
When the D/R ratio is greater than 1, a listener is considered to be in the "near" or "direct field." When the D/R ratio is below 1, the listener is in the "far" or "reverberant field." Acoustical control technologies aim to locate as many listeners as possible in the near field.
The most basic technologies for controlling sound are passive, involving the design of the space and the choice of surface materials.
Consider a classical opera hall, for instance. The "wedding cake" design minimizes the distance sound has to travel from the stage to the seats. Abundant decorations scatter the secondary sound, turning echoes into reverberation.
Operatic singing technique is designed for this environment: Consonants are almost inaudible, because they will be lost to most of the audience in any event. It is assumed that the audience knows the libretto and even the lyrics, so that they are able to fill in any information that is missing from an unamplified performance - and there is plenty of filling in to be done.
Symphonic halls and cathedrals function acoustically in a similar fashion: The reverberant field begins quite close to the stage or altar, allowing most of the audience to know when the music or the service of worship is being performed, without necessarily delivering all the details of the content. This type of architecture functioned reasonably well for centuries before the development of amplification. Acoustically, the emphasis is on diffusion and reflection: hard materials such as stone or plaster, with complex surfaces and many hollows and protrusions.
Modern sound systems can extend the direct field far beyond the limits of the human voice or acoustic instruments. In response, most modern architecture emphasizes unbroken planes and uses more absorptive materials to control sound.
But no matter what the architectural style, physical control of acoustic energy is limited and imposes constraints on the architect.
The first limitation is that physical control techniques are built in and fixed (with the exception of a few concert halls that have opted for expensive physical structures to vary the acoustics). If absorption and diffusion are required, design will have to be compromised in the choice of surface materials and even the shape of the room itself.
Acousticians can certainly help to find a reasonable balance between the competing claims of the eye and the ear, but they cannot eliminate the necessity to trade one off against the other.
Sound systems can help, both by amplifying sound and by actively controlling it. The basic principle of active control is to focus sound on the audience and keep it away from boundary surfaces: This maximizes the D/R ratio and helps keep listeners in the near field, where speech can be understood and music appreciated.
Line Arrays: The Modern "Horn and Box"
Originally, active control technologies were physical. An example is the acoustical horn, a technology that has been in use since the first wind instruments and a key element of the pioneering sound systems developed for "talkies" in the early 1930s.
Horns prevent sound from radiating in all directions at once - the sound on axis is louder and gets progressively softer as you move off axis.
Horns are physical devices, so they must be large in order to work well at lower frequencies where sound waves are long. They must be aimed and located properly in order to do their job, which usually dictates a spot in midair above the audience. They can also be in multiple spots if there are multiple seating areas: one horn (or array of horns) for the orchestra seating and another for each balcony.
These requirements make it difficult or impossible to integrate horns into the design of the space. Some of these systems are concealed behind cloth scrims - but then a suitable structure must be available in the walls or ceilings. In some cases, these structures are added onto the basic room at considerable cost, solely to accommodate a modern sound system without making it too visually intrusive.
Over the past 50 years, professional audio technology has progressed from the "horn and box" model to packaged loudspeaker systems that integrate the horn and the box into a single enclosure. A modern variant, the so-called "line array," deploys a number of packaged loudspeakers in a curved vertical array.
Each enclosure may have one or more horns or other waveguide technologies based on diffractive, refractive, or reflective models. While they may be more compact than their predecessors, these arrays don't work well in semi-enclosed mounting locations; they are typically hung in free space where they can interfere with sightlines and will certainly create a visually imposing presence in the room.
Another active technique for controlling sound involves putting a sound source near each listener. Seat-back speakers
or distributed wall-mounted or column-mounted speakers have been used to avoid the problems created by a smaller number of louder (and larger) loudspeaker systems. But the cost of installing all these speakers and distributing audio signals to them rapidly becomes prohibitive.
The resulting compromise often leaves most listeners outside the direct field. At the same time, the many enclosures, although small, can create visual clutter through force of numbers.
Electrical technologies for controlling sound usually go hand in hand with physical techniques. It is possible to build a "dead" room with many absorptive surfaces and then add reverberant sound back in to taste using a system of several microphones and many small loudspeakers. This technology is costly and imposes design limitations because the room must be highly absorptive in order for it to work.
