Home -> Magazines -> Issues -> Articles in this issue -> View
Effective Reverb (Part 1) | |
VocalsArticle from Recording Musician, March 1993 |
Paul White discusses how to use reverb in the context of a mix, this month focussing on lead and backing vocals.
Reverb is still the most powerful of all studio effects — yet if used carelessly, it can ruin an otherwise successful recording. Paul White examines the key parameters of reverberation and relates them to studio production.
Reverberation is much more than just an effect; it is an integral part of the world around us and plays a large part in our survival mechanisms. If this last statement seems a little strong, consider the plight of prehistoric man, stuck in a cave after closing time with no torch! The acoustics of an enclosed space tell us a lot about the size and character of that space, even when we can't see it, and if our prehistoric friend happens to yell at the floor at the right time and be greeted by a particularly long reverb coming up from below, the chances are that he's just saved himself from falling into an inconveniently placed bottomless pit! For millions of years, natural reverb did its stuff; then, a scant decade or so ago, someone decided that a digital, electronic equivalent would be a nice idea, and since that time our perception of reverb has changed drastically.
Current digital reverb units can produce a passable approximation of real acoustic spaces, but they also have the capacity to create reverberant characters that simply could not occur in nature — what bizarre set of natural parameters would result in a reverse or gated reverb sound? Gimmicks aside, digital systems also allow us to produce reverbs that are longer and brighter than anything normally encountered in the real world, so how do we use it without allowing it to take over our music entirely?
Back in the '60s and 70s when serious multitrack recording really got off the ground, the standard source of reverb was the so-called echo plate. This was basically a simple mechanical device comprising a large sheet of thin metal suspended in a framework and driven into vibration by a voice-coil assembly similar to the back end of a loudspeaker. Contact mics glued to different places on the surface of the plate could be used to pick up the vibrations which, after amplification, could be fed back into the effects returns of the mixer. Because the mics were stuck in different positions, their outputs were slightly different, so panning them left and right would produce the illusion of stereo reverb. The way in which a flat plate vibrates isn't quite like a real room, as the reverberation builds up very quickly with very little in the way of early reflections. The mechanical resonance of any freely suspended plate also meant that the sound had a slight metallic coloration. Nevertheless, the plate produced a very dense and musically satisfying reverb that didn't suffer from the twanginess of simple spring systems, and with a touch of EQ it could be made to sound surprisingly natural.
Without any form of damping, a typical plate might produce a reverb of around three or four seconds, but if shorter reverb times are needed, it can be mechanically damped. The more professional systems were fitted with remote-controlled mechanical dampers comprising felt pads or similar, while the simpler plates were damped by hanging a tea towel over them! In this apparently low-tech climate, some of the best pop and rock records ever made burst upon the scene.
Even before digital reverb appeared, our perception of reality had been changed because we'd come to accept the plate reverb sound as being 'right' for music. And so it was that when the first digital systems came upon the scene, customer demand saw to it that they included plate simulations, as well as the ability to mimic rooms and concert halls.
Because of the way in which digital reverberation works, it is easy to produce very strong early reflections patterns, something a plate didn't do, and though these reflections are an integral part of the character of any natural enclosed space, digital systems allow us to over-emphasise them quite dramatically. Additionally, the early reflections in a digital system — at least in all the affordable ones — seem much more distinct than those that occur naturally. This can be both a curse and a blessing, depending on the application; over-emphasised early reflections on percussive sounds, for example, can result in a very grainy sound that is quite unlike real life. On the other hand, smooth sustained sounds such as flutes, synth pads, string samples or even vocals can be considerably enhanced by using a reverb patch that is deliberately coarse.
Used as a mono-in, stereo-out device, the reverb unit produces a close approximation of the way in which reverb is perceived in real life, with the original sound appearing to emanate from a single point and the reverb appearing to come from all sides. This creates the effect of space in one dimension — width — but it is also possible to create the illusion of front-to-back perspective or depth, albeit in a less precise manner. To see how this might be done, imagine yourself in a concert hall with a solo musician playing on the stage in front of you. Because you are close to the performer, you will hear a higher proportion of direct sound than you would if you were at the back of the room. One reason for this is that reverberation in an enclosed space doesn't follow the normal inverse square law for loudness with distance but instead is at nominally the same level throughout the room. The direct sound, however, does obey the inverse square law, so the further you are from the source, the quieter it sounds.
