There are a huge number of variations in the controls on a reverb. It’s a bewildering array of controls, in fact, with few reverbs sharing the same exact set of parameters.
This is, for the most part, due to the differences between the various methods of creating reverb.
Modern reverbs, in particular, have been following a trend of adding features such as distortion, compression and modulation controls. An example is the recent Waves plug-in from their Manny Marroquin collection, which offers the user some measure of control over all of these aspects of the sound.
There may be various EQ and filtering options that are globally applied to the reverb, as well as more subtle and more realistic LF/HF Damping controls that are used to increase either high or low frequencies gradually during the time the reverb tail takes to die away.
The possible combinations of controls on reverb is staggering. Fortunately, most reverbs boil it all down to a limited control set, and it varies with type of reverb process.
HARDWARE DIGITAL REVERBS AND WORD CLOCK
It’s really important to be aware that any digital hardware reverb you connect with a digital send/return cabling arrangement should be correctly synced up with the studio digital master clock reference, whatever that may be.
Typically, the digital reverb will automatically sync to the incoming digital signal from the DAW, but in almost every case the DAW must be told in advance to sync to the incoming reverb return(s) or else you will have a serious problem that renders recordings either unusable (due to random clicks and pops appearing in the audio) or downright impossible to make in the first place (refusal to operate by the devices).
Using digital connections means that the clocks must run in sync on both digital devices, the DAW interface and the reverb.
You ignore Word Clock features on any digital audio device at your peril. Seriously.
This can be problematic in situations where multiple devices are to transfer digital audio in a system simultaneously. There are only two possible solutions that will prevent destructive clicks and pops appearing in your audio recordings.
First, connect analog and scrap the digital connection back to the DAW. Problem solved, at the cost of going analog and converting more than necessary.
It’s much better if you can stay digitally connected in both directions and not convert the file to analog at all. The caveat is that you MUST be in digital sync to pass audio correctly. It may not pass audio at all, in many cases, when the appropriate digital clock settings are incorrectly set somewhere in the signal chain.
The usual sync method is to use digital Word Clock, a signal which runs on 75 ohm BNC cables. It’s basically mic cable with a BNC connector on each end. Connect the devices WC connectors up to a common Master clock, and set the appropriate device as Master and the rest as Slaves. Now you’re good to go.
Possibly the only control that all reverbs have in common is the Wet/Dry Mix control, which might be called the Wet/Dry Balance or even Ratio on some units. Here’s a screen grab of the PSP EasyVerb, a simple but useful reverb plug-in from the excellent Polish company PSP, and they use yet another name, calling it simply the Mix control instead.
These controls can be used in two ways for two different applications. Full on, or just a little on, in short.
You can set Wet to 100%, and Dry signal will not appear at the reverb’s outputs. This is the SEND method.
This is ideal for treating multiple signals with a single reverb plug-in instance or hardware outboard reverb. This is the preferred method due to it’s flexibility and it’s dramatically better use of precious system resources. Using individual send level controls for each sound to be treated makes this a great approach to using reverb in a mix.
Alternatively, you can set a ratio more like 20/80 in favour of dry, or a ratio of 1:4 wet-to-dry balance. This would give you 20% reverb mixed in at the reverb’s output(s) with the original dry signal. This is the INSERT method.
This is ideal for processing only a single sound, or a single submixed group of sounds, mono or stereo, with a single reverb plug-in instance or hardware outboard reverb.
APPLYING REVERB WITH SENDS
For multiple signals to send/return (in individually adjustable amounts) to/from a single reverb
You get a lot more bang for your buck with this method. Usually the winning choice.
If the reverb Wet/Dry Mix control is set to 100% Wet, as is the usual scenario, then you use it as an auxiliary send and only the reverberation is presented to the output. This is the way you will find it easiest to use in a mix, and will maximize your use of system resources in the DAW.
The key to balancing the returned reverb in your mix is to set it so that it is not really noticeable when present, but once removed, it’s immediately apparent that something has changed.
Another important factor is that you are better off, in most cases, using reverb on a very limited set of instruments and vocals, possibly having just one thing sounding noticeably wet with reverb. The contrast with the dry sounds around it sets off the reverb like “sparkling gems against a black cloth”, to borrow an analogy from Pat Pattison of Berklee College of Music.
Contrasts are extremely important to generate on a continuous basis in most musical genres today. The public as a whole are also used to a minimalist approach to reverb these days, now that the thunderous caverns of reverb from the bloated 1980’s snare drums and guitars are far behind us. For the moment, anyway.
