Compression Tips and Techniques John Freyermuth
According to Jerry Finn in The Mixing Engineer’s Handbook, “the sound of modern records
today is compression”. 1 Sadly for modern music, compressors are one of the most misunderstood and misused processors available today. Many engineers and audio enthusiasts do not totally understand what a compressor is, how it works, what the controls mean, what type of compressor to use, what to use compression for, and alternatives to using compression. With all of the technology and plug-ins at people’s fingertips, users are just slapping any compressor on a track or a mix bus, turning up the output and trying to make their tracks as loud as possible. In the war for the extra db, dynamic range is the first victim, and unbeknown to the thousands of audio engineers and musicians misapplying compression, their mixes are the ones that really suffer. In this paper the mysteries of compression will begin to be unraveled and slowly a working vocabulary of compression terms will be created. This paper will examine what a compressor is, and what the different controls for the settings on different types of compressor’s do, and how those controls affect your audio signal. It will also examine when and why you should reach for a compressor and when you should not. It will discuss alternative mixing strategies to compressing every track. Finally, the paper will examine some classical compression techniques, time tested by pros. The purpose of this paper is to give you a working knowledge of what a compressor does and introduce you to some new and interesting compression techniques that will allow you to get the most out of your mixing abilities. Dynamic control was not a major part of classical and jazz recordings. Classical and jazz engineers did not really pay much attention to shaping and controlling the dynamics of the piece with processing; they mastered the craft of microphone placement. But as far as modern recordings go, compression and all the side effects that come along with it are forever a part of modern recording. Compression is a major part of modern records, and it is of the utmost importance to have an in-depth working knowledge of what a compressor is and what it does.
2 According to Bobby Owsinski in his book, The Mixing Engineer’s Handbook,
“Compression is an automated level control using the input signal to determine the output level, set by using the threshold and ratio controls.”2 So a compressor is an automatic fader that proportionately reduces the dynamics of a signal that rise above a user defined level so that the loudest portions of the signal are reduced and the lower level portions of the signal can be heard.3 Compressors are very difficult for many people to understand and are viewed to a degree as being mysterious. This is partly due to the fact that if you compress a signal or a program correctly your work should basically go unnoticed. David Miles Huber and Robert E. Runstein provide another great definition of a compressor in their book Modern Recording Techniques. A compressor reduces input levels that exceed a selected threshold by a specific amount. This reduced dynamic range signal can then be boosted in level at the output, thereby allowing the softer signals to be raised above the program or background sounds.4 Compressors come in all shapes, sizes and controller layouts, ranging from simple two knob peak reduction compressors like Antares Painkiller or UREI LA-3, which only has a gain knob and a peak reduction knob, to the fifteen knobs on the Fairchild 670 compressor, which allow, for far greater user control, but not always better results. To get the best results out of compressors, we will define the different controls that can be found on different compressors. The input gain control on a compressor is used to determine how much signal will be sent to the compressor’s input stage. The threshold setting determines the level at which the compressor will begin to proportionately reduce the signal from the input gain control. Basically, threshold is the level at which the compressor considers the signal too loud. If you set the threshold to its maximum setting very little of the incoming signal will be compressed because the compressor will not recognize the signal is too loud. If you set the threshold to the lowest setting, even the softest parts of the signal will be considered too loud and will be compressed. The maximum setting for most compressors is 0db while the minimum setting can vary greatly
but is usually a number below 0db. An example of the threshold control in action is if the
threshold is set to -20db, all signals that fall below this level will be unaffected while the signals that are above this level will be proportionately attenuated.5 Some devices do not have a threshold and the input gain on these devices will correspondingly control the threshold level. On devices without a threshold control, raising the input level will lower the threshold point and thus reduce the overall dynamic range of the program.6 Some compressors have threshold controls that are referred to as hard knee or soft knee. A soft knee setting gently fades in the beginning of the gain reduction when the signal crosses the threshold creating a smoother transition from uncompressed to compressed signal, making the compression less noticeable. A hard knee setting causes the gain reduction to quickly kick in as soon as the input crosses the threshold, this effect