Music is meant to be heard in an appropriate acoustical environment. But many times, we have to record in a space that does not have the right sound to it. Artificial reverb is often the solution.
In this episode, I talk about good room sound in a studio, and the earliest methods of creating a reverberation sound. That started with spring reverb, then acoustic echo chambers, tape delay, plate reverbs, and finally the digital reverbs that are used today on most recordings.
I discuss the origins of those techniques for creating reverb, and the strengths and weaknesses of each, along with my impression of how each of them sounds.
I also provide my own insights into using reverb. What kind? How can we modify the basic sound? And how much reverb is appropriate. And maybe no reverb at all is best.
Thanks for listening, subscribing, and commenting. I am always looking for your questions and suggestions for future topics.
86 Reverb December 19, 2023
I’m Doug Fearn and this is My Take on Music recording.
First, thanks for sending in your questions and suggestions for future topics. You can email me at email@example.com
We live in world of complex acoustical environments. If you spend a little time really listening to the spaces where you spend your days, you will become aware of the many variations in reverberation that surround you.
You probably already know that your bathroom sounds different from your living room, and the kitchen sounds different from both. Your garage has yet another sound.
Outside sounds are different from inside. Outside could be in a city with stone, brick, and glass everywhere you look. That will be an acoustically-confusing environment. The middle of a forest has a totally different sound.
In the context of this episode, I refer to reverb as any kind of effect the room or space has on the sound. Even a closet full of clothes has a “sound” to it. It’s a broad definition of reverb, I know, but I want to include all the acoustical environmental things that influence how a recording sound. This does not include noise of various types, of course.
You probably know the fundamentals of what this is about. The sound source is reflected off of the various surfaces differently in every room, large or small, and every different outdoor environment. And variations in absorptive materials in the space influence the decay time of the reverb and affect the frequency response of the reverb. Diffusion generally improves the sound of the natural reverb.
Everyone knows what bad acoustics sound like. There are many ways to create a bad-sounding space. The sound might be annoyingly bright, or depressingly dull. There might be distinct echoes, or no discernable echoes at all. I covered a lot of this in Episode 84 on Designing Acoustic Spaces.
Studios are a special case, since the goal of recording is to serve the musical intent of the artist and to provide the listener with an enjoyable experience. We don’t want the listener to be annoyed or distracted by the sound of the space.
Similar requirements apply to our control room. We need to have an accurate picture of how the music is going to sound to the listener.
The best studios in the world tend to be fairly large and have a pleasant acoustic signature. The sound of the room is distinctive, and readily recognizable to anyone listening to records made in that space, even if they have no way to explain it.
Some studios are designed to be as dead as possible. That was a popular sound in the 1970s but not so much today. Many great records were made in very dead rooms. They had a sound that was different from those made in great-sounding rooms.
Today, many recordings are made in adapted spaces that are not very big. The sound of a small room is generally dreadful, so people often make those spaces as dead as possible. That’s a good solution, even though it may not be comfortable to perform in that space, compared to a room that provides the performer with some acoustic reinforcement.
Adding artificial reverberation has been part of music recording for over 70 years. There are many ways to do it, some better than others, some more practical than others.
The earliest artificial reverb device was invented by Bell Labs in the 1930s, to simulate the type of echo that occurred on long-distance telephone lines. That helped them design circuitry that would cancel out the echoes. You can still hear this on long distance phone calls, especially overseas calls.
The Bell Labs reverb used a helical spring, with transducers coupled to either end. The spring reverb sounded like a spring, with the “boing” sound you would expect. It wasn’t a good emulation of real reverb, but it did later find a place in electric organs and guitar amps. The sound of an electric guitar through a spring reverb is classic, and partially defines the sound of electric guitar. It’s lousy reverb, but like the crummy audio amplifiers in the guitar amp, it became part of the sound. It is what the instrument should sound like, for many players and engineers. I do not disagree.
But it’s not a good sound for almost anything else, in my opinion. It draws attention to itself as an effect, and if that is what you want, it works well. But not for the kind of recording that appeals to me.
A couple of methods to tame the spring reverb sound were tried over the years. The springs could be encased in a tube of oil. That meant the reverb unit had to oriented vertically, to prevent oil spill from the open end of the tube. That changed the sound, but it still sounded like a spring to me.
When I worked at WPEN in Philadelphia, there was an oil-filled spring reverb in one of the production studios. I never liked it, but I found that it made a great percussion instrument if you tapped on the oil tubes.
