- Posted: 14/03/26
One of the more complicated and misused elements in a mix is often dynamic range compression, simply because there are so many variations on its use. At least, that’s one of the main things that I attribute to the misuse of compression. Maybe a lot of beginners don’t understand it, or are overwhelmed by the various controls and options. At any rate, it goes without saying that to really grasp your mix (or any mix for that matter) a comprehensive knowledge of what compression is used for, and what the various controls do to your sound is compulsory. For the fourth installment of the SPS let’s delve into it, and possibly demystify some of the everyday elements. In this article we will focus on the use of compression as a whole as I think articles based on certain types of dynamic range compression are great and informative; this article’s goal is to leave the mixrevu.com community with an overall understanding of the controls and why we use compression at all in our recording and mixing. I believe that way we can walk away able to approach even the most complicated emulations or variations with at least a foundation of where and how to use this tool to get the results we want.
“If it sounds good, it is good.”
I can’t remember who said that above quote, but I just want to start the article with that statement. I think we’ve all heard in audio that at the end of the day “there are no rules”. That statement goes for compression as well. There are always new ways to use the tolls at our disposal, so I don’t want to ever convey that “there is only on way to skin a cat…or mic a guitar cab, lol”. If it is working for you, by all means do what works. It’s also important to note that some of the most inspiring new uses of different technologies in the pro audio domain come from engineers who don’t just come by their devices by happenstance, but develop new and odd ways to use their tools because their depth of knowledge is that much more comprehensive. (i.e. If you know the ins and outs of compression you are more likely to find new ways to use it based on your comprehensive understanding of the tool and the scientific bases that its founded upon.) Just a little to think about…
Let’s tackle the compression topic by defining why we use compression first and then we can move into the many parts of a compressor that we might come across during our use of plugins or analog gear.
So, why do we use compression and what can it offer our tracks?
(JDK API 500 series compressors, courtesy of Rayford Smith)
Compressors like the ones above are used to control the dynamic range in our audio signals. When listening back to recorded tracks that are unprocessed you may notice large differences between the loudest and quietest portions of your audio signal. Often fader movement and track automation just won’t work to make things as consistent as needed, or if they will work, are too cumbersome a task. This is where compression kicks in, at the most basic level it’s used to lower the dynamic range of your signals. Just a quick detour here:
Audio Dynamic Range – The difference between the loudest and softest parts of your audio signal would be its dynamic range. So this would NOT BE an AVERAGE difference, but an ABSOLUTE difference between the peak signal level and the lowest signal level of the track, ex. that whisper in the second verse and the belting vamp in the outro of the song.
When you listen to your favorite record on the radio, or on iTunes, or when you listen to a morning radio show, or to your local T.V. anchor broadcast the news, you will notice a consistency in dB level to their audio. If you listen closely you will notice very little variation in the quiet “whispers” and the loudest “yells” in volume. This is the use of compression in action. Compression helps keep a lead vocal tamed and helps keep the strums of your rhythm guitar at the same relative level as the screaming solo. If we are looking to make professional sounding records, compression will almost always be used to some extent.
Now that we know why we use compression and where to look for it (often time’s compression is one of the most difficult effects to identify aurally), let’s describe the elements that make up a compressor.
Input gain- This, of course, is the level that you are sending to the compressor. If you alter this after the other controls are set this absolutely will affect the amount of signal that passes the threshold and is therefore compressed. I’ve found it beneficial to set the input gain where I want it and then move on to processing the signal. Of course on certain compressors (mainly VCA’s) you use your input gain to actually control the amount of compression, but that’s down the road…
Threshold- This is the control that once set, any signal level that rises above the threshold has processing applied to it. So you can set a high threshold to just effect the peaks of the signal or you could set a low threshold to have the majority of the signal processed.
Attack- This function adjusts the time the signal (that has passed the threshold) has before the compressor acts upon it. This is described in milliseconds. Adjusting this control can help to add “impact” on percussive sounds and allow transients to slip through on signals that might be attenuated and yet still maintain a sound that isn’t “dulled”.
Release- This is the time that it takes for the compressor to return the signal to pre-processed level after it has been attenuated by processing. This control can be set to auto, therein the compressor releases relative to the source material. If the release is too short you can get a “pumping effect” that if not purposefully used on percussive sounds may be undesirable. If this is set too long it is possible to miss the attack of the next signal because the compressor hasn’t released yet and so your subsequent note attacks are being detrimentally effected.
Compression Ratio- This is where the math comes in!!!! Well, not really for the end user. It’s better explained by an example; if your ratio is 4:1 (a ratio I often use depending on the compressor) this signifies four dB of compressed signal relative to one dB of signal output above threshold. So you are net 3dB of attenuation at your output stage. A ratio of 8:1 would then be 8dB of attenuation to every 1dB of signal output above threshold, with a net 7dB attenuation.
Knee- This describes the way in which the dynamic compression takes effect on the signal. A hard knee means a quick compression once the signal crosses the threshold and the attack time is reached. A soft knee will mean a gradual compression once the attack time is reached. For instance, a hard knee lends well to quick signals like percussive sounds and a soft knee might lend well to inconsistent signal lengths like vocals. Of course these aren’t hard and fast rules (pun?), there are many times where you might use a hard knee on sounds that aren’t percussive etc.
BITE- The above functions that were described are functions you will commonly see on most, if not all, compressors. The BITE (Bi-Directional Transient Enhancement) control I have only seen on a certain type of processor, but when I first used them I was really interested in this control and so I thought it’d be helpful to paraphrase how this works. The BITE control basically allows transients of your signal to pass through uncompressed while the rest of the sound remains processed. How is this unlike the attack control you ask? Well the attack control is a predefined time before the entire signal is processed, whereas the BITE control allows transients to peak through throughout the signal processing. I’d have a look at McDSP plugins if you are interested in learning more.
