The most common dynamics processors are compressors.
You choose settings appropriate to the incoming signal you want to compress, and then it will behave according to those parameters while you focus on other issues in your mix.
You can use compression as a spare hand to control the fader levels that won’t work if they stay in one place during the mix. This was the original intention of the inventors of the various early compression circuits. They are awesome for creative purposes as well, thankfully, so they see a lot action in the studio.
The way a compressor behaves is determined by a combination of not only the settings made on the compressor controls, but also by the characteristics of the input signal being compressed. You also have to consider the interactions between these controls, since they are inter-related and moving a control often affects the way another should be set.
You can go in circles somewhat without a good understanding of what is happening. Tomorrow, I’ll give you the rundown on each of the usual controls for dynamics processors, including compressors, and what they each do.
I’ll be sure to tell you how they can be set up on any compressor to get in the ballpark quickly.
Setting a compressor properly takes practice, and some are very much easier to use than others. But don’t despair – they are not so bad once you get the basics clearly in mind.
My personal recommendations are for an 1176 style compressor as a good novice-friendly plug-in to learn on. The controls are minimal, and the effect is easy to get and to hear.
Try setting the big knobs on the front panel to ten o’clock and two o’clock respectively. This has been called the “Dr. Pepper” setting by some wags, and often works pretty well for vocals, given a sensible ratio setting for the singer, maybe 4:1 even if they do have good mic technique, and on a bass guitar or snare or kick drum they can sound fantastic anywhere up to 8:1 ratio.
The classic Jeff Lynne snare sound from all those great ELO and Traveling Wilbury albums, and his solo records? That is a quality snare played well (usually on it’s own, no other drums played in the take, so perfectly isolated) going through an 1176 heavily squashing the snare. Crushing it to death, as he puts it. Using any of the numerous 1176 plug-in emulations out there with your best snare and an SM57, you can get a decent result from this same technique.
Personally, I don’t use any dynamics processing during recording. I only add it later, and I record at conservative levels so that clipping is not going to happen on my vocals.
Compression, then, evens out the level of unruly faders in a mix. It’s useful but optional during recording passes, and downright wonderful on live signals at a gig, as you bellow “Hello, Vancouver” into your vocal microphone then start the set, perhaps unwisely, with a tender love ballad.
With vocals and bass guitar, compression is usually called for in part because we hear these two things with compression all around us constantly. Everywhere we hear music, the bass and the vocal will be compressed in a way that you can tell they are compressed – it’s an audible change in the character of the sound, and one that is deliberate. You can paint tonal colours with compression, and we do it all the time.
WHEN TO AVOID DYNAMICS PROCESSING
There are two times not to use dynamics processing.
First, when you don’t need to because the fader can be left in one position throughout a mix and it sounds good all the way.
Second, when it sounds wrong to use it, because the style is firmly naturalistic, and it’s audience is opposed to an engineer intruding on the musician’s intentions. There are “photographic” styles of music, that document as a snapshot if you will. There are less realistic styles that are hybrids of technology and realism.
The Dixie Chicks are a perfect example of artists using natural musicianship and instrumentation alongside the best the modern studio methods can offer. It’s a gorgeous marriage.
If you are doing certain types of jazz and/or classical music, you will be a lot less likely to use compression, if at all. If you do use dynamics processing at all, it will be the transparent type you use 99% of the time. These styles of music have internal dynamics generated by the players as they play together simultaneously – at the same time in the same room making eye contact as they do so.
They read each other’s body language, facial expressions, and they see what the other person’s hands are doing.
They react in the moment to what is being played and felt by the other players. This is great! For many, this is ideal because it is quintessentially human.
Putting aside music styles which do not call for compression, let’s turn to those that do.
The elements of bass and lead vocal in a song, each one critical to the overall result, have a lot of variation in their dynamic range during the performance. If not, the song is rather tedious to listen to.
Bass and lead vocals almost invariably get compressed, sometimes several times over, in a series of dynamics processor operations of different types.
Compression and limiting, for instance, are very common to find together on vocals. The former to even things out, the latter to catch instantaneous peaks.
Compression can also add sustain, body, fattening up sounds. This makes it great for beefy drums and for walls of powerful rock guitar. The trick is not to over-do the compression effect, since it gets ever smaller and more lifeless the more you add. Just firm it up, no more.
Ratio and threshold interactions are big influences here. Lower the ratio, raise the threshold, either of these moves should lessen the effect.
COMPRESSING BASS GUITAR AND VOCALS
Let’s look at bass, since bass is almost always heard with compression in a mix.