AKG tried a different approach to spring reverb. Their unit used torsional excitation of the springs, rather than the longitudinal method used in guitar amps. That was a minor improvement and made the spring reverb marginally more acceptable to my ear. I had one of those from the early 1970s, but it didn’t get much use after I got the EMT plate reverb, which I will talk about in a minute.
Professional tape machines can be used to create an echo effect. The heads on a tape machine are, from left to right, erase, record, and playback. If you record something on the tape machine, and take just the playback output, the signal will be delayed by less than a second. The exact delay depends on the head spacing and the tape speed. In the early days of recording music on tape, 15 inches per second was standard. That provided a delay of a bit over 100 milliseconds. The exact interval depended on the spacing of the heads, which varied by manufacturer. If you switched the tape machine to 7.5 inches per second, the delay would be doubled.
Not many machines were capable of 30 inches per second back then, but if you had one that did, the delay would be around 50mS.
It was common back in those days to have a system to change the tape speed to anything between the basic settings. The capstan motor, which controlled the tape speed, was designed to lock onto the AC line frequency, 60Hz in North America and 50Hz in the rest of the world. If you used a variable frequency audio oscillator, fed into a power amplifier that could drive the motor, you could change the tape speed over a wide range. That was used a lot on vocals back then. But it could also be used to change the speed of the tape echo to match the tempo of the song.
Used this way, there would be one distinct echo of the original sound. That was commonly used on vocals in the 1950s. However, if you fed a bit of that delayed signal back into the tape machine, the echo would have multiple copies. If the level was low enough, the echoes would die out after a couple of repeats. Too much feedback signal and the echo would grow in volume and quickly overload the tape machine and studio console. It was a fine balance to achieve the effect you wanted.
The playback of a tape is always different from what went into it. With good tape machines, properly aligned and maintained, the difference was small. But when you fed the tape signal back into the machine, each successive echo would sound worse and worse, as the imperfections of tape system were magnified over and over.
Eventually, dedicated tape echo devices were developed. Often these had multiple tape heads, some of which could be moved to fine-tune the delay. The more sophisticated of these machines had a multitude of controls to mix the playback heads in various proportions, and equalizers to tune the character of the repeats. And, of course, controls to adjust the amount of feedback.
In my early studio, I had one of these devices, made by a company called Surround Sound. It had a moveable record head and several adjustable playback heads. The tape was a continuous loop that went around a large wheel. The record head could be placed anywhere around the circumference of the wheel, to achieve echoes of only a few milliseconds up to several seconds. The tape speed could be either 10 or 20 inches per second. It was a very versatile machine that got a lot of use in my studio. You could even create bizarre effects by moving the record head during operation, stretching or compressing the sound.
Tape echo can be an interesting effect when used properly. You can achieve something similar with a digital delay, but there was something unique about the sound of tape echo that made it a classic sound.
Another early technique for creating artificial reverb was the echo chamber, dating back to the 1940s. This method uses a room, small or large, designed to have a moderately long reverberation time. That might range from one second to several seconds. Inside the echo chamber is a loudspeaker, and one or more microphones to pick up the reverberant sound. One major design goal is to minimize the direct pickup of the speaker. All we really want are the echoes in the space.
Echo chambers can have many variables. The size is important, although some echo chambers are surprisingly small. The Studio 2 echo chamber at Abbey Road is not much bigger than a small household room, with a low ceiling. Others, like those a Capitol Studios, are considerably larger. The size affects the sound and reverb time.
The chamber will usually be designed with very hard surfaces, to maximize the reverb time. Ceramic tile is a common surface. It is used on all the walls and ceiling, and often the floor as well. That treatment tends to make the echo chamber favor the high frequencies – the reverb is very bright.
Other dense materials can be used, like concrete or stone. Those surfaces are painted with a glossy paint, to promote the sound reflections.
To minimize the pickup of the loudspeaker in the chamber, highly-directional speakers may be used, often horn-type. We think of horns as the high-frequency – tweeter -- part of a speaker system. Generally, we do not want a lot of low-frequency reverberation on most pop music, since it will muddy up the sound.
Or you could use a typical monitor speaker with a horn and use equalization to tailor the frequency response. The equalization can be used in the speaker feed, or on the microphone, or both.
The speaker is typically aimed at a corner of the chamber, and the microphones are aimed at the opposite corners. You would think that a cardioid mic would be best to minimize direct pickup of the speaker, but almost all cardioid mics have strange off-axis frequency response, which will color the sound in an undesirable way. Omni mics are usually chosen, although bi-directional mics work well, too, since they have a deep null off the sides, and mostly flat response to off-axis sound.