Look-Ahead- This function splits the input signal and delays the processed signal. Fairly simply, it monitors the split signal and then processes the delayed signal so it “knows” what’s coming. The look-ahead function does delay the output signal as this is the signal that is processed so that should be kept in mind. If you are using a D.A.W with delay compensation then you shouldn’t notice a hitch but it’s important to know what is happening to your signal either way.
Make-up Gain- Remember the conversation we had with the compression ratio? Well we talked about net attenuation as it relates to the amount of compressed signal versus the amount of signal that is output over the threshold, well with the make-up gain we can “make-up” the difference here. This is where if we have 4:1 compression we know we have a net 3dB signal attenuation, well in our make-up gain we can bring the output to +3dB to bring our signal back to relative unity. So, we have achieved the audio dynamic range reduction through the processing of the compressor and still kept the signal at the same volume! Pretty cool hunh? This is where we have made one step closer to the consistency we discussed earlier.
Whew! Now that we know why we use compressors, what compressors do to our signal, and the controls you’ll see on most compressors, we should have a nice grasp of what to do when we come across any compressor in our gear or plugin library.
What about the different styles of compression?
When I first learned how to use a compressor effectively I thought I had the “compression thing” down, but there is one other area that I would come across periodically that would have me guessing a little more then I’d like to admit. Styles of dynamic compression, as well as output amps. So let’s quickly make that area a little less mysterious.
Feed Forward- If you take a look at the below diagram it pretty much identifies how the feed forward design differs from the Feed Backward, but let’s be more explicit. The Feed Forward design (as most modern compressors are) splits the input and sends one signal directly to the output amp while the other signal is sent to be processed.
Feed Backward- The Feed Backward design as shown in the right section of the below diagram goes directly to the output amp and then part of that signal from the amp is sent to the processing unit and that is combined after processing with the original amped non processed signal. So, you can see the difference lies in where the signal hits the output amp and where it travels from there.
(Compressor Feed Design)
VCA (Voltage Controlled Amplifier) - Ahhh! I feel at home now. At home with the old VCA! This is my favorite type of compressor (SSL G-Series, API 2500, and Neve 33609). This compressor determines the compression applied to the signal based on the input gain. As you adjust the input gain you can overdrive the compressor to what I call “sweetness” which is just a pleasurable way to describe VCA distortion. These compressors are often mix bus compressors, but I’ve tried them on a variety of sources.
Optical- I’ve often found optical compressors perfect for vocals because of their smooth gradual gain reduction. This lends to a more transparent sound. The photo optical detector circuit is triggered by the input signal, and gain is applied according to the stimulation of the optical circuit.
FET (Field Electric Transistor) - One of my other favorite compressor types simply because I have enjoyed the sound of the 1176 and its emulations. While older FET designs were known for their high signal to noise ratio the more modern revisions have brought the SN ratio under control. The gain reduction in FET’s are controlled by this transistor and are known to be faster acting and often significantly more accurate than say an optical compressor that actually varies with the use of the optical circuit’s wear over time.
Valve- You might have heard of “variable-mu” a time or two in the compression space, well these compressors are tube based and often give a nice harmonic distortion. This variable-mu/gain design is famous in the Manley and Fairchild 670 lines. Like all tube based gear these aren’t cheap. The compression is based on a re-biasing of the tubes which has to deal with clearing electrons across the electric grid of the tube if I’m not mistaken, but I am not an expert here, so if you are interested in knowing more about how valve tube compression works, I invite you to get back to us in the comment section with your findings.
Peak vs. RMS- You may see this option in your compressor so I think it would be helpful to know the difference. Peak reduction is obviously just using your compressor as you usually would, once the signal rises above the threshold the peak is reduced based on the ratio. RMS (Root Mean Squared) gain reduction is in laymen’s terms “the average”. It is gain reduction based on the average signal level compared to gain reduction based on the peak signal level. Often RMS gain reduction might be used on the mix bus with a low ratio to raise the overall level of the audio, versus trying to contain signal peaking.
Class A- The most widely used output amp style. The Class A output amp boosts both the peaks and valleys of the output signal and is thereby more efficient in its amplification. Class A’s tend to be simpler than both subsequent designs. According to Wikipedia, Class A tends to perform better in the high frequency range as well
Class B- This style of output amp boosts one side of the output signal, either the peak or the valley but not both, this can lead to distortion, but at the same time the amplification can be greater and more efficient.
Class AB- The Class AB design boosts the output signal separately like the Class B but boosts both sides like the Class A, this can lead to less distortion than the Class B and slightly more efficiency in amplification than the Class A. More about all of these output amps are listed here for your edification pleasure.
Aside from touching on particular compressors and particular usage situations, I think we have definitely composed a comprehensive understanding on what to expect when approaching compression in your mixing environment. We know why we use compression, what controls are at our disposal in addition to the various types of compressors we might come across. Compression is a never ending subject of discussion when it comes to pro audio, so if you look around the web or elsewhere you’ll notice discussions galore on what are the best compressors for each situation and what are the best ways to record i.e. recording to tape with compression or dry. I have my preferences, but before any of us can improvise or become creative with the tools, it’s important that we understand how to use them on a foundational level and I hope we’ve furthered that cause here. As always:
“If it sounds good, it is good.”
**If you have any comments on this or any other article, catch us in the forum and let’s discuss! **
Sources referenced : Sound On Sound – The SOS Guide to Mix Compression, Dynamic Range Compression, The Stereo Bus.com, McDSP, Practical Music Production, Class A-wiki.