What this post is: first, a discussion of what compression is and how it works, touching on studio compression and focussing on compression as a guitar effect.
Second (and most importantly), an extended audio comparison of five guitar stompbox compressors: original MXR Dyna Comp, DOD Milk Box, Jangle Box, Barber Tone Press, and the obscure Stigtronics Compressor (an enhanced Ross clone).
Why you might want to read it: A, you're a relative newcomer to compression or a complete skeptic, and you want to know what it's about.
B, you're an expert and either want to challenge what I have to say or collaboratively correct and expand on it.
C, in either case, you want to compare the performance of these compressors. (To get straight to the sound samples, scroll way down.)
D, you don't mind my playing.
Why you might not want to read it: you don't care about and will never care about compression; you already have the ultimate compressor and have more important things to do with your time; my posts are way too long; you hate my playing.
Folksy introduction to goals, considerations, and results. Conceptually, the shootout is simple: play the same guitar material through several different compressors, at a couple of different settings, and record the results – so we can compare the performance of the compressors.
Hardest shootout I've ever attempted.
I intended to include enough variety of guitar techniques to cover the various effects compressors can have, which differ with pickups, pick intensity, effects, musical application, and compressor settings. It had to include single coils and humbuckers, light and heavy fingerpicking, clean chording, crunch, fast and slow lead both clean and dirty, etc. It was a challenge to come up a program covering all that.
To keep the comparison useful, the program had to be iDENTical through all the compressors. The solution was to play into a looper pedal first, then feed the signal chain from it. I didn't want to try punching in or patching it together in the looper, so I had to play the entire program acceptably in one pass. Since I'm lucky to play for 2 or 3 minutes without embarrassing myself technically, it meant lots of takes.
The final bit came together in a night, but I've been working at it for weeks. Then I changed the program at the last minute anyway.
Worst, when I was done I realized I'd left out two most typical compressor applications: clean chicken-pickin' and endless lead sustain.
Nonetheless, there's still plenty here to listen to: approximately four minutes of music employing various techniques, played back "naked" (with no compressor), then at two different settings through each of five compressors. That's a total of 11 tracks, and 44 minutes of compressor madness.
Not for the faint-of-patience.
Production Protocol
Signal chain: anonymous electric guitar --> Boss RC-20XL Loop Station --> Boss BD2 Blues Driver (Keeley mod) --> Fender Concert tube amp. Recorded at 16" from the amp by Edirol R-09 stereo "24-bit WAVE/MP3 Recorder" as 192k mp3 files.
Volume and tone pots on the guitar were wide open throughout; only pickup selections were changed. You'll hear all the switching. No settings were changed on the Blues Driver, at the amp, or on the recorder throughout.
After the program was played into the looper, it was used as the sound source (instead of live guitar) for the compressor tests. The dirt pedal was stomped "manually" every time (and not always at the same time).
The compressors were swapped into the same position in the signal chain one at a time (they were not all connected simultaneously, though that might be fun). Each got two passes, first with subtle settings and then more extreme.
The mp3 files were opened in Audacity on the Mac, and trimmed at front and end for consistency. The files were normalized to -1.5dB peaks so that all volume differences in the files result only from the performance of the compressors.
About Compression
Seems whenever there's a discussion of compression in a mixed audience of guitarists, there are some who have a lot of experience with it and some with none. Everyone knows what a distortion pedal does, but, as a more transparent effect, compression is less well understood. Part of the purpose of the exercise is to clearly demonstrate what a compressor DOES (by OVERdoing it), as well as how compressors can be used to improve tone without being glaringly obvious.
A micro-overview of compression might get us all on the same page. Parts of the following are intended for compressor newbies; other parts are for old hands. I mean to insult neither. Take or leave any part of the discussion for what is useful to you.
In the simplest form, compression is simply automatic volume control. If you had an assistant with impossibly fast and precise reflexes, and you could assign him to watch a level meter and make sure your signal never went above "X" on that meter, he would "ride the gain" on the channel and "compress" your loudest peaks down to your limit.