Bass players who can play consistently and evenly are in great demand, and a great deal of the sound comes from the player through nuances of their touch on their instrument and their internal time feel for the music.
Hand the bass to another player to have a go, and you’ll hear a different sound again, even though it’s the same instrument and the same bass part. A similar type of sound, but definitely now different.
Only so many bass players have the experience and technique necessary to pull off a consistent performance in level terms on a recording without the aid of compression.
When recording, every note is “under the microscope” and you know you’re playing for keeps. Being able to do another take makes things less stressful, but you still want an even, consistent bass part.
We mere mortal bass players must employ some compression to even out the part.
This might be better done after you record, rather than before, but to avoid unwanted distortion you must set your gain structure carefully in either case. That means check levels all the way down the chain into the recorder, wherever they can be altered or set.
Make the level as good as you can as early as possible in any analog signal chains. Too low adds noise, too high overloads the input. Clean, loud signals mean less level is needed at the pre-amp stage, which keeps the noise down.
Whatever you add to the chain adds noise and distortion in the analog world, which makes things sound more blurry and out-of-focus. Stay in focus and punchy, minimize processing the signal and get good levels that don’t clip.
With a bass guitar, if you are not recording with any compression, set the recording level to the DAW so it is hovering around the -10dBFS mark most of the time. This should prevent unwanted clips on the loudest bits, and not be too quiet either. I don’t think it would do you any harm to allow a few dB of gain reduction on a bass guitar, just to even out the levels. you could follow it with a limiter to catch sudden peaks caused by momentary transients. This works very well on vocals too, although gentler ratios at recording time are advised.
Stick to 2 to 1, or maybe 3 to 1 and be sure to use moderate attack and release times. That should work very well on many sounds.
Watch out for pumping noises, that means you’ll want to lower the ratio or raise the threshold or both if you don’t like the pumping, breathing sound.
Think of Ringo’s sound on the Beatles’ “I’m Only Sleeping”, because that’s the weird sucking thing the drums do with heavy compression (high ratio) and the incoming signal moving above and below the threshold setting regularly with the release time set to make a musical and rhythmic feel to the onset and release of the compression effect in tempo with the track. Creative compression!
You can compress after the fact, which allows you not only to get the optimum settings, but also to adjust them moment by moment in the song using automation of the parameters themselves, which is very, very simple in a DAW.
If you check the manual or online help for your DAW software, you will find it’s easy to enter and to edit automation data onscreen. Once you try it, your mixing will be revolutionized.
I’ll blog more about automating your dynamics processing during my mixing episodes of this blog series, to be posted around the middle of June. Imagine how handy it is to be able to automate a plug-in to step on all the “esses” in your song about the shells by the seashore and be smart enough not to affect anything else in that vocal.
There is no noise penalty in recording well below the clip mark on the fader, provided you are making a digital recording (into any DAW, for instance). The more bits you use up the better, true, but there is no noise penalty, and you can change the gain later with no penalty if you work in a decent DAW that processes at 32-bit or 48-bit.
Cubase, Logic, ProTools, they all do this at the necessary enhanced “up-sampling” resolution. For example, converting the 24-bit file you want to process to 48-bit while they do the processing, then converting back to 24-bit for the result delivered back to you by your plug-in.
There is no harm in playing it safe with recording levels. Just keep at least 6dB below the -0dBFS mark on your fader for the loudest stuff in the takes of any instrument that is not a static pre-set level each time it sounds – such as a recurring MIDI drum sound like a triangle programmed at the same velocity throughout the song. Don’t get any lower than you can help, though, since every bit helps.
Obviously, don’t use noisy low levels with analog equipment in the chain before the digital input, unless you want noise levels to rise in the mix.
Confirm the gain structure is set sensibly all the way through, checking it’s good at each step where you can adjust a level, until you get to the DAW from your source signal. Your sounds will be strong and in focus.
With a little care, there should be no noise present in a recording to a DAW except noise deliberately added, or the inevitable noise added by things like very-low-cost analog equipment, analog guitar pedals or samples taken from older analog recordings. Even that penalty can be minimized by setting the appropriate levels at each stage of the signal path from source to storage medium.
Ideally, your mic or instrument enters the console or preamp or interface input, and it doesn’t require a pad or a trim control to be applied. The preamp does it’s thing and raises the level to line level for the DAW to convert it to digital, process it if desired, and store it (record it).
It’s also smarter to avoid EQ until you have recorded the sound. Do it later in context with other sounds in your mix and with reference to other similar recordings in your genre that are commercially available.
