A Little Crush

Shaping A Drum Mix With Compression

Few things are more satisfying than well-recorded drums dialed up fat and punchy in a good mix. Naturally, that all starts with a nicely tuned kit, great playing, and good miking practices, but other factors are involved in achieving a good mix for your recorded drum tracks. The compressor is a powerful tool, which can even out levels (such as an inconsistent kick drum performance) or help certain tracks have more presence. However, if it’s overdone it can totally kill the dynamic life of your mix, removing the sense of depth and dimension and making it sound flat.

But there is more to compression than just reducing dynamic range. If you become familiar with the tool and use some imagination, you’ll discover creative approaches that can actually fatten a drum sound or make cymbal hits blossom. And there’s also a time and a place to get heavy-handed (in a tasteful way) to really jack up a drum sound and give a recorded track something different.

Compressor Basics

Before we begin goosing sounds, let’s take a quick look at a compressor’s basic functions and standard controls (be it a piece of gear or a software plug-in). A compressor is essentially an automatic volume controller that uses the input signal to control the level of the output. It can make loud sounds softer and bring up the quieter stuff, which is why it’s referred to as a dynamics processor.

A compressor usually gets “inserted” on a mixer channel via the insert send/return jack (or on a DAW by selecting the insert option on the desired channel or subgroup). Typical adjustable parameters include threshold, ratio, attack, release, and make-up gain. Some models have fewer user-adjustable options, such as the dbx163, an older unit that has a lone slider on the front that simply goes from “less” to “more.”

The threshold determines at what level the compressor will kick in while the ratio controls the input-to-output level for signals that go over the threshold. If you were to select 6:1 as your ratio, an input signal that goes 6dB above the threshold would result in an output that is 1dB above the threshold. These two controls work in close association with each other, as adjusting one will affect what is happening with the other.

Attack determines how fast the compressor will react to input levels that go over the threshold. Slower attack times will let a transient through to sound more natural, while a faster attack may squash that initial peak. Release sets the rate at which the compressor will stop compressing once the signal drops below the threshold. This parameter determines the effect the compressor has on the decay of the sound. Make-up gain is used to bring the level back up after the compressor has applied gain reduction to the initial signal. This has the effect of making the quiet stuff louder.

Compression Angles

There are several approaches to applying compression to drums. One way is to compress individual tracks on a multi-miked kit. For instance, a snare drum track that was played inconsistently could become more solid with compression. Another way would be to sum the tracks to stereo and compress the entire drum set, which would bring the kit forward in the overall mix. Or you could go for a hybrid approach and compress individual tracks as well as compressing the entire mix. While this is over-compressing to a degree, it is pretty much the basis of the modern rock sound.

And then there’s the angle of using compression as a creative tool to shape and affect the overall sound of the recorded drums – individual tracks, the mix, or both – which is where it becomes exciting. To get a broad range of techniques, we spoke with several recording engineers who agreed to share their drum compression tricks: Steve Orlando, John Scanlon, and George Borden (who are also drummers and percussionists), Timo Preece and Drew Webster (who both delve into electronic music in different ways), and Scott Theakston (a bassist who is a veritable encyclopedia of recording and mixing know-how).

The Conservative Route

Some folks prefer to use only a slight amount of compression to give the drums a bit more presence in the mix while preserving the better part of the original dynamic range of the tracks. Borden subscribes to this camp, opting to compress only the kick drum track. Although, in order to bring the sound of the whole kit together, he’ll then assign the entire drum mix to a subgroup and send that through a compressor with conservative settings. A ratio between 2:1 and 4:1 with a carefully set threshold will push down the peaks a bit, reducing the crest factor so that he can turn the overall level back up with make-up gain. This helps the drums sound fuller and more forward in the mix without being obviously compressed.

Compression will make the transient lower in value in relation to the body of the sound, so it can be used to highlight body and tone. The other side of getting a thicker sound is that the attack may be lacking. To keep the defining transient, you could lengthen the attack time to let the leading edge of the hit through before the compressor kicks in.

Parallel Compression

Here’s another way to enhance the body and punch of a recorded drum track path while maintaining the dynamics. Rather than using the compressor to make loud sounds quieter for more headroom, so that you can bring level back up, this approach allows you essentially to make the quiet sounds louder with some creative routing, processing, and mixing.