On a regular hardware console, you would use Auxiliary Send, either returning to dedicated Auxiliary Return inputs or to spare input channels (which are usually more full-featured than Auxiliary Returns). Individual sends let you balance how wet a particular sound will be in the 100% wet returns.
In a DAW, you would use the Sends from the individual channels you want to include, sending via an internal send buss to an Auxiliary Input type of track which has the reverb inserted as a plug-in, set to 100% wet, on one it’s inserts.
APPLYING REVERB WITH INSERTS
For a single sound to/from a single reverb
For this, you will be setting the reverb’s Wet/Dry Mix control to taste, typically, for me at least, at around 18/82 in favour of dry.
If the reverb is inserted on a channel, it has the disadvantage that an instance of the reverb is required for every channel you wish to put a reverb on. To use one and the same reverb instance or outboard unit to process multiple signals at once, you will need to connect via Sends instead.
On a regular hardware console, you would use Auxiliary Send, either returning to dedicated Auxiliary Return inputs or to spare input channels (which are usually more full-featured than Auxiliary Returns).
If you have a nice outboard reverb to use, then you can do that too with a DAW.
In your DAW’s I/O settings page, you can assign specific hardware inputs and outputs on your audio interface to route signal to and from a hardware reverb unit connected to the interface, setting software routing in the DAW channel’s inserts.
There will be a slight hardware latency involved in the case of external outboard processors, and it’s usually more than enough to cause phase problems, but with reverb it does not matter due to the notion of pre-delay, the time it takes for a sound to reach a boundary and reflect back to your position. This marks the onset of reverberation as opposed to having only direct sound within the soundfield.
REVERB DECAY TIME or SIZE
This is the second most important parameter in defining the character of a reverb next to the Wet/Dry Balance control.
This is the amount of time it takes the reverb tail to die away.
Although furnishings and materials will impact the results, in general a big room takes longer for it’s reverberant field to die away after a sonic impulse than a small room does. The decay time is an indicator of room size.
Big is not always better. The reverb is capable of masking other sounds in your mix, including the sound it is being added to. If the reverb is still dying away and quite loud in the mix at the point that the next note or beat appears from the sound in question, then it may be masked by the effected version.
Masking is a phenomenon whereby some sounds are obscured to us by virtue of competing sounds with similar characteristics. There are ways to minimize masking, or even entirely unmask sounds, lifting a veil from the audio. In my blogs on mixing later in this series during June, I’ll cover the topic of masking. Combating masking is critical if you want great results that sound clear, present and in focus.
The battle against masking actually begins with deciding which instruments will play what parts. This is arrangement, in the traditional sense of orchestrating a piece of music for an ensemble performance. Get that right, and you are winning half the battle. Have the guitarist play in a different octave to the keyboard player and any other guitarist, for example.
The Decay Time is measured by how long the sound pressure levels take to drop by 60dB after the original sound has stopped. This is called the RT60 value of a an acoustic space, whether it be a kitchen cupboard in a New York City loft, or the cloisters of Canterbury Cathedral. Every space has a measurable RT60 value, and studio designers and builders are wise to determine it for their spaces.
60dB of negative gain change is a huge reduction, since being quieter by 60dB is equivalent to falling to one-millionth of the original level.
Remember, decibels are a logarithmic measurement unit. The reverb is effectively gone at the RT60 value for a particular room.
It’s important to note that you may not have the option of setting a decay time or size. You can’t change a spring reverb’s decay time without using a different size spring, for instance. You also can’t change a plate’s decay time without using a different plate, in the real world, because the decay time is a function of the physical size of the plate (or spring).
The decay time for a convolution reverb is also a fixed amount due to the IR (impulse response) file properties, and so you won’t always find a means of adjusting decay time or size other than changing reverb presets, thus changing to a different IR file with a different (fixed) decay time. You can always try pitch-shifting or timestretching the IR file, but that’s about it, unless you are given an envelope and/or modulation control that can at least give you a dynamic effect, rather than a static wet/dry mix. Turning the amplitude down would simulate a shorter decay time whenever the amplitude of the reverb was lowered, but on the other hand it couldn’t be set to a longer audible decay time than the wet/dry balance setting allowed you to hear.
Sometimes it’s only a Size control, in which case you have a simple means of adjusting decay time that probably adjusts a few other subtle nuances too. Keeps things simple, I guess, but it’s nice to know what is being changed by a parameter in more detail than just “Size”.
I’ve mentioned this quite a bit already, but here’s a quick recap. Not all reverbs offer control over pre-delay, and some have no pre-delay anyway. It’s a really useful feature, though, so do read on.
