Reverb is one of the most complicated effects to recreate, unlike echo which simply comprises multiple repeats of the original sound. Artificial reverberation is based on the way we hear sound in an acoustically reflective environment where our ears capture both the direct sound from the source plus thousands of closely-spaced reflections as the sound bounces between walls and other solid objects. Because these reflections arrive later than the original sound and also come from different directions, our brains process the sound to determine the type and size of the space.
Studio recording often takes place in an acoustically ‘dry’ acoustic reverberation so some form of artificial reverberation must usually be added to give the recorded sound a sense of space and to help it sit comfortably in a mix. Early reverb devices included springs, plates and even specially designed reflective rooms though in today’s studio nearly all artificial reverb is generated using digital electronics.
To appreciate what a tough job an artificial reverberator has, just think about what happens when a real sound occurs in a confined space. The sound moves outwards in the form of spherical wavefronts travelling at around 340 metres per second, rather like ripples in a pond, until it encounters a reflective surface where some of the energy is absorbed (no surface is perfectly reflective) and some is reflected back into the room. These reflections subsequently re-reflects from any surface they encounter, getting quieter all the time as their energy is absorbed. In larger spaces such as cathedrals, a significant amount of acoustic energy is also absorbed by the air itself.
The spacing between the first few reflections is important in helping or brains deduce the size of the space as the larger the space, the further the sound travels before it meets a reflective surface, other than the floor of course. The larger the space, the longer the delay between the original sound and the first reflection and the first few reflections also tend to be spaced further apart than in a smaller room. In reality we only perceive a few separate reflections before the multiple reflection paths cause the reverberant sound to blur into a dense reverb tail. We usually call these two components Early Reflections and the Reverb Tail, and though they are all created as part of the same process in real life, reverb algorithm designers often choose to synthesise them separately before blending them. The designers of more sophisticated artificial reverb units realised that when sound bounces from a surface, it is also diffused or scattered making the reflections sound less distinct and contributing to the complexity of the reverb build up.
Most structural materials reflect sound more efficiently at some frequencies than at others where most of the materials found inside typical buildings reflect low frequencies more efficiently than high frequencies. This means that in a typical ‘real’ space, the low frequencies continue to reverberate for longer than the high frequencies and this is particularly noticeable in large spaces with a long reverberation time.
We’re pretty much all familiar with spring reverbs from guitar amps that have them built in, but the first serious artificial reverb unit was the Plate where a thin sheet of metal hanging in a frame is driven into vibration using a magnetic transducer rather like a loudspeaker. Contact mics on the surface of the plate pick up the vibrations, and because of the way the sound reflects around the plate, these vibrations sound much like real room reverb, albeit with less well-defined early reflections. Remote-controlled mechanical dampers adjusted the reverb decay time while EQ is used to tweak the tone. These units were large, very susceptible to outside noise, and at the time, very expensive. Perhaps the most famous of all the studio plates is the EMT model 140.
The first digital reverbs created over 30 years ago really stretched the available DSP power where the designers generally created a complex network of multi-tapped delays, recirculating delays and EQ to approximate what happens in real life. A simple multi-tapped delay makes it easy to recreate very strong early reflections patterns where the spacing, relative level and tonality of the individual delays can be changed to emulate different room types. The recirculating delays recreate the dense reverb tail. This type of approach is said to be algorithmic and every designer has their own recipe for creating their own reverb sound. The most famous names in this area are Lexicon, TC and more recently Bricasti with notable vintage reverb units being made by EMT, AMS and Klark Teknik.
Today’s algorithmic reverb can come in either hardware or plug-in form where modern computers make light work of what was once a very demanding set of calculations. Note that while some reverb devices have true stereo inputs, many derive their reverb feed from a mono mix of the two inputs and then process this in slightly different ways to create left and right outputs that give the impression of stereo width. This mono-in, stereo-out approach makes sense since in an enclosed space reverb comes from all directions, even if the original sound source, such as a voice, exists at only one (mono) point within the room.
Different models offer different levels of user programmability but I’ll describe only the most important parameters. Reverb decay time is an obvious one but most units also have an adjustable pre-delay time to place the reverb further behind the original sound. No early reflections are hear until the pre-delay time has elapsed where values of between 50 and 120mS are often used in setting up vocal treatments. The room type is emulated by offering different early reflections patterns combined with different tonalities of reverb tail and there’s usually a way to affect the high and low frequency decay times separately to recreate the way natural materials absorbing high frequencies more effectively than low frequencies. Most reverb plug-ins, and hardware units, offer emulations of rooms, halls, chambers, plates and small room ambience (where the reverb is so short as to be barely noticeable) as well as some artificial effects such as gated reverb and reverse reverb. There may also be control over diffusion, which is the speed with which the early reflections build up into the dense reverb tail. Some units also include a density control to adjust the complexity of the reverb tail.
Lexicon reverbs incorporate a chorus-like modulation inside the reverb algorithm as this adds warmth and a sense of lushness to the reverb tail as well as preventing it from sounding metallic and ringy. Other algorithmic reverb designers have also adopted a variation on this idea.
Next week we’ll be looking at gated and convolution reverbs, and how to actually use all of these properly in your mix.