This technique involves taking a copy of the track, sending it through a compressor, and returning the processed version back into an open channel to mix in with the unprocessed track. In order to send a version of the track(s) to a compressor while preserving the original, you could use a splitter, make a track copy in a DAW, or use a subgroup or aux send.

{pagebreak}

Theakston beefs up the sound of the entire drum mix by sending it via a subgroup out to a compressor, and returning the processed mix on a parallel path, such as another channel or two on the mixing board. ( Fig. 1) Blending the processed mix with the unprocessed drums adds body and presence without trouncing the original dynamics.

(Above) Fig. 1. A stereo mix of the drum set is sent to the compressor inputs via Bus 7-8. The stereo mix is compressed and the output sent back into a stereo channel and blended with the unaffected mix. This beefs up the sound without affecting the transients of the original mix.

Sometimes he opts to process only a couple of elements that he wishes to emphasize, such as the snare and overheads, and perhaps the room mikes. Adjusting the settings on the compressor to accentuate resonance and sustain on these tracks, he’ll blend them back in with the original sound. This can result in epic-sounding cymbals and an explosive room sound without influencing the kick drum sound. For that big Bonham sound, try a simple miking setup of a stereo pair of overheads and a kick drum mike, and take a copy of the overhead tracks to a compressor, squash them, and blend them back in with the unprocessed tracks.

Orlando, a punk drummer, needs his kit to cut through a mass of screaming guys and distorted guitars, so he gets a bit more heavy-handed. Compressing a subgroup of the drum mix using a fairly radical ratio of 5:1 and above, or even 10:1 or higher for a super-squashed sound, he adjusts the attack and release time to get the sound he wants. Returning this uber-compressed mix to some open channels and adding it in with the original tracks allows the drums to punch through a mix while preserving the original character and dimension of the drums.

Hip-hop producer Scanlon often uses parallel compression on the kick drum to get more punch. The low end that the compressor eats up is still present in the unprocessed track, so mixing the two together maintains the lows while accentuating the pop. EQ can be used to add more snap to the kick mix.

Scanlon will also process the snare drum to make it louder and sit forward in the mix. He goes for 10dB—12dB of gain reduction (peak squashing) by the compressor and uses the make-up gain to boost the level back up for an ultra-present snare. To add extra fatness to a snare drum, he sometimes splits the track and takes one of the two through a slight delay, and then pans the two tracks at 10:00 o’clock and 2:00 o’clock.

Another mixing trick involves layering samples underneath existing drum sounds and mixing them together for sonic variation during changes in a tune, such as a shift to a chorus. Adding, say, a brighter sample underneath the original snare or kick drum signal will add dimension to the song. Scanlon will often take samples and squash them with compression before layering them with the original track. You could either match up the samples manually, which could be tedious, or with a program like the Massey DTM (Drums To MIDI) plug-in, which analyzes drum tracks and puts out MIDI notes with velocity information.

Sound Shaping

Compression can also be used to shape sound by softening the transient and extending sustain. Overall perceived loudness decreases, and make-up gain brings it back, accentuating the midrange on quieter passages. Using a compressor to shape and fit your sound in a mix has the extra bonus of not messing with the signal phase like equalization can.

Scanlon shared a tip on how to make cymbal washes blossom using a compressor’s sound-shaping power. He cites a session where he used three room mikes 3' apart in front of the kit, and a pair of cymbal overhead mikes about a foot above the cymbals in a split configuration. During mix-down, he applied compression using a short attack time and a fast release. The short attack time squashed the peak of the initial cymbal hit, and the fast release let up on the compression quickly, bringing the resonance back up and letting the cymbal wash ring out, swelling upwards in kind of a reverse sound envelope.

Using a compressor to affect tone will invariably affect the dynamic range to a certain degree. Theakston sometimes uses Transient Designer on drums to get around this phenomenon. This is a device/plug-in that allows you to shape the attack-and-release envelope on a sound source without changing the dynamic range. SPL offers both hardware and software versions, and comparable plug-ins, include Waves’ TransX and Oxford’s Transient Modulator.

(Left) Fig. 2. This transient shaping plug-in and others like it can accentuate or soften transient attacks and affect the length of sustain, effectively bringing a sound forward in a mix or pushing it further back. It can enhance the sound without typical compression artifacts.