Pre-delay time is the time it takes a sound to get to the first boundary, whether wall, floor, ceiling, console, or whatever surface it may be, and come back to the listener. This is why it gives a very real sense of how large a space is.
It’s the pre-delay control that sets this “first boundary and back” time, usually in milliseconds. Increasing pre-delay to settings above 30ms will increase intelligibility and clarity, unmasking the sounds the reverb is applied to in large part, since the original signal remains dry until the pre-delay time has elapsed.
You can really clear up a vocal nicely with a healthy pre-delay setting, but don’t overdo it, or the walls will be half a mile away before you know it! Even worse, the reverb will begin to sound disconnected from the dry signal.
The speed of sound in air is roughly one foot per millisecond (actually, 1080 feet per second, or so), so this makes it easy to do quick mental calculations of the rough size of a space from where the pre-delay time is set in a reverb patch.
Once the pre-delay time has passed, the first reflections begin to travel away from the nearest boundary and the build-up of the reverberant field begins within the space, whether real or simulated.
Using algorithmic reverbs, you can create spaces that would not occur in the real world. You could combine long pre-delay times with very short reverb decay times, for example.
If there’s no pre-delay control for your reverb, use it as an AUX SEND reverb, and simply insert a delay plug-in in an earlier insert on the Auxiliary Input track ahead of where the reverb plug-in is inserted. Set the delay plug-in’s delay time to be whatever pre-delay time you need on the reverb. Simple.
The distance to a boundary affects the tone of a reflected sound, since highs diminish in energy (relative to lower frequencies) with distance traveled by a sound wave through air. This makes things darker-sounding if they are further away. Our brains expect this, so it is helpful to provide a means of altering the tone of the reverb.
Ignoring the possibility of automating EQ settings, a static EQ will change the tone of a reverb. This is good for simulating the roll-off of highs in a signal, but ideally we want the sound to get darker over time, not in a permanent EQ setting, since the highs are dying away more quickly than the lows. This is a dynamic EQ change.
You can do it with automation, but hopefully you have a separate damping control for highs and for lows. The higher the damping value, the faster those frequencies will die away relative to others.
This means not only that you can simulate natural reverberation, but also that you can make the tail of your reverb get more and more wispy and shimmering and thin, or alternatively more and more ominous, dark and threatening. Sweet!
This control really helps with realism in more naturalistic styles of music that bravely consider adding reverb, but it doesn’t have to be used for realism. Sometimes it’s a consideration of retaining clarity in a sound through separating the dry and wet signals in time a little, so any frequency amplitude masking is reduced. Boosting the LF Damping control way up high will help you simulate a large, open space, for instance.
There will be a crossover point for LF or HF settings relating to the frequency setting for the damping or roll-off control. Set the crossover frequency for lows to be in a note that’s the key of the song, and you might get more space in the sound at that frequency area. This is because there is a dip in levels at the crossover point, wherever it is set. It’s far less important to worry about this in the high frequency area above 16kHz.
Adjusting these LF/HF Damping parameters (thoughtfully) can help you simulate the materials in a room or space. This is yet another good reason to learn a little about the acoustic properties of different materials like wood, glass, stone, etc. Intuitively, you already know what to expect in terms of brightness and reverberant qualities.
It’s also really helpful to roll off (discard) the lowest frequencies from reverb, since it sounds very thick and muddy on the lowest frequencies. The last thing you need is a huge wash of reverb floating about the bottom end of your uptempo dance mix. This is done, of course, with an HPF (high pass filter) set to keep all frequencies above the set frequency, often well above 120Hz on reverbs.
The more lows you turn down or discard, the less dark the reverb will be, and the thinner and more ethereal it can become.
Low mids are also boomy, so you may want to use a gentle EQ to turn down a broad area of low mids by a dB or two if the mix is sounding too tubby with the reverb left in.
HF/LF Damping controls go a long way towards realistically simulating the materials in a room and their effect on the reverb characteristics. In the absence of damping controls, you may find a Materials section instead, that you can experiment with.
The UAD-2 RealVerb-Pro is just such a reverb (see image below).
ROOM SHAPE and ROOM MATERIALS
These two are variations on the theme of HF/LF Damping controls in a lot of ways.
Room Shape is possibly the least useful of the two, since very few rooms will remain (if they ever were) simple geometric shapes, once populated with fixtures, furnishings, equipment and/or and people.
The Room Materials control, arguably, can prove rather more useful, being less subtle when changing from carpet, say, to glass or wood.