can be more noticeable when the gain reduction begins, but it is better at handling large amounts of gain reduction. Another important control is output or makeup-gain. This control is used to determine how much signal will be sent to the compressor’s output. It is used to boost the reduced signal into a range where it can best match the level of a medium or be better heard in a mix.7 It is basically used to boost the level of the reduced signal back up to its original level or beyond. The ratio, or control slope, determines the slope of the input to output gain ratios. In simpler terms, it determines the amount of input signal that’s needed to cause a 1db increase at the compressor’s output.8 The compressor’s ratio effectively sets how firmly the compressor reins in signals that overshoot the threshold level.9 For example, with a ratio of 2:1 with every 2db increase at the input there will be a 1db increase at the output. With the same 2:1 ratio a 4db increase at the input would yield a 2db increase at the output. Not all compressors react the same to incoming signal at the same ratio settings, so it is important to watch the gain reduction meter display to see how much gain reduction is occurring. This leads us to another very important
4 control, meter display. This control changes the compressor’s meter display to read the device’s output or gain reduction levels.10 some devices do not need to switch displays because they display gain reduction and output simultaneously. There are two more very important controls to cover that determine the speed at which the compressor’s gain reduction reacts to changes in the input signal level.11 They are two of the controls that can help your compression sound its best, and also the two controls that if not used properly can wreak the most havoc on your signal. They are the attack and release controls. The attack control determines the how fast or slow the compressor will turn down signals that exceed the threshold. It is defined as the time it takes, in milliseconds, for the gain to decrease to a percentage (63%) of its final gain value.12 Attack is basically the speed at which fast the compressor reacts in reducing gain. Release is used to determine how slowly or quickly, in milliseconds, the device will restore a signal to its original dynamic level once it has fallen below the threshold point.13 This point is defined as the time required for the gain to return to 63% of its original value.14 Release is basically the speed at which the gain reduction resets to its original level. There are two main styles of compressors, peak reduction and RMS (root mean square), that control and affect the detection of the compressor’s signal input.16 Peak reduction looks for peaks in the incoming audio signal and when those peaks are detected and cross the threshold they are accordingly reduced by the user defined ratio. Peak reduction works much quicker than RMS and is great for handling the sharp transients and the short peaks of percussion tracks. Peak reduction also works most effectively over short time periods. A good example of when to use peak reduction is while working with drum tracks and you do not want the peaks of the kick to overload the dynamic range of your program material and the medium you are working in. Another good example of when to use peak reduction is when you are trying to tame the initial
5 transients of a slap bass line or a drum beat where the drummer becomes a little over zealous on some strikes. You would use peak reduction to bring the signal level closer to that of the rest of the material in the program in this case. RMS, works a lot like our ears when they hear and adjust to loud sounds. RMS compression works by detecting a signal’s average level and reducing the signal based on the average level of the material.17 In RMS mode, short peaks do not cross the threshold and cause the compressor to reduce the overall level of the signal. This allows for a more dynamic signal and more transparent compression that still prevent the signal from being too loud. RMS is often used if you are trying to raise the apparent volume of the track during mix down or mastering because it does not react to every small peak and therefore does not squish the dynamic range of the material as much as peak reduction. There are four main types of compressors that are found in the analog and digital world. Most of the compressors in the digital world are based on or are emulations of their famous analog brothers. The four main types of compressors are optical compressors, variable mu tube compressors, VCA, voltage controlled amplifier, compressors, and FET, field-effect-transistor compressors. A variable mu compressor uses a valve, instead of vacuum tubes, in its signal chain to achieve compression, by re-biasing the tube to control the amount of gain reduction.18 In the audio world, the word mu refers to gain, so a variable mu compressor is basically called a variable gain compressor. The sound characteristics associated with a variable mu compressor are that of well designed tube equipment, warm, with a slight but altogether pleasing harmonic distortion and a smoothness that is hard to achieve with other compressor designs.19 Many people associate the sound of variable mu compressors with that of the classic analog recordings of the 1960’s and 70’s. Some famous examples of variable mu compressors are the Fairchild 660 and 670, the Manley Variable Mu and Pendulum’s ES8.
