Creative Compression

Publish date:
Social count:
Image placeholder title

Over the years, much has been written and discussed about compression. Everyone has a passionate opinion about how and when it should be applied, and what types of compressors should be used. Some people have gone as far as to say that the overuse of compression has destroyed the process of recording beyond repair. Personally, I love compression. To me, it's one of the great wonders of the production world.

My aim here is not to restate the already well-covered conventional compression techniques. Instead, I will offer up some less-common and more aggressive applications for compression that use it to heavily manipulate audio.

Image placeholder title
Image placeholder title
Image placeholder title
Image placeholder title

FIG. 1: This example, using a stereo drum loop, shows the steps for reverse compression: Take the original loop (a) and reverse it (b). Then compress the reversed loop (c). Finally, reverse it one more time (d) so that it''s facing forward again.

Dynamics in Reverse

This first technique has been around for quite a while, and it is a vastly underused form of colorful compression. I refer to it simply as reverse compression. That name means exactly what it implies. Though this method has been in use for many years, beginning in the old days of analog tape recording, reverse compression is much easier to achieve in your DAW.

To start with, take a drum track or some other source with dramatic dynamic changes. In your DAW, reverse the audio file. Next, insert a compressor of your liking on that channel and play around with the attack and release times a bit. This is the tricky part, but also where all the magic is. Of course, you can just throw caution to the wind and try an extreme setting, such as a fast attack, fast release and about a 10:1 ratio.

Either way, when you're feeling good about it, you can either process the track with the plug-in or bus it in real time, using a send, to another track. When you have printed the effect, select your processed audio and reverse it yet again so you're back to the normal audio (see Fig. 1). Now you'll hear the compression you printed while it was reversed. With luck, you will have completely mangled the audio into something way beyond what normal compression is capable of. If you're not happy with the results, change the compression parameters and try again.

I tend to use the most extreme forms of this technique on drums, where the transients cause a tremendous amount of artifacts. On drums, you'll hear the typical breathing and pumping that you can get with extreme compression, but depending on how you set the compressor, you can also achieve exaggerated envelopes that seem to suck into each drum hit rather than after the hits, as you would with normal compression. On a recent session with Finnish band Lapko, I used this technique most notably on one particular drum track that needed to be a bit special (see Web Clip 1).

While it's used mostly for extreme manipulation, reverse compression can serve as a more subtle effect on sources such as vocals and guitars. In certain instances, you can get a vocal to compress heavily without the major artifacts you would normally hear. Remember, because your audio is reversed when you apply the compression, the compression artifacts will also happen in reverse when the audio is played back in its original direction.

The key to getting good results is to experiment with the compressor's parameter settings. It's hard to generalize about what the best settings are because it depends a lot on the source. However, I have found that short release times and slow attack times together will give you an almost gate-like effect, whereas longer release times will sound smoother and the effect more subtle.

Image placeholder title

Here''s the signal routing for a sidechain-compression application that makes it possible for reverb on an aux send on a guitar track to be ducked when the guitar plays loud notes.

Illustration: Chuck Dahmer

Chain, Chain, Chain

If you listen to pop music, especially of the dance variety, you've no doubt heard the effects of sidechaining. It amazes me that this technique, which is an extremely powerful method of creating wild textures with otherwise lackluster tracks, hasn't become more popular as a mangling tool in rock music. I've been on a mission for some time to bring extreme sidechaining into rock music, manipulating more than just the usual suspects such as bass and kick drum.

In the land of the DAW, sidechaining is easy to accomplish, and there are many shaping options available. One of the classic struggles in the studio is how to take a dull track and give it some explosiveness and verve. Many times, when I get asked to mix a project, the tracks I get are exactly that: dull and lifeless. Of course, drums typically suffer the most, but guitars can also fall prey. One way to liven up a guitar track and create some space is either to reamp it into a real space or to create an aux channel and insert a good room-type reverb on it. The reverb plug-in should be set to output wet signal only, with none of the original dry signal. Of course, I'm sending the guitar to the reverb send.

After the reverb, I like to insert a fairly aggressive compressor or limiter that has a sidechain input. I use Digidesign Pro Tools primarily, and one of my favorite plug-ins is the Waves SSL G-Master Bus Compressor because it both sounds fantastic and has a sidechain input. Now, set up another send from the guitar so that it feeds the sidechain input of the compressor. Once that's all routed (see Fig. 2), I like to set the compressor so that it aggressively ducks the reverb while the guitar is at its loudest and quickly releases as the guitar quiets down. This is obviously most effective with guitars that are dynamically active or staccato because the reverb plug-in excites whenever the guitar stops, creating a more lively and active space around the guitar. And unlike simply putting reverb on the guitar, the result is rhythmic pumping from the compressor as it ducks and lifts. As an additional benefit, it allows you to retain some of the clarity that reverb tends to obscure in a dense mix.

I used this technique most recently on the new LP from New London Fire. We had one guitar track that just didn't have enough excitement (see Web Clip 2a), so I set up a similar chain and the guitar started to come to life (see Web Clip 2b). It saved the guitar track and helped the mix immensely.

The above technique is not just for guitars; it can be handy on multiple sources. This past summer, I worked on the new record by The Bravery (Stir the Blood, Island Def Jam, 2009) with producer John Hill. Throughout the tracking, we set up a variation of this chain in both the analog realm and the digital realm. First, we set up a send to an old spring reverb, then we ran it through my Smart Research C2 compressor and fed the sidechain input with various sources while we worked.

We also set up a chain in Pro Tools using the wonderful Universal Audio UAD Plate 140 and the Waves API 2500. We used the digital setup for more precision rhythmic effects, such as sidechaining from a click track or specific drum pattern. In some cases, we would send something like a snare to a fairly distinct reverb and then sidechain the compressor from another rhythmic element, such as the kick drum or the hi-hat. This can get fun pretty quickly, and doing it while the musicians are playing allows them to tailor their performance to the sound even more to exploit the effect.