EARLY REFLECTIONS and EARLY/LATE BALANCE
As a reverb field builds up due to the increasing number of reflected waves moving through the space, like ripples in a pond around a thrown pebble, you will find that the earliest reflections are a series of closely spaced echoes, that cluster closer and closer until the field reaches a point where it is a complex web of intermingled echoes that can no longer be individually distinguished.
The time during which reflections are detectable as separate and discrete initial echoes is the early reflections period, and only then do we perceive the late reflections.
Due to many humans being unable to distinguish sounds arriving within anywhere up to 50ms apart as separate sounds (some of us do manage 30ms or so with critical listening training), the early reflections are heard in a different way than the late reflections.
Most reverbs give you a way to balance these two properties of reverb. This control is usually called Early/Late Balance.
Using all early reflections and no late reflections, you would get the sense of being indoors somewhere but without having any noticeable sense of reverb in the space.
If you only use the late reflections, you lost many of the spatial cues we use to localize sounds, which are primarily from early reflections. The result is a sense of being in a diffuse sound field, a large and open space without being sure where you are in it relative to any of it’s boundaries.
Early reflections can be great for livening up an electric guitar or piano without wetting it down too much.
Diffusion is a control that affects the space between the various echoes that together make up the reverb effect.
Acoustic reflections build up over time, gradually growing into a more and more dense reverberant field, as discussed above, but dying away through the decay time as they bounce more and more often from surfaces (at the expense of their inherent energy) until the RT60 time for the virtual space is reached and they are too quiet to be heard any more (60dB down and falling).
The Early Reflections are spaced further apart than the the dense mass of reflections that make up the Late Reflections.
It’s still possible to hear separate echoes in the Early Reflections of the reverb, so it’s easier to hear the effect of diffusion in the Early Reflections too. Sometimes the Diffusion control only affects the Early Reflections and yet, on other reverbs, there may be a separate control for diffusion in the Late Reflections as well. Most often, the Diffusion affects both Early and Late Reflections but it’s impact will always be easier to discern in the Early Reflection character of the reverb sound.
Lower diffusion values make for more regularly spaced echoes, whereas higher diffusion values will randomize the spacing of the echoes to some extent, making them sound smoother. The bulk of diffuse reverb in real-world reverberant environments will be heard in the later part of the reverb, during the tail, after the Early Reflections have concluded. The more diffuse, the smoother the reverb.
This effect can be pretty subtle, so it may take you a while to be sure what you are listening for. Play with it for a while to get used to listening for diffusion changes. Similarly, practice a bit with density control changes (see below) to get used to identifying them too.
Density is found on many reverbs, and is often found together with Diffusion, although they are extremely similar in operation.
Whereas diffusion controls spacing and build-up of echoes, density controls the amount and spacing of echoes, which is subtly different.
Raising density can appear to bring a sound closer, so lower densities tend to push sounds towards the background of the mix. This happens in nature.
If you use low density settings on a reverb, you’ll probably want to turn the diffusion control up a fair bit to prevent the reverb sounding too much like a resonant delay feeding back on itself with a really short delay time.
Higher density reverbs tend to be darker, richer and more complex than a low-density reverb.
Density can be thought of as an audio equivalent of playing with depth of field in an image. You can blur things a little with added reverb density, and it tends to move things towards the background of the mix, allowing a contrast with other elements of your mix that you choose to have clearly focused instead, right there in the front of the mix.
NON-LINEAR REVERB AND REVERSE REVERB
There are non-linear reverbs, non-linear being an artificially generated type of algorithmic reverb which does not appear in nature. These make rather alien, inhuman sounds that tend towards resonant, metallic sounds.
They are cool on all drums and guitars, and very distinctive. The work of Peter Gabriel has often featured plenty of non-linear reverb, for instance, often using the AMS Reverb so popular in the 1980’s and on into the early 1990’s. Many large studios (those that survive) still have the AMS reverb and delay units in their outboard racks, since they are very popular, despite the limited bit-depth of their digital sampling rate.
These early digital units could also be triggered via MIDI to play back a sound that had been played into them, which was another neat trick often used.
Reversing reverb in your DAW is yet another very cool effect, requiring nothing more difficult than for you to record the reverb returns to a new audio track. You can then reverse the resulting audio file and position it where you want it on the timeline, usually ending on a downbeat so as to provide a “sucking upwards” effect that ends right at the start of a new bar or fill. It sounds great on tom-toms and long sustained piano chords.