Another way to help isolate the direct speaker sound is to build an L-shaped room, where the speaker sound does not have a direct path to the microphones.
More than one pickup microphone can be used to provide a stereo reverberation. I don’t know if has been done for mixing immersive formats like Dolby Atmos, but additional mics could be used to provide a more surrounding reverb effect.
On a recent session I was involved with, at Sweetwater Studios in Indiana, we used their indoor racquetball court as an echo chamber. The room had a beautiful, smooth and bright sound that was perfect for what we wanted. We used six omnidirectional mics around the room.
Over the years, people have used various rooms with different characteristics to make an impromptu echo chamber. I have used a large garage adjacent to my studio for that. It sounds like a garage, which is only appropriate for some types of music.
Most chambers had internal structures to diffuse the sound. Large concrete pipes, a foot or more in diameter and capped at the top was a cost-effective way to break up the sound in the room for a more realistic representation of a larger-sounding space. The pipe would have to be painted with several coats of hard, glossy paint.
Other things could go into the room to improve the sound. Wedges, angled surfaces, or even miscellaneous other stuff might help, as long as it was a hard reflector of sound and wouldn’t rattle or impose a resonance in the chamber.
As you can imagine, there are a lot of things to experiment with in an echo chamber. The speakers and their placement. The microphones, and where they are placed. And the equalization applied to either of those are just a few of the variables that a studio owner can experiment with. Often, the setup will be different for different types of music.
Rather than a room designed to be an echo chamber, people have used existing rooms, usually a bathroom. A typical household bathroom would not work too well in most situations, but a larger, commercial bathroom with tile walls can provide a usable chamber for a short and bright reverberation.
The amount of noise created in an echo chamber can be a problem. In the case of Abbey Road’s Studio 2 chamber, which was constructed outside the main building, neighbors complained of the noise during late-night sessions when that residential neighborhood in London was quiet and people found the strange sound of an odd combination of instruments and voices booming for hours during a mix to be intolerable.
In another case, at Sigma Sound Studios in Philadelphia, the echo chamber was constructed along a hallway just outside Studio A. It was L-shaped. People in the building were not particularly annoyed by the sound, but the chamber was not well isolated from the people in the hallway, so voices would sometimes be heard during a mix. That, of course, ruined the mix.
Another odd thing about Sigma’s echo chamber was that it was not well isolated from the studio. The music in the studio could leak into the chamber and confuse the sound of the desired reverberation. Sigma founder Joe Tarsia described an unusual situation in our conversation in Episode 26. The studio was built in the 1950s, prior to Joe acquiring it. Back then, everything was recorded live, in mono. There was no overdubbing and no mixing. The echo chamber needed to be used during the session, and had a unique sound due to bleed from the live performance.
But when Sigma opened in 1968, it was an 8-track facility and mixing was done after all the tracking. The echo chamber was not used during tracking.
Joe wondered why he was not getting the sound that the previous occupant was getting, until he realized that the live sound from the studio was leaking significantly into the chamber. Multitrack recording and mixing had eliminated that prior technique, and the sound was different.
Echo chambers can be a great sound and often defined the records made in studios that used them. But they took up considerable space. They had to be isolated from the studio facilities and the outside world. That made the acoustic echo chamber impractical for many smaller studios, and that includes most home studios, too. Unless it is a major facility, with a construction budget to match, the echo chamber is out of reach for most recording situations.
In the 1950s, the EMT company in Germany invented a new way to create artificial reverberation – the plate. This method uses a thin metal plate, usually stainless steel, roughly 40 by 80 inches The plate is suspended by springs inside a heavy wooden case. A special transducer is attached to the plate, at a point determined to be optimal, and driven by an amplifier. This causes the plate to vibrate.
One or two transducers, similar to contact microphones, convert these vibrations into an electrical signal.
Plate reverbs have a distinctive sound. The reverberation can be very good, mimicking a real, reverberant acoustic space. But setting a plate up properly takes a lot of work to get the optimum sound out of it.
I bought a used EMT 140 plate from a studio in New York that was closing down. We moved it down to my studio in Philadelphia in a pickup truck. It was heavy and awkward, but we got it moved into the studio basement and wired up.
The EMT uses a panel of sound-absorbing material, like fiberglass, parallel to the plate, to adjust the reverb time. The position of this sound damper could be adjusted from less than half an inch to several inches. The adjustment was made with a large metal wheel on the top of the case, with a pointer that moved across a calibrated scale to indicate reverb time.