6 Optical compressors usually have a very transparent sound because of their particular
construction. The photo-optical detector circuit that controls the amount of gain reduction can work with a minimal number of components, allowing the signal to be altered by less material, therefore imparting less of the compressor’s sound characteristics on the signal, allowing it to be more transparent.20 Because of their transparent sound, optical compressors have found a place in many modern recordings. Popular optical compressor models are the Avalon AD2044, Joemeek SC2, and Millennia Media TCL2, which are very popular in classical recordings, where transparency is a necessity. VCA, voltage controlled amplifier, compressors apply more or less gain depending on the level of control voltage they receive. When that control voltage is derived from the audio input itself, it can be used as a compressor.21 VCA compressors have very unique sound quality that differs greatly from the aforementioned compressor models. According to Paul White of Sound
on Sound, and the many topflight engineers he interviewed for his May 2008 article “The SOS Guide to Mix Compression”, “VCA designs have a colorful sound quality, but the color is quite unlike that obtained from tube designs”. 22 This colorful sound quality of VCA compressors has made them a very popular choice amongst modern engineers for mix-bus compression. Some very famous VCA model compressors can be found in famous solid-state mix-bus compressors including the Neve 33609, SSL G-Series Compressor, API 2500 and the heralded Focusrite Red 3.23 Field-effect-transistor, FET compressors, use a transistor to control the amount of gain reduction, similar to the way a triode tube works.24 FET compressors have a distinctive bright and focused sound quality that has often been described as aggressive. FET compressors have the fastest attack and release time of any type of analog compressor.25 Their super fast attack and release make them a favorite choice for precision compression and for dealing signals with very
Â sharp and quick transients. Famous examples of this compressor are the UA 2-1176 and the
Cranesong STC8. Along with these analog compressors there are countless digital emulations of these classic hardware designs, some of the best being made by Universal Audio. Also there are unique digital designs and hardware designs that are based on the four analog device types and combine characteristics of all or some of the four types of compressors sonic signatures. There are also digital and hardware compressors that allow you to switch between different compressor types, giving you the Swiss army knife of the compressor world. Now that we understand the controls of a compressor and the basic types of compressors, we can now examine when to use a compressor in a mix situation. A compressor should be used for two main reasons. The first being to control the dynamics of a signal or of an entire mix. The second reason to use a compressor is as an effect. For controlling the dynamics of a mix, the goal of the mixer is to achieve a static fader level for each of the individual track and for the main mix-bus. If you can find a static fader level and it sounds good then you do not have to compress. If it sounds good just leave it alone. If you cannot achieve a static fader level by automating the volume of a track or mix, or by EQ or multing, splitting the track out over multiple tracks and adjusting the fader level for the new tracks, then you can reach for a compressor. You can use a compressor in tandem with fader riding/automation, EQ and multing, as long as you achieve the right balance for that mix. There are many different schools of mixing that have all different types of rules for when and where to use compression or if to use compression at all. In the end you can break every rule or make up a new crazy rule that works for you, as long as you achieve your desired result. On that note here is a short list of suggestions on when you should compress to control dynamics if automation, EQ or multing is not enough.