If that was ALL he was doing, he'd really be acting as a limiter (which is related to, but not the same thing, as a compressor).
Only he can also take the extra energy from your over-limit peak and save it to release as your note falls below the limit, pulling that level up – providing a more consistent level throughout a performance.
In fact, what if you could also have him make sure your level never fell BELOW "X" on the meter, no matter how lightly you picked? Technically, he'd then be operating as an expander. (Another related device.) And, again, he would serve to provide a more consistent level.
On the positive side, your magical golden-eared, sharp-eyed, instantly-reflexive assistant could help you stay consistently audible throughout a performance, without ever overbearing, unintentionally blurting or whispering a note, or getting too quiet.
On the negative side (depending on how you look at it), he's limiting and compressing your dynamic range.
Both things are true.
In practical application, most studio rack-mount compressors – and all guitar stomp-box compressors I know of - provide all these functions together: limiting, compression, expansion. They can boost low-level signals and squash high-level signals, and because of the design of their control circuits, they often do these things interactively. More about the control circuits in a moment.
Transparent vs "Special Effect" Compression
But I want to think more about how compressors bring the simple concept of automatic level control into reality, and the way they sound AS they do so – because understanding that is key to understanding why some guys like compressors, used in certain ways, and other guys don't.
Go back to your audio assistant. Let's assume he can't see into the future (some actually can), so he can't know when you're going to fall below or go above the volume threshold you've set. He can only react when you do it. This means that, no matter how fast he is, there will always be a split microsecond (at least) where the guitar gets through at its uncompressed level, and then you'll hear the result as he twists the knob or pulls the slider to adjust you to the "right" level.
Obviously these adjustments will be more or less audible. He can be told to yank it down (or push it up) very suddenly as soon as a compressible event happens – or he can be told to do it more slowly and gradually. Remember too that he's saving the extra electrical energy from your "overs" and releasing it to raise your "unders." What if you game the results by really pushing the volume, so that you're practically ALWAYS over, so that you're distorting the input amplifier on the channel and no matter what he does at the output, you're clipping a bit – and so that when you DO fall back below the threshold, he has a lot of collected juice to release, and has to keep pushing the slider up to let it out?
These "artifacts" of the way compressors work, the audible effect of their chasing your level in order to control it, are what we hear when we do hear compression. And sometimes those are desirable. I'll come back to THAT in a moment.
First I want to say that sometimes we DON'T want to hear the effects of compression. We want our parts to lay in consistently, with even attack and a Chet-like smoothness, to be audible without eating the mix. Maybe we want a little more sustain at the end of chords and notes. We want sudden overs to be invisibly controlled. We want to shape the volume envelope of our sound without it being audibly obvious that we are doing so. We'll call this transparent compression. It's frequently what we want out of high-quality rack-mount studio compressors. Sometimes we want that from stomp-box compressors too.
The truth is that most stomp-box compressors, and all but the best studio comps, do not do their jobs entirely transparently. Besides sometimes making it obvious that levels are being adjusted on the fly, they also have a subtle, grainy, fuzzy overtone especially at extreme settings. They can filter out highs. These are not usually desirable "artifacts," but they are usually there. Suffice to say that if we're after transparent compression, we'd rather not hear them.
But now let me come back to those times we want obvious compression. We want the effect, we WANT to hear magic fingers pushing and pulling the volume, pumping the channel. This is generally the domain of stompbox compressors. Most compressor-freaks are attracted to the rubbery squeezy sound of extreme compression before we appreciate its more subtle and refined uses.
All well and good. Distortion was not originally a sought-after "good thing" in the performance of audio gear, but we've learned to use it musically, seek it, and love it. Compression was originally designed just to squeeze big sound into small signal containers (like mechanical records and tape); during WW.II, it also became useful to keep voices in an audible range during radio and signal broadcasts. Originally, no one was interested in the accidental side effects – that we could hear the volume changes as they happened.
Audio engineers and producers originally used compression as transparently and INvisibly as possible. As the 50s wore into the 60s, though, some heard musical value in more apparent, even extreme settings. Studio compression is crucial to the sound of 60s pop. And it was during the mid-late 70s that special-effect compression was designed "down" into stomp boxes specifically for guitarists.
