The Sweet Sound of Reverb

Issue #28
November 02, 2005

Is it Music to our ears or just another overused studio toy?

Re·verb[r-vûrb] The remainder of sound that exists in a room after the source of the sound has stopped is called reverberation, sometimes mistakenly called echo (which is an entirely different sounding phenomenon). We've all heard it when doing something like clapping our hands (or bouncing a basketball) in a large enclosed space (like a gym). All rooms have some reverberation, even though we may not always notice it as such. The characteristics of the reverberation are a big part of the subjective quality of the sound of any room in which we are located. Our brains learn to derive a great deal of information about our surroundings from the sound of a room and it's reverberation. Consequently it is necessary to have the proper type and amount of reverberation on recordings in order for them to be aesthetically pleasing or to sound natural to us. This can be accomplished with careful microphone placement, but it is often necessary to employ artificially created reverb. To create reverb, a device known as a reverb unit is employed. Reverb units have historically come in many shapes and sizes, and have used many different techniques to create the reverberation. These days most of the reverb units employed throughout the world are digital, where the sound of the reverb is generated by a computer algorithm and mixed with the original signal. We will be discussing other types of reverb units in the future. View other terms

In the real world, reverberation is an important tool our ears use each day. Sound waves bouncing around man-made structures, which can range from back street alleys to gothic cathedrals, give us important clues to the size, complexity and proximity of our environment.

You'll notice I wrote "man made structures" here. The reason is simple: Reverb is pretty rare in the natural world. If you stand on a beach or walk out into a meadow and then pop a balloon or fire a starter's pistol, well you're not going to hear anything most of us would think of as reverb. Maybe some sort of distant echo. It's really not until you wander into a natural enclosure - let's say a small canyon or a cavern - that significant reverberation clues occur naturally.

Basically, reverberation is the reflection of a sound wave from solid surfaces to our ears. When you think about it, reverb has nothing to do with the original sound at all. In the studio, we'd call any sound without reverb "dry." Once you start adding reverb to a particular sound - and let's just agree that, for the sake of this discussion, our original sound is a drumstick hitting a snare drum - you change the character of the sound. You can change it subtly or dramatically. In the studio, we call the reverberation the "wet" signal, which can then be combined with the original dry sound.

In the real world, reverberation is actually a whole lot of primary and secondary reflections hitting our ears. If a particular environment is sufficiently complex, such as would be the case in a huge Baroque cathedral for instance, all the reflected sound waves produce a wash of echoes that decay slowly, so that our ears hear the phenomenon we call reverberation. Such reverb consists of hundreds or even thousands of echoes. In musical terms, we tend to think of echo and reverb as separate and distinct effects, however they are just different ways of hearing reflected sound.

ACOUSTICS AND REVERB: A BRIEF HISTORY

If you're really only interested in the theory and application of convolution reverb, you can skip a lot of this and go right down to the "Convolution Revolution" section. For the rest of you (particularly those fairly new to recording technology), I think you'll gain some valuable insights into just how we got to this point in music technology, so I'd definitely encourage you to at least scan the next few sections.

Naturally, the earliest reverberation was produced when musicians played in concert halls, churches or even palaces. We're talking about many hundreds or even thousands of years ago. Eventually, audiences not only came to judge the quality of the musical performances, but also the environments in which musicians would play. It probably didn't take mankind long to figure out that a richly reverberant hall was far preferable to a small, acoustically dry parlor.

Of course, there is no way of knowing exactly when this was. We do know that the earliest room designed to function as a public concert hall was in London in 1678. It was built by a group of musicians and could accommodate an audience of some 200 people. The magnificent Pantheon, located on Oxford Street, opened its doors in 1722, while the Holywell Music Room in Oxford (which was restored in 1960) opened in 1748. For the next few centuries, composers and musicians, along with the wealthiest "patrons of the arts" began to seek out the best venues for performances.

No discussion of concert hall acoustics and reverberation would be complete without mention of the physicist, Wallace Sabine, the first man to actually study the science of room acoustics. While teaching at Harvard University in 1895, he undertook the challenge of turning an "acoustically impossible" lecture hall into a usable room. Based on information gathered during his research, he actually developed an equation that made it possible, for the very first time, to consistently predict reverberation time in any room and thus the "musicality" of such a space. Personally, I find it amazing that such a discovery took so long to occur.

In any case, armed with this newfound knowledge, in 1900 Sabine took a job as "consultant for the New Boston Music Hall." After analyzing the initial designs, he quickly determined that it would be an acoustical nightmare. Based on his experience, a new design was developed which led to the construction of Boston Symphony Hall, which is still considered today as being among the most musical concert halls in the world.

EARLY USE OF ARTIFICIAL REVERB

In the earliest days of what we would call modern recorded music, reverberation was rare. Some studios actually built small, highly reflective rooms in which a separate speaker and microphone could be placed. As music was played, the original signal was sent to the speaker and then returned via microphone to the "mixing desk" where it could be mixed with the dry signal. Bill Putnam Sr. of Universal Audio fame was the first to use the technique on a 1947 recording by The Harmonicats called "Peg o' My Heart." The acoustic chamber used for the reverb? His bathroom! Eventually, Putnum constructed a more sophisticated chamber in his Chicago studio.