8 Â A compressor should be used to minimize the changes in volume of an instrument or
vocal performance when they are too great for the mix.26 A compressor can also be used to balance out the dynamics of a single instrument, like a bass, where output volume can vary greatly from string to string. A compressor can be used to balance out momentary changes in the distance from mic to source. A compressor can also be used on a stereo mix-bus to reduce the dynamic range or boost the average volume of a mix so that it appears to be significantly louder.27 Compressors are also used in the mastering process, but we are not going to address that. Compressing to control gain should be done in moderation because you can quickly suck the life out of your track, or even worse an entire mix, if you over compress a signal. But do not be afraid of compression because when used correctly, it can solve a whole host of issues. For example, if you have a vocal track were you cannot quite understand all of the lyrics, because some are far too soft, a compressor can be used to allow you to reduce the difference between the loudest and softest parts allowing every word to be intelligible. When using a compressor to control the dynamics of the performance it is always a good idea to use small amounts of compression at first, like a ratio of 2:1 or 4:1. When controlling dynamics, compressors can be used as inserts on a channel strip affecting a single instrument, placed on a mix-bus, affecting a selection of signals sent to the bus, or placed on an auxiliary track where they can be blended with other signals as an effect. A compressor can also be placed on the master mix-bus to control the overall dynamics of a program. There are also many different combinations of the aforementioned methods used to achieve the desired results. For example, some engineers when working with upright bass tracks will use serial compression, two compressors inserted on a track one after the other to achieve the needed balance. The bass signal will pass through the first compressor that reduces the peaks
9 of the signal. Then the compressed signal passes directly into the next compressor that reins in the dynamics of the instrument, creating a smooth and controlled bass line that has undergone large amounts of compression but has not been squished to death. The two compressors usually used in serial compression configuration like this are a peak reduction compressor followed by an RMS compressor. The effects that can be created by compression range from subtle distortion to all out sonic mayhem. According to Andy Johns, engineer of Led Zeppelin and The Rolling Stones fame, states that, “I use compression because it’s the only way that you can truly modify a sound.”28 With the right compressor and the proper settings you can alter the volume envelope of the sound, making it cut through the mix and sound more aggressive and assertive. You can then alter the decay, making the sound longer, and making the sound “fatter”. With the correct compressor and the correct settings you can make it seem like all of the elements of the mix are moving to the front of the mix (at the listeners) creating a very exciting effect and bringing a lot of energy to the piece. Other engineers like Joe Chiiccarelli use compressors like the Teletronix LA2A to create beautiful distortion on vocal tracks by overloading the final tube stage to create a nice distortion effect. This distortion effect created by the LA2A worked particularly well with Joe Chiccarelli’s approach to working with Jack White’s, of the White Stripes, vocals.29 Other engineers take advantage of the unique colors the sonic pallets of all the different types of compressors, using them to process single instruments, vocals and groups of instruments and vocals. They run the signals through the compressor, sometimes while the compression section of the unit is bypassed, just so the unique sonic qualities of the unit can enhance the signal. Other times they slam the signal with compression and alter the attack and release to really get the signal to pump,
Â usually an unwanted side-effect, but in this instance a desired effect to get the signal to
breathe with the groove of the music and really be in the listeners face. An often-used effect when mixing percussion is to take the signal from the overhead mics on the drum kit and the signal from the room mics and really slam them with a ton of compression. Then you gently mix the compressed signal back in to the mix against the discretely miced percussion elements. If you recorded in a good sounding room this really adds to the mix and creates a really great sounding and natural reverb effect. But beware because if you record in a room that sounds bad and you try this effect it will most likely sound really bad. This is because this effect is bringing the normally lower levels of the sound of the room to a higher level in the mix allowing them to be heard. Another highly utilized effect is parallel compression, also known as the New York
Compression Trick. Which is used to get a very thick and rich sound, that sounds much bigger but not overly compressed. To use parallel compression you bus your signal to a stereo compressor and compress it with at least 10db of gain reduction, even more if you like. Then send it back to the mixer or DAW where it will return on two separate tracks. On one track add some high end, usually 10khz or above, and on the other track boost the low end around 125Hz. Now mix these signals against the original uncompressed signal. You should have a much richer and fuller signal, without any side effects of over compression. This trick works because of the difference in the compression curve of parallel compression verse insert style compression. With insert style compression the loud parts of the signal are turned down to reduce the difference between the level of the soft and loud signals. When using parallel compressing the compression curve works inversely and the soft parts are turned up to match the louder parts of the signal.30 So that when the parallel compressed signal is mixed back into the uncompressed signal the soft parts of the signal are raised and the louder parts are not reduced allowing for a fuller richer
11 sounding signal. In the digital world phasing can be a problem with parallel compression but as long as you keep an eye on the amount of delay your signal processing and A-D and D-A conversion is creating, and if you utilize some systems automatic delay compensation, you should be able to tame your delay problem and reap the benefits of parallel compression. In my experience I find that parallel compression works great on an array of signals ranging from percussion and bass to lead vocals. Try it out and go nuts. Mike Senior describes another classic effect that harkens back to the analog days of yore, in his article Classic Compressors for “Sound on Sound”. Senior describes the use of precompression fader rides and chaining compressors to get a similar sound that is associated with the days of all analog recording. Senior describes how back in the days of analog recording the signal was routinely compressed multiple times. This started in the recording process where most signals were lightly compressed to increase the signal to noise ratio. Then the tracks would be further compressed in mix down with another model compressor imparting “a generous helping of tasty analog-processing side-effects.”31 Also engineers rode the faders either into or out of compressors during tracking, creating super smooth compression, a signature of the great analogue recordings. Senior points out that fader rides in and out of a compressor during tracking is all but a lost on young engineers because of the advent of digital recording and 24bit A-D converters and the massive amounts of head rooms they allow.32 Fader riding is a lost art and in my opinion is one of the reasons people are having trouble creating the life and feeling that used to be imparted in analogue mixes. We are missing out on the performance of the engineer and how he interacted and fed off of the performers energy from the control room. That is part of what I feel gave so much life to analogue mixes and would be an awesome enhancement to digital mixing.