What ARE those special effects, and what do you listen for? Always, listen for the attack and decay characteristics of notes and chords. A very compressed note will have a quick pop at the beginning, then a sudden "suck" down, and usually a quick swell up at the end as well. A very expanded note will sound very faint at first, then gradually (or quickly) swell up "artificially."
Listen for chords that keep swelling as they ring, while they should be decaying, that seem to feed back into themselves and just keep growing.
Listen for long held lead notes that seem to sustain forever. These are the kinds of effects that more gain can produce – but compressors can produce them without saturation or the grainy fuzz of distortion.
Notes with unnaturally percussive, clicky attack are compressed.
Listen for phrases where the attack of every note seems overly and supernaturally cleanly articulated – and phrases that sound, from the articulation and tone, as if they are being played very lightly and softly, but which are as audible as if they were played loudly. At the other end, listen for notes whose tones sound like the strings are being ripped from the guitar, but which are no louder than the soft bits.
Ultra-even, very detailed and articulate power chords often owe a lot to compression. (Tom Scholz of Boston, and his Rockman, rely heavily on compression.)
Compressors are also very useful to hit a dirt pedal or the front end of an amp harder, and get the illusion of more volume, and more saturation, without shredding ears.
A very rubbery, elastic guitar sound is probably compressed, as is classic 70s-present Nashville chicken-pickin'. (I don't know about rubber chickens.) Funk guitar is often compressed. These have in common that a Telecaster, played hard, has a naturally compressed tone – but stomp boxes greatly magnify it.
Compressor Control
Studio compressors have a fairly consistent, and musically/electronically logical set of controls. Not all compressors have all of these (some do); most have most. (And this is a very rough overview, provided only because it breaks the "steps" of compression into logical components.)
• input gain: to adjust the input, either simply for signal matching or to force the device to compress the signal more or lessGuitar compressor designers and builders have reached no similar consensus about what knobs ought to be on their boxes, or what exactly those knobs do. They know that guitarists quickly glaze over if there are more than two or three knobs to deal with, so they combine various sets of the above functions in the action of a single knob. The characteristic sound of different compressors result from just HOW the functions have been combined.
• threshold: the level at which you want the device to start expanding or compressing; usually expressed in dB
• attack: how quickly the compressor reacts when it reads a signal either falling below the threshold to be expanded, or above the threshold to be compressed; usually in ms
• slope/contour: related to attack, but controlling how gradually and consistently level is adjusted from the threshold to the target level; also called "knee"
• ratio: how exuberantly the expander amplifies a qualifying signal, or how hard the compressor squeezes it; expressed in ratios from 2:1 to infinity:1. (I've always thought "infinity" was probably more a theoretical ideal than something a mortal electronic circuit can deliver. Joe Meek excepted.)
• release: how quickly the device lets go of the signal when its input returns below the preset threshold; also in ms
• output gain (also called makeup gain): the output volume of the device, usually to bring the volume of highly compressed signals back up to the level they would have had without compression
Each classic design takes its own approach, and there is no consistency in nomenclature. Sometimes you don't know WHAT the hell they mean by the label till you twist the thing. In general, though, one knob is usually a "compression amount" control, and another is overall volume.
Because studio compressors have such a complete complement of controls, it's possible to make different brands "sound" very similar to each other (especially the more "transparent" ones). Guitar compressors, on the other hand, embodying their designer's decisions on our behalf, sound as distinct from each other as dirt boxes.
It follows that some excel at some things, and others at different things. And past the seemingly requisite two knobs, it's the additional controls on various stomp compressors that dial in their unique capabilities.
OK, meet the contestants.
Some of the compressors in the shootout tend to be pretty obvious-sounding; others tend to transparency; all will operate across at least a range of that continuum. They range in price (new) from 50.00 to 225.00. All are available today.
And a few words about each one. But not about their SOUNDS – you'll make up your own mind about that.