When Laurens Hammond introduced the very first Hammond Organs in the late 1930s, the sound was an immediate sensation, but Hammond realized that the sound was too dry. Using an electromechanical device developed by Bell Laboratories that used oil-damped springs to produce delays, he eventually had a unit that stood four feet high. This wasn't a problem at the beginning, since the early Hammond Organs had huge "tone cabinets." But as time went on and the cabinets became more compact, these original oil-filled units became impractical.

Ultimately, another Hammond engineer took on the task of creating an artificial reverb that was "no larger than a brief case." So successful was his design that in 1960 it became the Accutronics Type 4 Reverb, Now Hammond had a great reverb for his organs and a design he could market to other companies that needed a compact reverb unit. One of his first customers was Leo Fender, who first put the Accutronics Reverb into his new Fender Twin Reverb. Many early home studios employed spring reverbs derived from the original Hammond designs. My own utilized an ElectroVoice unit that had a fairly decent sound for the late 1970s.

The first real breakthrough in professional quality artificial reverbs came in 1957 when a German company, EMT, introduced the Plate Reverb. As the name implies, the unit is made from a very large, thin piece of sheet metal suspended from a steel frame by spring tension at each corner. An electrical transducer mounted at the center of the suspended plate induces vibration, which creates the effect. A damping plate controlled by a servo motor allowed adjustment of the reverb.

Exactly how this particular form of reverb came into being may well be lost in the mists of time, but this was clearly the best artificial reverb available in the late 1950s and all through the 1960s. Listen to almost any recording by The Beatles and odds are you are hearing the classic EMT 140 plate reverb. But you will also hear what is clearly some tape delay on many of the group's vocal tracks.

Tape delay results in a discrete echo or series of echoes and was first pioneered by Les Paul and used in his 1950 recording, "How High the Moon." That's right, the man whose name may be on your guitar was the first to realize that the space between a tape recorder's heads caused a delay in the sound. To increase the delay, he actually tied together two tape recorders and with the advent of variable speed motors, Les found he could match the echoes to the tempo of the music. While this isn't really reverberation, the tape delay effect - sometimes called "slapback echo" - quickly became a music industry standard and was often used in conjunction with the early plate reverbs.

FROM ANALOG TO DIGITAL

By the late 1970s, analog delay units became a reality, though their signal-to-noise ratio and frequency response was nothing to write home about. I was living in Green Bay, Wisconsin when I bought my very first analog delay, so I know that was in either 1979 or 1980. By the early 1980s, there were lots of delays available, in both rackmount and stompbox configurations. By 1984, Ibanez released the DM1000, the first digital delay affordable enough for us guitar players. I still have the unit and it sounds great.

It's a shame that nobody thought to carefully document the evolution of effects processors from their initial introduction to the era of the modern multieffects modules. As best as I could determine from hours of scanning an endless series of web sites, the first commercially available digital reverb appears to be from Lexicon and it made its debut at the 1978 AES Show. Most pro studios were quite content to stick with their plate reverbs, but that would quickly change.

I do recall the Yamaha Rev1 making an appearance not long after the initial introduction of the "hotter-than-a-firecracker" DX7, which would have been about 1984, but I can say for certain that the first digital reverb that was available to most home studio owners was the Alesis XTc (ecstasy, get it?) of 1986. Eventually, Alesis unleashed a wide assortment of digital reverbs, the most ubiquitous of which was the MIDIverb. To this day, almost everyone I know who has a studio has some incarnation of that particular unit.

In fairly short order, almost every manufacturer jumped on the bandwagon and by the late 1980s, there were digital reverbs to fit every budget and by the early 1990s, those original EMT plate reverbs were becoming museum relics, though a few professional studios have held onto theirs, particularly for vocal applications. Not surprisingly, there are engineers and producers who still swear by the sweetness of a real plate reverb.

Today, hardware reverbs come in all sizes and price ranges. Some are dedicated reverbs while others are multieffects processors that include all the popular effects flavors like chorus, delay, phaser and flanger. You'd be hard pressed to find even a modest project studio that doesn't have a rack full of outboard processors. And now with the advent of powerful personal computers, high quality reverb is available by using plug-ins or even from within host applications.

Reverb Plug-ins:

Waves IR-1 Native	TASCAM GigaPulse	Audio Ease Altiverb	TC Electronic DVR2	TC Electronic NonLin2
Convolution Reverb Plug-In with Full Traditional Parameter Control	Convolving Reverb Processor with Microphone Modeling for VST	Sampled Acoustics Processor Native Plug-In. MAS/VST/RTAS/AU (Mac)	DVR2 Vintage Plate Reverb Emulation Plug-in from System 6000 for Pro Tools	Stereo Effects Reverb Plug-in from System 6000 for TC PowerCore
View	View	View	View	View

Sonnox Reverb	Digidesign ReverbOne	Digidesign ReVibe	Trillium Lane Space	TC Electronic NonLin2
"Oxford" Reverb Plug-In For RTAS and AU	Stereo Reverb Processing TDM Plug-in	Room Modeling Reverb Plug-in for Pro Tools|HD Accel - TDM	Convolution Reverb Plug-In for Pro Tools LE systems	Stereo Source Reverb Plug-in from System 6000 for PowerCore
View	View	View	View	View

THE PAYOFF: CONVOLUTION REVERB

Gang, here's where things really start getting interesting. Most of you have heard about convolution reverb and others have even tried it out on their own systems. Without going into a long discussion about the way convolution technology can accurately capture all the nuances of live music played in a specific recording venue, let's all agree that with convolution reverb, we can create recordings that sound very much like they were made in some famous acoustic spaces, such as Boston's Symphony Hall or the Amsterdam Concertgebouw.