Most of the time, this will only change the width at the stereo outputs. If it’s a really clever reverb, it may let you decide how wide the source itself is within a given virtual space, but that’s a lot less common.
Virtually every reverb will treat the source as being in the centre of a space ,with an equal distance assumed from the source to the virtual left/right boundaries.
This makes reverbs typically display a close symmetry in their L/R imaging, which can be defeated by panning so that each side of a stereo reverb is at a different distance from the centre position of the overall stereo image of your mix.
The wider apart the sides are panned, the closer to full stereo width the reverb’s output will be, but the different offset for each side is the important factor in defeating unwanted symmetry in the reverb’s stereo outputs.
Some budget hardware reverbs, such as guitar pedal reverbs, might not even be true stereo at all. They can easily be cheaper mono units posing as two-channel units. The left side may simply be a copy of the right side with the phase reversed.
If this is the case, summing the reverb returns to mono and listening will demonstrate the sound disappears in mono, due to phase cancellation between the two identical but opposite-phase sides.
The workaround is to use two of the reverbs, using only the left output from each unit.
SURROUND SOUND REVERBS
I doubt that many of you reading this will be using a 5.1., 6.1 or 7.1 surround sound multichannel reverb at home. Still, they do exist and I use an algorithmic eight-channel hardware reverb myself.
To be wholly accurate, I use it in a mode that gives me up to four different stereo reverbs, rather than as a single surround sound 7.1 system. It is very common to require at least three types of reverb in a mix.
The Lexicon 960LD and the TC Electronics’ TC6000 are both very capable, full-featured professional multichannel surround reverbs, but they both cost well into five figures, which is why you are unlikely to be using them in the average home studio.
10 QUICK TIPS ON USING REVERBS
I’ll be talking about reverbs and delays again, in the mixing part of the blog series later in June, but for now consider the following ten tips on suing reverbs effectively.
1 Use different types of reverb in your mixes for sonic variety. If you use one room reverb, don’t use a second similar room. Maybe use a Hall or a Chamber instead.
2 Try reverbs in different applications, such as blending things, adding size or sustain to things, just to see which reverbs are best for which purposes. I recommend using plate reverbs for snare and vocals, but ideally a different plate for one than the other. If you do use two algorithmic reverbs in the same mix, use a different manufacturer for the second one to avoid using the same underlying algorithm to generate the reverb.
3 Slower songs can handle a lot more low end in the reverb, and, obviously, a longer decay time. Faster tempo songs prefer shorter decay times and brighter reverbs.
4 Use very short delays instead if the reverb is clouding your mix and you find that playing with parameter settings does not resolve this.
5 Set the pre-delay to be a multiple of the BPM of the song. If it’s at least 50ms, definitely there will be an audible delay before reverb occurs. If the pre-delay is set to more than a few hundred milliseconds, it might start to sound a lot more like a delay effect than it does a reverb.
6 Always set your various reverbs to have different decay times from each other, unless, perhaps, you are blending two reverbs to act as a composite reverb on a particular source or group of sources.
7 EQ the reverb returns, or EQ the reverb sends, see what you get. If you cut a frequency in the send, then boost it in the returns again by the same amount, or vice versa, you can get some cool results despite the fact it shouldn’t really make much difference. Experiment.
8 Compress the sends to the reverbs for a smoother sound. Less transients are there to excite the reverb, so things are generally mellow. This really suits ballads.
9 If you must use a noise gate on a reverb, and want it’s action to be as transparent as possible, it’s best to gate sounds before they hit the reverb (in the sends) so as to avoid the unnatural cut-off in the reverb that gating the reverb returns would cause.
10 Make your own reverb by playing audio in real rooms and capturing it with microphones at suitably reverberant locations in the room (meaning at a sufficient distance from the source). Use playback speakers and microphones if you have already recorded the sound you want to add reverb to, and mic up a playback.
That’s about it for reverb, until we get to the mixing section of this blog series towards the end of June 2013.
Join me tomorrow for a quick tour of the other assorted oddities of the processing world, everything from bitcrushers to exciters, from distortion and amp simulation to transient designers. I’ll also look at time-stretch functions, and the ubiquitous pitch correction we are treated to these days at every turn. I’ll look at Celemony’s Melodyne as well as Antares Auto-Tune +Time.
We’ll be looking quickly at each of these topics, because it’s simply too vast an area to cover in a single blogpost, but I will be doing my best to cover all the areas I can think of.
It’ll be quite a long post, no doubt, due to the sheer range of interesting oddities that are worth a mention.
I hope you’ll join me for the guided tour! See you there.