Some models had an electric motor that moved the damper plate. That allowed remote adjustment from the control room.
An operating plate did not produce much sound that could be heard nearby, so they needed to be isolated in a separate room. Their bulk and weight also made them difficult to situate in small studios. Still, a plate reverb was much smaller than an echo chamber.
After trying it out on a variety of sounds, I was hooked on the sound. It was bright, like an acoustic echo chamber, even without any eq. The reverb time was very short, and moving the adjustment wheel didn’t have much effect on the sound until it was at the extreme long reverb time.
I was curious to see the inside of it, so I took off one of the side panels to take a look. I saw that one of the many separate panels of fiberglass had broken free of its mounting frame and a corner of it was touching the plate. It stayed in contact with the plate until you turned the wheel to longest reverb time. That explained what I was hearing.
I got some glue that seemed compatible with the materials and glued the loose panel back in its proper position.
When I listened to the EMT in its repaired state, its character changed entirely. It was not nearly as bright and reverb time was much longer. I didn’t like it as much.
So I took the unit apart again and un-did my repair. It stayed that way for the next 15 years that I owned it and gave my mixes a distinctive sound. One of the early digital reverb manufacturers even sampled my plate and incorporated it into one of their plate programs.
Plate reverb causes the reverb to become audible very quicky after the initial sound hits it. That is contrary to the sound in most real reverberant rooms, where there is a delay before the reverberation is heard. Most rooms also have a distinct “first reflection” that occurs before the diffuse echoes begin.
In order to simulate real spaces, it is often helpful to delay the signal going into the plate reverb by a short amount. The longer the delay, the larger the imagined room becomes. It was a common practice to delay the send to the plate reverb by 50 to 200 milliseconds, to enhance the effect. The only way to do this back then was with a tape machine. Adjustable tape speed allowed the engineer to tune the reverb delay to exactly what was needed to better simulate a real space, like a concert hall.
You could also take the delay from the tape machine and feed a bit of it into the mix to provide the “first reflection” sound of a real space.
In my studio back then, we always delayed the send to the EMT plate. We could use a tape machine, since we had several mono and two-track machines in the control room. Or we could use the Surround Sound tape delay machine, with its large number of adjustments.
For some very dense and powerful music, like early punk, I would often over-drive the delay tape machine to saturate the tape. This was like a limiter in the echo send, which increased the density of the reverb and added tape saturation to its sound. The result was an aggressive reverb that fit the intense music of that genre.
There was one more innovation in the basic plate reverb sound that finished the heyday of electromechanical reverb devices. That was the EMT 240 reverb that used a gold foil instead of a metal plate. The foil was similar to a condenser microphone element, only much larger. The gold foil reverb was a fraction of the size of the plate reverb. The gold foil was about a square foot, compared to around 22 square feet of the steel plate reverb.
And the sound was very different. I vividly recall first hearing the gold foil reverb at an AES Show in the 1970s. I was astounded by its realistic recreation of a concert hall sound. The reverb was much smoother than the plate, and warmer sounding, too. It was a very familiar sound to me, after growing up hearing a symphony orchestra in a great concert hall.
Alas, I never had one of those gold foil reverbs. Initially, the company had great difficulty shipping them without damage. The foil was very fragile. Eventually, that problem was mostly solved.
The gold foil reverb had just one sound, but it was a magnificent sound. It was less sensitive to ambient noise levels than the plate, but still needed to be isolated. It was so small that locating it in a quiet spot was much easier than with the 670-pound, bulky plate.
The era of that reverb was short, since EMT, and others, developed digital reverbs that made the old style seem far less desirable.
Spring reverbs, tape echo, acoustic echo chambers, and plates can sound magnificent. But each is mostly a one-trick pony. They produce a great sound, but there isn’t much variation available to the engineer without some additional work. When combined with delay, limiting or saturation, and equalization, a unique reverb sound can be created.
When analog noise gates first became available in the early 1970s, yet another tool was available to modify the basic sound of these hardware reverb devices.
Everything changed in the late 1970s when the first digital emulations of reverb became practical. A sales engineer from Lexicon was making the rounds of studios in our area and he brought the first iteration of the 224 digital reverb. It was big compared to digital reverbs that came later. A large rack-mount box contained an early microcomputer, extremely rudimentary compared to even the simplest microprocessor device today. It had large and loud fans to cool the digital circuits. It was not possible to mount the box in the control room because it was noisy.