At the front of the hall it's also noticeable that early reflections build up quickly and seem to be more strongly defined — whereas at the back of the room the early reflections appear to build up more slowly, giving the reverb a noticeably slower attack time. The reflections heard from the back of the hall also appear to be more diffuse and softer in tone, as they have had to travel further and encounter more obstacles before reaching you. To confirm the psychoacoustic validity of this observation, try adding a single delay with a 60mS delay time to a vocal or guitar sound. This creates a slapback effect that immediately conjures up the image of being close to the sound source in a reflective room.
If a sound source is relatively distant, the reverb will have less high frequency content, as it suffers from air absorption and, as a rule, most structural materials absorb high frequency sound energy more effectively than they absorb low frequency energy. That's why naturally occurring reverberation in buildings becomes noticeably less bright as it decays.
A sense of distance and size can also be created by putting a delay between the original sound and the reverb that follows. This convinces the hearing system that the reflected sound has had to travel a long way before reaching the listener, and is a simple but very effective means of adding extra depth and dimension to a sound. The delay can be created by putting an ordinary DDL (Digital Delay Line) before the reverb, though many programmable reverb units have the facility to add pre-delay as one of the user adjustable parameters. In general, pre-delay times of a quarter of a second or more are useful, though the most useful range is arguably between 60 and 120mS.
General theory is useful when it comes to understanding why things sound the way they do, but now the time has come to look at how reverberation can be applied to the various sounds in a typical mix.
For some reason, vocals sound rather better with reverb than they do without it. One theory is that we humans like to be able to hear some kind of reference pitch, and by adding echo or reverb to a sound, we are constantly hearing the new notes being sung overlaid on the reverb or delay of the immediately preceding sounds. This is borne out in practice, as there's something very uncomfortable about listening to a totally dry voice, even an exceptionally good one, whereas even an average singer sounds tolerable when singing in the bath!
With lead vocals, there are several conflicting factors that need to be balanced when we add reverb.
- In the case of lead vocals, it is important that the words themselves are clear — but adding reverberation reduces intelligibility. The longer the reverb, the more the intelligibility is reduced.
- Bright reverbs can sound exciting but emphasise sibilance and can be annoyingly distracting if mixed too loud.
- Too little reverb may make the voice sound unattractive, and may also make it sound as though it's 'stuck on' to the backing rather than sitting comfortably in the mix.
- Long reverb times sound flattering, but they tend to fill up the spaces in a mix that create the all-important contrast; the space is just as important as the music.
- The traditional place for a vocalist is at the front of a band, so we normally try to make the singer stand out at the front of the mix. However, adding a lot of reverb creates the impression of distance, which tends to push the singer back.
As you listen to different records, you'll come across dozens of different reverb treatments, all of which create a specific effect.
- To make a vocal sound close and intimate, you may need to choose a relatively short setting (less than 1 second), and depending on the kind of environment you want to create, it may need to be fairly bright to suggest an intimate club rather than a large concert hall. However, if the reverb is too bright, it will bring out any sibilance in the voice and may become annoyingly obtrusive.
- A high level of early reflections with a relatively low density (large spacing) can also suggest an intimate environment, as it fools us into perceiving the existence of nearby reflective objects. A plate or small room simulation works well in this context, as do programs that provide mainly early reflections with little or no following reverb.
- If a bigger sound is sought, then it may help to introduce a degree of pre-delay to provide a little separation between the original sound and the following reverb which, typically, may have a decay time of anything from 1.5 to 4 seconds. If the original sound is kept quite dry, this can still maintain a degree of intimacy providing the reverb is mixed low enough not to bury the vocal. It may also help to keep the early reflection level low on reverb units that provide parameter access to the early reflection level.
- Those who have the luxury of more than one reverb unit might like to try using one of them to produce an intimate, (short, bright with little or no pre-delay) effect and a second unit with maybe 200mS of pre-delay and a longer reverb time to create the sense of depth. It may even help to keep the short reverb in stereo but pan the longer reverb to one side or the other to lend a little movement to the sound.