12 Today the master of compression for dynamic control and for effects is mix engineer
Michael Brauer. Brauer’s all encompassing and liberal use of compression can be heard on the new Coldplay album Viva La Vida. Michael Brauer is the master of parallel compression and multi-bus compression, where he is know to process everything from multed vocal tracks, to groups of instruments, through his unique multi-bus, four-way stereo output-bus system, compression system he has deemed ‘Brauerize’, a term which he has trademarked.33 Brauer takes the signals from the four buses, which are labeled A, B, C or D and sends each one to one of the five different racks he has routed with compressors and other signal processing equipment. Brauer then brings the four buses, A, B, C, or D, back up in the center of his SSL J9000 console. Brauer also has a five compressor serial chain set up exclusively for compressing vocal tracks.34 The thing that sets Brauer apart from other mixers, aside from his esoteric use of his massive collection of rare and strange outboard gear and his willingness to compress anything, is that he mixes with post compression, while most other people mix with pre fader compression.35 What this means is that Brauer mixes into compression, meaning that all his fader moves happen before compression and that his fader moves feed the compressors inputs. This is one of the things that help to define his style and give him his signature sound.36 Now that we have gone over basic compressor controls, the basic types of compressors and their sonic characteristics, looked at when and why you compress, as well as alternatives to compression and reviewed some basic examples of compression for dynamic control and compression for an effect, we are ready to go out and start mixing. Now that the basics of compressing are covered we are free to go out and apply the rules we like and throw away what we do not like. There is no set rule for compression only suggestions. My suggestion is to try it all and find what works for you.
Haas, Will. "The SOS Guide to Mix Compression." Sound on Sound, May 2008. http://www.soundonsound.com/may08/articles/mixcompression.htm (accessed January 24, 2010). Houghton, Matt. "Analogue Warmth." Sound on Sound, January 2010. http://www.soundonsound.com/feb10/articles/analoguewarmth.html (accessed January 20, 2010). Huber, David Miles, and Robert E. Runstein. Modern Recording Techniques, Sixth Edition. New York: Focal Press, 2005. Moylan, William. Understanding and Crafting the Mix, Second Edition The Art of Recording. New York: Focal Press, 2006. Owsinski, Bobby. The Mixing Engineer's Handbook. Belmont: Thomson Course Technology, 2006. Robjohns, Hugh. "Mix Mastery." Sound on Sound, June 1999. http://www.soundonsound.com/jun99/articles/mixcomp.htm (accessed January 20, 2010). Senior, Mike. "Classic Compressors." Sound on Sound, September 2009. Senior, Mike. "Compression Made Easy." Sound on Sound, September 2009. Tingen, Pual. "Secrets of the Mix Engineers: Michael Brauer." Sound on Sound, November 2009. http://www.soundonsound.com/nov08/articles/itbrauer.html (accessed January 24, 2010). White, Paul. "Advanced Compression Techniques." Sound on Sound, January 2001. http://www.soundonsound.com/jan01/advanced.asp?print=yes (accessed January 20, 2010). White, Paul. "Advanced Compression Techniques." Sound on Sound, January 2001. http://www.soundonsound.com/jan01/advanced.asp?print=yes (accessed January 20, 2010). White, Paul. "How does Parallel Compression Work?" Sound on Sound, March 2009. http://www.soundonsound.com/sos.mar09/articles/qa0309_3htm?print=yes (accessed January 24, 2010).
Published on May 9, 2011