Reverb is arguably one of the mixing engineer’s most-used effects. Though its popularity waxes and wanes (there was a time when singers sounded like they were in the bottom of a well, and a time when reverbs were treated as tools of the devil; decent people wouldn’t dream of using them), in general, they’re enormously useful at adding a sense of space to a recording. The earliest presence of reverb on recordings wasn’t necessarily planned — it was simply an artifact created by recording a group of musicians in a live room. The first artificial reverb, the chamber, came along later.
CHAMBERS
A reverb chamber is really a simple invention; take an empty room, put a speaker at one end and a microphone at the other. (Point the speaker at a reflective wall rather than directly at the mic to minimize the direct signal). Play a track through the speaker and the microphone picks up the sound of the track after it bounces around the room. These “echo chambers” (as they were first known) didn’t provide echo as we use the term today (that is, discrete repeats of the original sound); instead, the effect was a wash of diffuse reflections from all the walls of the chamber. The dry signal (or signals) were sent from the mixing board (through an aux send named, simply enough, “echo send,”) and the signal from the mic was returned through the “echo return,” with the level of the reverb controlled by the return. The larger and more reflective the room, the longer the reverb would last, so chambers were built with hard walls (often cement) painted with reflective paint. Quite a number of chambers were built without parallel walls to minimize flutter echo, and had signature sounds that are still recognizable today. For many engineers, the most famous (and possibly the best sounding) chambers are those of Capitol Studios in Los Angeles. Gold Star Studios’ chambers were a huge part of Phil Spector’s “Wall of Sound,” and the chambers at Abbey Road (specifically Chamber 2) was used on the Beatles’ records.
PLATES
A big advance in artificial reverbs was the development of the plate reverb in 1957 by a German company called EMT. Though the original plate reverb weighed close to 600 pounds and was eight feet long, five feet high and a couple of feet wide, it still took up less space than a chamber. A plate reverb is a large metal plate with a driver mounted in the middle of it. Signals sent to the driver cause the plate to vibrate, and a transducer mounted close to the edge of the plate picks up those vibrations and returns them to the console. The reverb time could be adjusted on the EMT plate using an adjustable damper. Even if you’ve never seen a plate reverb, you’ve likely noticed that every software-based reverb has settings labeled “Plate.”
DAWN OF DIGITAL
EMT didn’t only create the first plate reverb; they also, in 1976, created the first successful digital reverb: the EMT 250. Along with its successor, the EMT 251, these units were arguably the first digital multi-effects units; echo and delay were available, along with modulation effects such as phase shift and chorus.
Though artificial reverb processors are primarily used to simulate the sounds of real acoustic spaces, chambers, plate reverbs and occasionally spring reverbs, they can also be used in creative ways that do not imply any sort of real space.
In the digital world, there are two main approaches to creating reverb: algorithm-based processors and convolution reverbs. Algorithm-based reverbs synthesize the sound of an acoustic space through mathematical calculations of delay times, while convolution reverbs are essentially samples of actual spaces that are captured using a series of impulse responses played in the space. These impulse responses are manipulated to make them suitable for use with the software and then saved as presets. When a a dry signal is fed into the convolution software, the dry signal and the impulse response are subjected to an algorithm that essentially combines — or convolves — the reverberant characteristics of the impulse response with the dry signal.
Each approach has advantages and disadvantages; algorithm-based reverbs offer a great deal of control over the parameters used to create the sound of a room (or the sound of a non-existent rooms), and though convolution reverbs typically have fewer adjustable parameters than algorithm based reverbs, they have the advantage of realism. Convolution reverbs can successfully sound like actual spaces because they are based on recordings done in actual spaces.
PARAMETER CONTROL
If you’re working with software-based reverbs, these are some of the parameters you’re likely to have available. A great deal of the fun of reverbs is creating your own settings; though in the context of a full mix, the reverb sounds aren’t likely to matter to most folks, as an engineer, you can rest easy knowing that even the smallest details are under your control.
Reverb Time
Reverb time is generally accepted as the amount of time it takes for a level of the reverb tail to decay by 60dB (oftern referred to as RT60). RT60 is a measurement of sound level, it doesn’t necessarily correspond to the size of a given room — small room with hard surfaces can have RT60s longer than a larger room that is covered by sound-absorbing surfaces. Because it measures RT60 rather than the absolute end of the reverb tail, you might want to treat the reverb tail as a soft number; adjust it to taste rather than to a given number.
Pre-Delay
Pre-delay is the time between the initial signal and the onset of reverberation. For example, if you were to stand in a room and clap your hands, the sound of the clapping hands must travel (at 1,130 feet per second) to the floor of the room, the ceiling, and the nearest walls and back (these initial echoes are called “early reflections,” and become “reverb” when the individual echoes become so dense that single echoes can no longer be discerned). In general, the larger the room, the longer the pre-delay. Pre-delay can also be used to separate the onset of the reverb sound from the dry signal; in the days of analog recording, engineers would often take the reverb send though a tape echo to create longer pre-delay. Of course, you can also use pre-delay, for example, on a snare, to create something similar to an echo; it’s a fertile area for experimentation. Setting pre-delay properly can also have a huge effect on the intelligibility of vocals and speech.
Early Reflections
Early reflections are the direct echoes one hears in an acoustic space before the later reverb (made of of indirect echoes that have reflected off of multiple surfaces) become prominent. Combined with pre-delay, early reflections offer the most apparent sonic information about the size of a space. By changing the early reflection time and level in relation to the diffusion and RT60 time, you can create the sound of spaces that couldn’t exist in the real world. That doesn’t mean that they’re “wrong,” of course; if it suits the track, it can be perfect, even if (or especially if) it isn’t a sound that exists outside of the recording.
Diffusion
Diffusion is a term used primarily to describe indirect reflections — reverb density is determined by the amount of diffusion; a greater number of indirect reflections correlates to more diffusion. After the onset of early reflections, the indirect reflections — the diffuse sound — takes over.
Room Size
Most software reverbs link a number of parameters together to emulate a given room size — a smaller room, for example, would have a shorter pre-delay, faster (and louder) early reflections, and a lower density than a large room. Nevertheless, you can manipulate the room size to taste, then tweak the other parameters to suit the needs of a given track.
While it’s perfectly acceptable to simply choose a reverb preset and move on, if you can create the “perfect” preset for a given song, building the perfect reverb can put your mix into a perfect spot. Experiment!