Theakston, who often mikes both the top and bottom of a tom and sums them together to get a deeper tone, will use Transient Designer to control the drum’s ring by backing off the sustain parameter. “It’s almost like a gate without the gate artifacts,” he says. Using Transient Designer, he takes out sustain and adds attack to make the mikes sound closer for a more intimate effect, or adds sustain and reduces the attack to make the mikes sound farther back and roomier. (Fig. 2)

{pagebreak}

Finding A Pulse

To add a pulse to the room sound, you could insert a compressor on the room mikes or overheads and adjust the compressor for fast attack and release times. The fast release opens back up the compressor and rapidly brings the level back to normal, so it sounds like it’s pumping. This is a classic rock sound – even Led Zeppelin employed this technique.

Another crafty trick is to use the side-chain input, which is sometimes called a key input. This is simply an input for a signal that will serve to control the compressor on another track. Setting the threshold level controls the amount of compression that will be triggered from the key (or side-chain) input.

Preece, who does a lot of electronic and dance music, employs the side-chain input to correct issues, such as frequencies that overlap between tracks, and also as a creative tool to generate a pumping effect in his mixes. In order for the kick and the bass guitar to move out of the way of each other, while both remain present, the side-chain can essentially duck one sound out of the way when the other one dominates, and bring it back up when it’s out of danger of tromping on and murking up the other sound.

Use the kick drum track to control the compressor on the bass to duck it out when the kick is hit. To get the signal on the track to the side chain input on the compressor, you could send it out on an aux send or subgroup. Or if you’re using a half-normalled patchbay, you could split it out from the insert send. Pro Tools gives you the option to use two outputs from the same track in order to send a copy of the signal to the compressor.

Preece generally starts with a moderate approach, setting the ratio to around 6:1. in order to preserve the transient, he adjusts the attack so it’s not too quick, and sets a quick release time to bring the sound back in fast. He then adds make-up gain to bring the sound back up. This also creates an overall pumping pulse to the mix. Setting a short, repetitive percussive sound to trigger the compressor on a track with a steadier, longer duration sound creates an interesting breathing effect in the mix.

Getting Radical

Webster, a sound-tweaking experimental madman, enjoys doing a Dr. Jekyll and Mr. Hyde job on drum tracks, ending up with something almost unrecognizable from the original. He often uses between two and ten plug-ins per track in Logic to achieve his nefarious aims. If you want to really go for radical impact with your drum mix, here’s a glimpse at his unsubtle methods.

He uses a compressor to initially even out a track and prepare it to be processed with reverb, distortion, delay, or whatever time-based processing he can think of to morph the sound. After processing the compressed track with a variety of effects, he’ll finish it off with a 4-band compression plug-in to highlight certain sonic aspects, such as the high end from distortion. (Fig. 3) Multi-band compression allows for frequency-specific dynamic control and provides separate compressor controls for several selectable frequency bands. This lets you affect only one section of the frequency content from the original signal or mix.

(Above) Fig. 3. This multi-band compressor plug-in separates the frequency spectrum into four bands. Here the crossover points are set at 100Hz, 1kHz, and 10kHz. This allows for variable compression of different frequency ranges of the signal.

To either emphasize or de-emphasize a specific frequency on a track without messing up the fundamental, Webster recommends employing the side chain for surgical control. Duplicate the audio track so that you have two identical tracks. On the first track, select the frequency to notch or boost with EQ. Send that equalized track into the side-chain input to drive the compressor that you have inserted on the second track.

Notching a frequency with EQ on the side-chain track will emphasize it on the original track. Boosting that frequency will sublimate it on the original track, because the compressor will clamp down on the boosted frequency that is controlling it. This technique keeps the dynamics alive on the track that’s being compressed by almost surgically altering a specific frequency with compression. Overall this has a less-destructive effect than compressing the whole track. This approach could also be used to emphasize natural reverb tails on the original track without adding artificial reverb, which could affect the fundamental frequency.

Serve The Master

When to compress and how to compress? Those are the questions. Whether ’tis nobler for the tune to squash the bejeezus out of your drum mix so it could jump out of the speakers in an F16 cockpit, or to tweak your recorded drums with a few nifty sound-shaping compressor tricks, and by so doing, achieve some cool effects while preserving the dynamic range. It all depends on what you’re going for with your drum sound and what will serve the music best.

Keep in mind that when your overall mix gets mastered, it will undergo an additional round of compression/limiting, so be sure to leave some of that dynamic range intact, even if you are going for audible radio play in an F16.