Use filters in the mix on every track to keep only those frequencies you actually need in your mix, and then you really won’t have to worry about having recorded those low rumbly bits of traffic noise. Filtering on the way in is great, but you can’t recover what you filter out, so be careful until you know what you like and need.
You could automate an even sound on every instrument and vocal by going through the track bar by bar and entering the fader moves one after the other that keep the signal at the level you want in the mix at all times.
This is all very well, but time is precious. You’ll be working on that mix for a very, very, very long time.
Do that for every track in the mix? Well, that’s a great education, because ti shows you how to deal with each type of problem. All the same…
…in the interests of making acceptable progress and finishing work while still inspired, it’s very important to trim back the time it all takes to a less “white coat” level of technicality.
You don’t want to work the life out of a recording. It can easily happen. Better to use dynamics processors to rein in the levels of your individual track elements as much as possible, and see what really HAS to be dealt with by personal intervention.
It’s only sensible to use a plug-in to make those changes based on where you set the controls and how loud the input sound is, moment by moment, rather than entering data for every track in every bar. Time is saved by letting the compressors get things under control, and seeing what isn’t being helped by the compression.
Which faders appear to have an inconsistent, unstable level no matter where you leave the fader? Compress those first, and then you will find it’s a lot less work to control a mix. The more unstable, the greater the compression ratio you may require, or the lower the threshold. Try to keep attack and release sounding musical.
Slower attacks may sound smoother, but will allow sudden transients to pass through unchanged. Longer release times may not allow compression to reset before the next note arrives, which might be perfect or unsuitable, depending on application.
In effect, compressors will act like an invisible hand that turns a track’s fader up and down for you while you are doing other things, but in a way that affects how much intensity or pressure the sound conveys.
Compressors can also sound really, really cool, especially the tube and photo-optical types, and you have a whole bunch of styles to choose from. Think of the drums on “Rain” by the Beatles and marvel at the genius of all concerned.
There are several methods of getting compression, and they all have things that make them useful for different situations or for getting different tone colours. Some compressors are amazingly obvious when you hear them, others are so transparent as to be “invisible”.
Many people like to use compression in series, chaining together several, each to perform a different function. This is a skillful use that takes effort to master, since adjusting one compressor can be confusing enough at first. Two in a row can be total chaos in no time. Save chaining compressors for when you are comfortable using compression. Then, it can be very helpful.
The aesthetic use of dynamics processing has ventured into new artistic territory many times since the birth of rock and roll in the 1950’s, and people are still breaking new ground today with inventive users in most genres. The myriad sub-genres of urban and pop music styles are especially rife with cool processing.
The most useful place for compression in the home recording studio is on the vocal. The human voice has a huge dynamic range and sounds can go from a whisper to an explosive shout almost instantly.
The loudest peaks in the singer’s performance need to be captured without overloading the pre-amp input (which might well be your DAW interface input). You don’t want to badly distort these louder parts of the vocal take. A compressor is very useful for this.
The ideal compressor on vocals is generally agreed to be analog, with tubes (valves), because the harmonic distortion introduced in these units is considered very flattering to vocals.
Often a solid-state mic will be matched with a tube pre, or vice versa. Both the mic and the preamp being the same type is less preferred, but works perfectly well. Two tube components in a signal chain can get noisy.
It is perfectly possible to get good results from plug-ins, however, especially those modelling analog compressors. Waves and UAD and many, many others make excellent emulations in software of classic tube compressors.
Compression really helps put sounds right in your face, up front and larger than life.
Compression fattens sound up – but only up to a point.
Over-use of compression sucks the life out of a sound. Do that on enough tracks in your song and the music will sound lifeless and drab and strangled.
It is a nice sound, but it takes a while to become familiar with the sound of compression.
The more you pile it on, the weedier the sound gets. Lifeless is a common description for how this outcome of over-compression leaves the audio sounding.
Once you are attuned to the unwanted artifacts that mis-set compression causes in the audio, you will be able to keep your ears open for damage occurring as you pile on the compression.
Go easy on gain reduction and don’t keep the threshold control any lower than you need to, since it is incredibly difficult to repair over-compressed audio, even in part, and usually it is completely impossible. It’s hopeless.
Better to set it how you like it, and then back off, just in case, at least until you are comfortable with using compression. It’s easy to miss a problem at first with the subtle nature of compression impacting at different frequencies.
You can always add more salt, but you can’t take it out of the soup.
Tomorrow, in “PROCESSING – DYNAMICS Part 3″, I’ll be covering the actual controls of a compressor, and how to set them up reliably and quickly.