The remote control was much more compact and, of course, silent.
The number of different reverb types the device emulated was mind-blowing back then. I ordered one after playing with it for an hour.
The various reverbs were created by sampling and modeling real acoustic spaces. I couldn’t wait to start using it.
Interestingly, after my engineers had used the new digital device for a month or two, all of us started gravitating back the plate reverb with our own enhancements. That was especially true for the featured tracks on a recording, vocals, in particular. It just sounded better to us than the digital equivalent.
We continued to use the digital reverb, especially on string sessions. But there was something about the organic, analog sound of our plate that kept drawing us back to it.
My opinion of digital reverb stayed that way for many years, until computing power allowed the devices to have higher sample rates and greater bit depth.
These days, all my reverb comes from dedicated hardware digital reverb devices. The technology, and the modeling have improved to the point where I am totally satisfied with the sound they provide.
Having hundreds of different reverbs available at the turn of a knob is something we could only dream about in the past. And the ability to fine-tune the basic sound to get even closer to what is needed is another major advantage.
I’m a big believer in cohesive reverb that uses the same sound for all the instruments and voices. That works best for me. But I do usually use two different digital reverbs on my mixes.
One of them is the main reverb, chosen for its appropriateness for the song. I use many different reverb presets for this, ranging from concert halls to churches, to plates, to acoustic echo chambers.
The second reverb sound is a good studio room sound emulation. I find I only need a bit of that, and often only on some tracks, to create the sense of a nice recording studio. My studio is relatively small and doesn’t have much usable character. The room emulation allows me to fine-tune the sound to create a better acoustic environment.
The amount of this “room sound” can often be vanishingly low in the mix. You would be hard-pressed to pick it out. But when I mute that sound, the mix suddenly sounds less interesting and appealing. I am amazed at how little of the room it takes to make a difference.
Today, most people use plug-in reverbs, which can sound very good. The era of electromechanical reverbs, like the EMT, is mostly over. The dedicated hardware digital reverb is still a viable alternative and offers some advantages over a plug-in.
Plug-ins are available that emulate all of the reverb sounds I have discussed. Some are sampled from classic echo chambers or unique plate sounds. Others duplicate the sound of the world’s best-sounding studios, concert halls, and churches. There are even emulations of various acoustic environments that can be used as an effect of certain sounds. I always liked the “Ice Rink” sound on the Quantec reverb for its long pre-delay, extremely bright sound, and fast decay. It’s not good for most things, but sometimes it is just the sound I need.
The hardware digital reverbs also have emulations of various outdoor spaces that can be useful. Or the sound of a huge enclosed space with reverb times that go on for tens of seconds. Special purpose, but it might be just what you need. I presume there are plug-in versions of many of these unusual acoustic spaces.
Almost everything you could want is out there to use, and at a tiny fraction of the price of the actual source. And plug-ins take up no room, need no isolation, and never need maintenance.
Regardless of source of your reverb, learning to use it properly takes some time and dedication. I would urge you to explore the settings available to see if you can tweak the stock reverb sound to better match what you want for a particular song. It’s easy to get lost in the parameters, but you can always re-load the stock sound if you find yourself drifting off into something that might sound cool but not appropriate for the music.
The amount of reverb is also important. It is easy to become enamored by a sound you discovered or modified. If a little sounds good, how about bringing it up a bit more? That may sound better, but you might regret that decision after some time has passed.
In most music, the reverb should just hint at the acoustic environment. It is the rare song that benefits from a large dose of reverb.
I have never listened to a mix I did, perhaps decades ago, and said to myself, “I wish I had used more reverb.” That just doesn’t happen. Sometimes I think, “Why did I use so much?” Other times I hear that the reverb I chose to use was not optimum. That may have been because I did not have access to the reverb that was perfect for the song. Today, we have so many different reverb sounds to choose from that you almost always can find what you are looking for.
And finally, there is always the option of no reverb at all. The sound of a totally dry recording is often the best choice for a particular effect you are seeking. This can work especially well in a situation where the contrast between a lot of reverb and none brings drama and interest to a mix.
With automation in your mixes, you can vary the amount of reverb through the song, when that is necessary. You can change from one reverb sound to another. You might use one sound for the verses and another sound for choruses. And maybe the bridge is completely dry. As always, the music tells you what it needs.
Thanks for listening, commenting, and subscribing. I can be reached at firstname.lastname@example.org
This is My Take on Music Recording. I’m Doug Fearn. See you next time.