Another electrical technology uses a few microphones to sense the ambient noise level and adjust the sound system's output to match. This technology has proven to be more effective in luxury autos (highly absorbent environments where the problem is exterior noise, not uncontrolled acoustical energy) than in restaurants or transit terminals.
The newest practical technology for controlling acoustical energy involves the use of digital signal processing (DSP). Like other silicon chips, the cost of DSP has dropped continually as its power has increased. Its cost/performance ratio recently crossed a threshold, making it practical to deploy this technology in a range of applications.
Acoustical and electronic technologies aim to provide enough amplification power by combining multiple sound sources while minimizing destructive interference between them. By contrast, DSP-driven arrays use interference between adjacent sources as a technique of control.
Loudspeakers are transducers: They turn electrical energy (voltage and current) into acoustic energy (sound pressure waves). If each transducer in an array receives a unique electrical signal, the interference effects can be predicted and used to control the behavior of the acoustical wavefront that is radiated from the array. The resulting performance can be quite impressive.
The latest generation of DSP-driven vertical arrays can produce multiple beams of acoustic energy from a single vertical column of transducers. Beams can be as narrow as 5 degrees in the vertical plane and they can be aimed up to 30 degrees above or below the horizontal axis of the array.
In addition, the "acoustical center," the apparent origin of the sound beam or beams, can be raised or lowered without moving the physical enclosure. Moreover, the shapes and aiming angles of the beams can be maintained over a range of frequencies.
The acoustical performance is produced by controlling the electrical signals driving each transducer, which is in turn controlled by a digital signal processor. The DSP applies "Finite Impulse Response" (FIR) filters, whose shapes are determined by complex mathematical calculations.
Again thanks to advances in silicon technology, laptop computers can run graphical user interface (GUI) software that makes it easy to define the desired pattern of acoustical radiation. All the math is done in the background, and the filters are uploaded to DSP firmware onboard the loudspeaker enclosures.
The enclosure combines the virtues of large-format loudspeaker systems (high output, meaning a few devices can fill a large room with sound) and small distributed systems (a small visual footprint). The columns can be quite tall: over 12 feet in some instances. But they are only half a foot wide. Because they mount flush to walls or columns and can be recessed without affecting performance, these column arrays can be almost invisible. Yet a single array can cover multiple seating areas and/or a very large audience.
Although each individual transducer is fairly small, there are lots of them, so the output can be quite high: up to 103 dB at 100 feet. The sound beams can be shaped to provide consistent sound pressure level (SPL) over distances as long as 300 feet.
While their performance makes them a prime choice for very large, highly reverberant spaces such as cathedrals, basilicas, or transit terminals, slightly smaller versions of the same technology can also be used in a much broader range of buildings.
For general applications, DSP-driven vertical arrays can allow the architect to use more reflective surface materials such as stone, steel, and glass without making it impossible to understand speech or enjoy music. Even where the design includes a good amount of absorption and the natural acoustics of the room would be considered well balanced for its range of functions, a DSP-driven array can be the preferred option because one or more can be installed without a significant visual impact.
Like every technology, DSP-driven arrays have limitations. Presently, the main constraint is SPL. If extremely high "rock 'n' roll" volume levels are desired, the amplification will have to be provided by large-format loudspeaker systems.
No matter the size or intended uses of a building, it must provide some control of acoustic energy in order to function well. Available technologies include: passive acoustics in the form of building structures and surface materials (which may, at some cost, be made variable); active acoustics, which today come in the form of packaged loudspeaker systems; active electronics, mainly used to compensate for "dead," highly absorbent passive acoustics; and DSP-driven arrays, which use computer technology to control sound.
For many applications where extremely high volume levels are not a requirement, DSP-driven arrays offer the most latitude to combine striking and impressive visual aesthetics with usable acoustical performance.
Ralph Heinz is senior vice president at Renkus-Heinz. He is in charge of product development and has developed several patented innovations in sound reinforcement technology.
- ISO 3382 - Room Acoustics Measurement Parameters
- Olson, Harry F, Music, Physics & Engineering
- Heinz, Ralph, DSP-Driven Vertical Arrays
ARCHI-TECH welcomes comments on the distance education program, including its content, format, and process for AIA/CES learners.
Please contact Maureen Patterson at (319) 364-6167 or [email protected].
Stamats Buildings Media, publisher of ARCHI-TECH, is the provider of ARCHI-TECH'S AIA/CES distance education programs. Stamats Buildings Media's provider number is J683.