The way this happens - and pardon me for simplifying - is that a recording is made of a loud percussive sound in a real hall, chamber, cathedral or heck, even a nicely tiled bathroom. Normally, engineers pop a balloon or fire off a starter's pistol to stimulate all the reverberant characteristics of a particular location. The resultant file is known as an impulse response or IR. This may be saved as a 16-, 24- or even 36-bit recording at sample rates of up to 96kHz.

To illustrate this concept - and here's where the real fun starts - I collected several percussive samples from my own Stratus Sounds archives: One is a plain old Ludwig maple-frame snare drum hit (that sample is called SNARE) and the other is the same drum struck "cross-stick" style (that sample is called RIM). Both were miked with a Crown SASS-P stereo PZM microphone. Then, using BIAS Peak's Impulseverb, I placed the dry recordings in a variety of sampled acoustic spaces. That should give you a general feel for just how impulse response files can change the character of a simple sound.

Next, I got out my hand carved, Peruvian pan flute and recorded a short "performance." Again, I used the BIAS Peak Impulseverb to process the audio, and that sample is called PAN. In particular, you'll want to listen to the way each sample's reverb "tail" decays. It's really interesting to hear how certain acoustic spaces can add their own unique tonal coloration to a particular sound.

If you want to give convolution reverb a try, no problem. More and more software products include it in their design (MOTU's Symphonic Instrument is just one example). There are also dedicated convolution reverb plug-ins available such as Altiverb from Audio Ease and IR from the folks at WAVES. Give your Sweetwater Sales Engineer a call for assistance in choosing the best product for your needs and budget!

REVERB; USES AND ABUSES

Music sounds better with some reverb. I won't argue that for a second. However, applying too much reverb creates a horrid mess. Even worse, when different reverbs are applied to different tracks in a multitrack recording and then all mixed together, the end result is sort of like putting a bunch of things you like to eat in a blender and hope you'll end up with something tasty. Odds are you won't.

The late Frank Zappa once said something to the effect of: When you put a control on something that goes up to 10, people will set that control on 10. This could be a tone control on a guitar amp or the dial that sets the amount of chorus on an effects pedal. Yes, sometimes 10 sounds great. But when you set everything on 10, you'll end up with the audio equivalent of a train wreck.

Over the years - okay, let's be honest, decades - that I have worked with people in their studios, I always wait for that moment when their hands reach over to the reverb control. Invariably, they add too much, perhaps in a vain attempt to cover up an inferior vocal track. I always try and patiently explain why too much reverb won't fix anything. I also tell them exactly why a hall reverb on the keyboard track doesn't make sense when there's a plate reverb on the vocal and a room reverb on the guitar and so forth. It's not that I'm a genius - I'm not (as many can attest) - but I've spent many decades learning how to produce a superior recording and a really good mix.

Sometimes they listen. Sometimes they don't. But I can say with complete confidence that the people who listen to my explanations and then really listen to their recordings actually get it! By applying disparate reverbs to any number of tracks, they totally confuse the listener. The final product comes out muddled and smeared all across the stereo field.

In the "old days," when there was only one reverb unit available (like the EMT plate), recordings had a more homogeneous sound. With the proliferation of reverb processors and their 100 or so presets, it's far too easy to choose something different for each track, simply for the novelty factor.

Hey, I know you love your reverbs and your multieffects processors. But do me a favor and go back and listen to an assortment of your various mixes and ask yourself, "Does this sound real? Do I get the sense that there's a group of musicians up on stage in front of me playing a really cool song?" If the answer is yes, congratulations! If the answer is no, I suggest you look no further than your liberal use of reverb as the biggest offender.

Don't feel bad. We all learn through experience. Heck, I made more than my share of mistakes during my 35 years of recording experience. As long as we learn from these mistakes, our recordings are only going to get better. And with the advent of convolution reverb, it's even more important now to consider exactly where you are placing each instrument in the final mix.

At this point, let me apologize for the length of this particular article. It's a topic that I feel is particularly relevant given the availability today of so many audio processing tools. Just like you, I often find myself wanting to use a particular effect simply because it sounds good or different or just plain weird. But like the proverbial "kid in a candy store," it's far too easy to have too much of a good thing.

Questions? Comments? Discuss this article in Jim Miller's Forum

Questions, comments, rants, suggestions, unwanted ‘62 Stratocasters and any other form of correspondence can be addressed to jim_miller@mindspring.com.