MXR Dyna Comp: one of the first mass market stomp compressors (along with the Orange Squeezer), and still arguably the leader in the market. The first version had "MXR" in script, and is more highly prized by pedal freaks. The next version has "MXR" in block letters, and that's the one I've had since 1979 (and for about 10 years, if I was plugged in, it was on). Controls: Output, Sensitivity.
(The first Dan Armstrong Orange Squeezer, circa mid-70s, wasn't really a pedal, just a small box, and had NO knobs. On or off, baby. But it became a classic effect all the same.)
DOD FX84 Milk Box: as many knobs as a boutique job, but at popular prices. The first version called the controls Quarts, Cream, Pasteurization, and Spill; the manual revealed the code, and later versions are labeled: Level, Compression, High Expansion, and Attack. The high control is useful, as it lets you bring back the high-end content that stompbox compressors often lose. On the second clip of the Milk Box, below, I played with both the High and Attack controls as it played.
Jangle Box: the most recent design here, and the darling of the Rickenbacker crowd. A good all-round compressor, its main trick is its "bright/dark" switch, which re-eqs the pedal to accentuate or mute highs. Most guys use it to push the highs, and it uncannily reproduces the EQ and compression used on McGuinn's 12-string on The Byrds' hits. If that's what you're after, there's nothing better – and you get a competent compressor in the bargain. Controls: Gain, Attack, Dark/Bright. On the second Jangle Box clip, the bright control is alternately engaged. You won't be able to miss it.
Barber Tone Press: a relatively recent design, the TP includes – along with the main controls Volume and Sustain – a clever Blend control which mixes the compressed signal with a variable ratio of the unprocessed signal, passed clear through. One of the dead giveaways of compression is the percussive suckdown at the beginning of the note. The Barber lets you mask this with the unprocessed tone – which then decays and lets the articulation, smoothness, and sustain of the compressed tone swell up through. Pretty smart.
Stigtronics Compressor. The other early leader in stomp compressors was the Ross (the MXR is actually based on the Ross, but outcompeted it in the market). Ross went away fairly early in the game, and now the classic blue/gray Ross design is the basis of most boutique and clone compressors. This particular version is from a small builder in Bloomington, Indiana. I've tried Mike Stiglitz's dirt boxes, and they all impress me; I was anxious to get his compressor next to these others. I plugged his in first, and lost about three days just playing through it. I didn't want to unplug it. It's a first-class Ross-inspired design to which he's added an attack control. In short, I think it's a great box – amazingly clean and transparent in tone, but pumpy enough in effect. Easily the equal of anything here. Controls: Volume, Sustain, Attack.
Random Notes about the program material.
One of the functions of stompbox compressors is to boost low-level signal. I've left a few seconds of dead air at the beginning of each clip so you hear how much "noise" this adds when there's no signal at ALL. (Natural by-product of compression, just part of the package.) By the same token, switch clicks and other low-level noises are sometimes boosted as well; you'll hear that in the clips.
I purposely alternated both "too light" and "too heavy" playing in the program in order to demonstrate the compressors' action under those conditions. The inconsistent and extreme dynamics are intentional, as are too quiet and too loud passages.
Finally, the audio clips!
Uncompressed original signal.
Dyna Comp, light compression.
Dyna Comp, heavy compression.
Milk Box, light.
Milk Box, heavy.
Jangle Box, light.
Jangle Box, heavy.
Tone Press, light.
Tone Press, heavy.
Stigtronics, light.
Stigtronics, heavy.
What compression looks like.
(You can click the pics to hear the clips; right-click or command-click to open the clips in new windows and keep the pics onscreen.)
Pics of the compressors in action on this bit of music. The timeline is across the top of the screen captures from Audacity; if you play one of the clips and watch its elapsed time in your audio player, you can track the progress against the timeline and see how the compressor tracks and modifies the dynamics of the material.
Simply by comparing the uncompressed original to any of the compressed programs, you'll see that, in general, quiet bits are boosted and loud bits are squeezed. But the differences in the way the compressors respond to the same material are pretty dramatic.
And that's all I have to say. Time to decompress.
But what guitar did I use?