- Really big reverb sounds can work in songs that have the space to give the reverb room to coexist with whatever else is going on, but a useful dodge is to feed the reverb output through a gate or compressor configured as a ducker. For more on ducking, see the article on compressors elsewhere in this issue. Figure 1 shows how a ducker can be patched in to pull down the level of the reverb when the vocalist is singing, but will allow it to swell back in during pauses between phrases or at the end of lines. The ducker should be set with an attack time of around 200-500 mS and a release time in the same range so that the reverb level changes don't occur too abruptly. The actual release setting will need to be fine-tuned by ear depending on the material — if it is too short, the gain changes will be too abrupt, but if it's too long, the reverb will have died away before the ducker can fade it back up again. The range of the ducker should be set somewhere between 10 and 15dB to start off with, but again, the actual value depends on your taste and the style of the song.
There are also some rather neat tricks involving the combination of ducker, reverb and delay, but I think those particular techniques deserve an article all of their own.
A somewhat different approach can be taken with backing vocals as, by definition, they don't need to sound as forward as the lead vocal.
- Using a setting with prominent, widely-spaced early reflections will thicken the sound considerably, and you can usually get away with a longer reverb time than you use on the main vocal. By the same token, choosing a slightly less bright reverb will help the backing vocals sit further back in the mix, creating that all-important sense of perspective.
- It can help to thicken the sound further if the feed to the reverb unit is first treated using a chorus unit or a pitch shifter set to produce a mild detuning effect. Even a flanger placed before the reverb isn't too unsubtle, as the characteristic sweep of the flanger is diffused by the reverb and the dry portion of the sound remains unchanged. The result is to add a rather nice shimmer to the sound but without making it too gimmicky.
- Another little-known technique is to connect a gate before the reverb unit and set the threshold so that only high level sounds are treated. I first heard of this trick being used during an interview with Tony Visconti a few years ago, when he described how he'd produced a vocal effect on a David Bowie album. He'd used real, acoustic reverb from the back of a large hall, but a digital reverberation unit makes the whole task so much easier. Aside from the novelty of having reverb only on high level sounds, there is a side benefit in that the mix becomes less cluttered. Again, the technique can be expanded upon if you have two reverb units by combining a small room sound from one reverb unit that is connected normally, with a longer setting on the second reverb unit brought in by the gate.
Amateur demos often suffer from too much reverb, which has the effect of filling all the available spaces in a mix, and preventing it from breathing. Reverb isn't musical Polyfilla; it's a very powerful effect which often works best when used sparingly. If you listen to a selection of well-produced records, you might be surprised at how restrained the use of reverb is on many tracks, and on those that use longer or more obvious reverb treatments, you'll probably find that the arrangement has been tailored to provide the necessary space for the reverb to work. There are always exceptions, which confirm the old adage that nothing succeeds like excess, but in general, reverb is like seasoning — too much spoils the dish (unless it's a curry, of course!).
I'll be looking at some more specialised vocal reverb and delay treatments in more detail in the near future, but next month I'll examine how reverb can be used within a mix to enhance the sound of drums, guitars and synthesizers.
Read the next part in this series:
Effective Reverb (Part 2)
(RM Apr 93)
All parts in this series:
Part 1 (Viewing) | Part 2
Sound Bites - Production Tips & Techniques |
Hands On: Lexicon PCM70 |
![]() Kit Noise |
Digital Effects - A Guide to Digital Reverbs, Delays and Multi-Effects Units |
Experimenting With Analog Delay |
![]() How To Do Tricks With Time |
Phase The Music |
Music Modules for NON-Keyboardists |
Workbench - Signal Processors - the saga continues |
Delaying Tactics |
History |
The 'Andy Summers' Chorus |
Browse by Topic:
Feature by Paul White
Previous article in this issue:
Next article in this issue:
mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.
If you value this resource, you can support this project - it really helps!
New issues that have been donated or scanned for us this month.
All donations and support are gratefully appreciated - thank you.
Do you have any of these magazine issues?
If so, and you can donate, lend or scan them to help complete our archive, please get in touch via the Contribute page - thanks!