The beauty of the sidechain is your ability to modify and manipulate what it's detecting. Because the compressor or limiter only grabs when the sidechain detects signal, you can aggressively alter the signal that is triggering it without mucking with the sound of the compressor itself. This is the basic principle of frequency-dependent compression, such as is used in de-essing or multi-band compression. But in this case, I'm not interested in the tried and true. I want to push some extremes.

One way to experiment with this technique is to set up an aux input with your console or DAW and send its output to your compressor's sidechain input. On the aux channel, experiment with different processors such as delays, tremolos, phasers, envelope filters, etc. Remember, the output is only going to the sidechain, so you will have to play around a bit with the compressor's threshold control and find a setting that interests you. Though it can be time-consuming, this approach can be enough to turn a track on its head and give you some needed inspiration.

You can also use it to fix problem areas in a particular track. For example, take the case of an overly muddy bass sound. By setting the pre-sidechain EQ to boost the areas of muddiness (say, boosting 220 Hz), the sidechain will react more aggressively to those frequencies. You can set the Q to be very tight so it grabs only on that frequency. It's basically like de-essing, but in the bass realm.

Image placeholder title
Image placeholder title
Image placeholder title

FIG. 3: Here are typical plug-in settings for a serial-compression-with-delay effect on a vocal track. All the plug-ins shown here are inserts. First comes a Massey compressor (a) set for mild compression, next is a Universal Audio UAD RE-201 (b) and finally a Chandler Limited TG 12413 limiter (c) set to a high ratio and quick release.

Serial Killer

When I need to sit a vocal track better in a mix, I often employ a technique called serial compression. It involves using multiple compressors — mostly with very low ratios and high thresholds — placed on the same source. I will usually set up the first compressor to hit the vocal mildly at, say, a 3:1 ratio with a slow-to-medium attack and a moderately fast release. Then I will insert a really short delay immediately after the compressor and set it up so that the signal is about 30-percent wet on the output. For the delay, I typically use something like a Roland RE-201 Space Echo, or even the UAD RE-201 plug-in (see Fig. 3).

After I set up the delay, I will insert an EQ if needed, but otherwise will go straight to inserting a limiter or a compressor with a very high ratio (20:1 or higher). I'll usually set the limiter to grab very aggressively and release very quickly so I can introduce a bit of distortion, as well as the breathing and pumping of the limiter while it exaggerates the delay on the vocal without sounding like a simple slapback. It will sometimes sound like a small room rather than a delay because the second limiter doesn't release until after the delay begins to taper off, creating a sense of excitement and space on the vocal itself. You can also do this with reverbs instead of delays.

Image placeholder title

Not only do old consumer cassette decks often have built-in limiters, but many times they also have an onboard noise-reduction circuit. After recording through it, try disabling the circuit for playback and you''ll get an interesting exciter-style compression.

Cheesy Squeezy

In the April 2009 issue of EM, I wrote an article (“Radical Recording Tips,” available at in which I talked about the use of old cassette machines as a way to manipulate individual tracks. Old cassette decks often have an onboard limiter or compressor built in. When you send tracks out to them, you can usually run through the line input and engage the limiter without even using a cassette tape and recording. As you might expect, these onboard units tend to operate poorly, so you can coax some extreme accidental sounds out of them, including distortion and filtering. What's even better, by running the signal onto a cassette tape, you can get some good old tape compression to boot.

Here's how to do it: Start by putting a beep or stick-hit tone at the beginning of your track because you're going to have to import the processed track back in afterward and line it up. Next, send your signal into the tape recorder and try to hit the tape pretty hard with your input so your meters drive into, or beyond, the red area. Cassette tapes have a pretty low threshold for compression, so you'll hear it clearly. In fact, you can drive a cassette tape into extreme compression very easily by cranking your input to the recorder. Purity and transparency aren't the goals of this process, so don't be worried if you get a little fuzz on your signal.

The last step is to record the track back into your DAW from your cassette deck and line it up with the original track using the beep you put in as a visual reference.

There's even more you can do with cassette machines. Many of the old recorders, as well as some open-reel decks, had integrated Dolby or DBX noise-reduction circuits (see Fig. 4). If your recorder has the ability to record with these circuits engaged, then the noise reduction will encode the cassette tape with its process. Then you can play the recording back with the noise reduction disengaged so that it doesn't decode the sound on the way out. Because these noise-reduction circuits are basically just enhanced high-frequency compressors, the undecoded playback will be very bright and aggressive. I've used this method to great effect on pianos and acoustic instruments as a way to make them sound a bit more alien and atmospheric, without the use of reverb.

This same principle can be implemented if you have an old modular Dolby or DBX multitrack studio unit around, such as a Dolby Type-A card with its own host rack. Simply engage the unit on the input, but not on the output. You now have yourself a complementary high-frequency compressor/expander.

Drive It Home

Finally, don't underestimate the use of overdrive and its resulting compression. Many times, I'll overdrive the input of a mic preamp or even a compressor itself to create a more distinct type of limiting that verges on the sound of electronics exploding.

For instance, if you overdrive certain mic preamps with a guitar, you can push it to a point where the circuit itself compresses the audio while distorting the life out if it, making the guitar sound as if it's being destroyed. It's a wonderful way to introduce extreme, heavily compressed fuzz on virtually any source.

D. James Goodwin is a producer/engineer in Woodstock, N.Y. He can be found twisting sounds at his studio, The Isokon. He has most recently worked with The Bravery, Norah Jones and Lapko. Visit him online