(This column originally appeared in the September 1977 issue of Contemporary Keyboard magazine.)
Reverberation is the series of closely spaced, overlapping echoes of sounds that imparts a feeling of space or ambience. When we listen in a small room such as a studio control room or our living room, there are few echoes and they die out rapidly. In a near-ideal performance space such as a properly designed concert hall, there are many echoes and they die out evenly. In a large performance space such as an enclosed sports arena, the echoes are so dense and long-lasting that the reverberant sound often overwhelms the direct (original) sound.
The quality of reverberation is crucial to our ability to listen to music without undergoing aural fatigue. Completely unreverberated sound, such as you might hear in an anechoic test chamber, is unpleasantly oppressive. Anechoic chambers are lined with a several-foot thickness of sound-absorbent material. Sounds that hit the walls are nearly completely absorbed. A handclap in an anechoic chamber is heard as a very sharp, thin, colorless impulse; speech seems disembodied and somewhat artificial. After a few minutes you become disoriented and begin to actually hear blood pulsating through your head. A vague sense of external pressure develops. The feeling of discomfort is as real as if you had to sit motionless in a hard chair for a long time.
Of course, anechoic chambers are extremely unreverberant sound environments, and are capable of producing greater discomfort and aural fatigue than almost any real performance environment. At the other extreme (maximum comfort and minimum aural fatigue) is the classical concert hall. Most of the world's great concert halls have similar reverberation characteristics: reverberation time (the time it takes an impulse of sound to die down to one millionth of its power—a drop of 60dB) of about 1.5 seconds, with a series of echoes so dense and randomly spaced that the decay seems completely smooth. When we listen in a concert hall, the relationship between the direct and the reverberant sound and the detailed character of the reverberant sound itself give us a sense of the performance space. We "hear the hall" through its reverberation, just as we see the hall through the reflection of light on the hall's surfaces.
An electronic music performer playing in a small club or recording in a studio has the opportunity to add artificial reverberation to his/her music, thereby generating an illusion of a more ideal performance space and presumably improving the listenability of the sounds. One who is recording may opt not to use artificial reverberation, reasoning that any playback of the recording, even in a small room, will include the reverberation of the playback space, and will therefore sound more "natural." This line of reasoning breaks down only when playback occurs via headphones. Completely unreverberated sound heard via earphones rivals the oppressive sound of an anechoic chamber, at least to my ears. So, how much and what kind of reverb to add is a matter of personal preference, subject to the generalization that the smaller the performance or playback space, the more desirable reverberation becomes. No reverb is required in Madison Square Garden—it has too much already. At the opposite extreme, headphone music requires some reverberation unless it is the performer's musical objective to give the listener a headache.
So what methods of generating artificial reverb are available? Let's start with the fanciest, most ideal reverb chambers and come down from there to devices that are practical for musicians to own and use. The original reverb chambers used by radio stations and recording studios were actually rooms with hard-surface (usually concrete) walls, ceiling, and floor, a speaker system at one end, and a pickup microphone at the other end. Although distances between reflections in such echo chambers were not as great as those in concert halls, acceptably long reverberation times were possible because concrete walls are good acoustical reflectors. Reverb chambers are still in use today; one well-known New York studio uses an elevator shaft. They sound good, but are somewhat impractical for musicians to carry from gig to gig.
Reverb plates are the most common professional studio reverberation devices. One popular device consists of a precisely made steel plate about four feet wide and eight feet long. A driver transducer is attached at one end and a pickup transducer at the other end. Sound waves in the plate are reflected from the edges, forming a complex series of echoes, which are, however, not quite as dense and uniform as those generated in a three-dimensional chamber. The reverberation time can be varied from one to five seconds by means of an acoustic damper adjacent to the plate, a feature of considerable utility to electronic musicians. Reverb plates have some basic drawbacks for gigging musicians: they are approximately the size, weight, and price of a grand piano.
Now we come to the ubiquitous spring reverbs, a class of devices that range from hobby-shop toys to fully professional chambers. A spring reverb consists basically of one or more wire coils, each with a driver transducer at one end and a pickup transducer at the other. The driver vibrates the wire by twisting it. The sound vibrations travel back and forth along the wire and are reflected at each end. The round trip generally takes about twenty-five milliseconds. The reverberation time is fixed by the properties of the wire and the amount of acoustic damping material at the transducer ends. The first spring reverbs were developed by Hammond nearly forty years ago, and used four-foot springs that were partially suspended in an oil bath! Modern reverb units are somewhat more compact.
Chambers reflect in three dimensions and plates reflect in two dimensions. Springs reflect in only one dimension, the axis of the wire. The reflections of a single spring are equally spaced in time. If, for instance, it takes twenty-five milliseconds for an impulse of sound to make a round trip in the wire, then that impulse will give rise to a series of decaying echoes, spaced twenty-five milliseconds apart, at the pickup transducer. Now suppose we feed the drive transducer with a 40Hz sine wave, which has a period of 25 milliseconds. The beginning of a cycle makes the round trip in exactly the time that a new cycle is starting. The reflections thus reinforce each other. This happens not only at 40Hz, but every whole-number multiple of 40Hz! At frequencies in between, the reflections are out of phase, and tend to cancel each other out. Thus the frequency response of a single-spring reverb in which the round-trip time of a vibration is 25 milliseconds is like that of a comb filter in which the peaks are spaced every 40Hz. This is shown in the graph. The peak heights are not constant, but are rolled off at low and high frequencies by the acoustic limitations of the spring reverb system.
The regularly spaced peaks of a single-spring reverb generally impart an artificial, metallic quality to the reverberation. This is in contrast to plates and chambers, whose frequency-response curves have dense, randomly spaced peaks, and whose quality of reverberation is much smoother and less metallic. My next column will discuss various types of good commercially available spring reverbs, and then go on tape and solid-state echo units.
For more articles by Bob Moog, please visit http://www.keyboardmag.com/analog/1318/moog-monday-hub-page/57761