redmeansrecording
patreon
Patreon Exclusive: Intro to Compression
🕑 Added 2024-08-05 14:50:58 +0000 UTCVideo here: https://vimeo.com/994568696/02b5707ee3?share=copy
I am going to be doing, on average, one of these a month for ya'll. We'll do some educational video on something that I wouldn't normally do for the main channel. In this case, it's an intro to compression.
I have a script for this that went into more detail that I never made a video for, and I'll paste it below. Hope this helps!
WHO
https://musictechstudent.co.uk/music-technology/history-and-development-of-compression/
Who uses compressors?
Audio engineers, guitarists, radio engineers, radar engineers, telephony services, mastering engineers, television broadcast, and more. It’s a ubiquitous tool wherever audio is found. You should use them too.
Radio engineers use them to keep program levels consistent across music, vocals, and commercials. If they didn’t, they could get kicked off the air.
Broadcasts used to use early versions of them, called leveling amplifiers, to help the announcer’s voice rise over the audience.
Audio and mastering engineers use them for everything from subtle adjustments to dynamics of a track or mix, to absolute obliteration of an instrument. It’s on the top level of most useful tools alongside EQ when you’re mixing.
WHAT
Compression and limiting fall into the category of “dynamic range compression” tools. They reduce the volume of loud sounds and increase the volume of quiet sounds. Dynamic range is the difference between quietest and loudest volume, and when you smoosh them like a soft dog, they compress, hence the term “compression”. A limiter is a compressor with a high ratio, and generally a fast attack time to catch peaks. A “brick wall” limiter is a compressor with an extremely high or infinite ratio.
There are a few different types of compressors, all based on different analog circuitry:
https://www.prosoundweb.com/the-different-types-of-compressors-and-when-to-use-them/
https://producerhive.com/ask-the-hive/types-of-audio-compressors-when-to-use-them/
Optical Style (like the LA-2A)
A light bulb and a photocell were used as the main components of the compression circuit. The time lag between the bulb and the photocell gave it a distinctive attack and release time.
Slow attack and release, use when large transients aren’t present (like vocals). Transparent tightens up the track without being noticed. Adds warmth. Limitations: Won’t control transients, pumps with low-end content.
FET Style (like the 1176 Compressor)
A Field Effect Transistor (FET) was used to vary the gain in the old analog hardware version, which had a much quicker response than the optical circuit.
Very fast attack and release with lots of control. Aggressive sounding. Best for punch and snap but adds the most color. Very warm and rich sounding, but won’t work on the mix bus. Limitation: not transparent.
VCA Style (like the dbx 160 or SSL bus compressor)
A Voltage Controlled Amplifier (VCA) circuit was a product of ‘80s technology and had both excellent response time and much more control over the various compression parameters.
Aggressive sounding with extreme settings and extended parameter functions. Works on the peaks of the program. Use when the transients are already controlled (mix bus). Excellent for adding punch to drums and bus compression. Doesn’t smooth out the volume. Limitation: Doesn’t smooth out the volume. Can be thin sounding.
Vari-Gain Style (like the Fairchild 670 or Manley Vari-Mu)
The Vari-Gain compressors are sort of a catch-all category because there are other ways to achieve compression besides the first three.
Takes time to react, which adds a “glue” to the mix (like the mix bus or subgroup). Ratio increases with gain reduction. The louder the transient, the harder it’s compressed. Good for a mix bus or to add warmth and fatness. Limitation: Slow attack and release. Won’t solve dynamic issues or increase the punch.
HOW
A compressor has four major controls and a few variations on extra controls:
Threshold: this is the volume at which the compressor starts compressing
Ratio: this is the amount the compressor will reduce the volume after the threshold is passed
Attack: this is the speed at which the compressor compresses the sound. A fast attack only lets a small duration of sound pass through uncompressed, then compresses the sound. A slow attack will take more time to fully compress the volume of the sound.
Release: this is the speed it takes the compressor to stop compressing and return to zero gain reduction. A fast release means we go back to zero compression quickly. A long release means that depending on the threshold, ration, and source material, we may never return to zero gain reduction.
This is most obvious when using a technique called sidechain compression, in which an external signal is used to compress another. Most people know it in dance music when the kick drum compresses the bass or other musical elements. It makes a “pumping” sound, where the mix ducks out of the way every time the kick hits. In this scenario, the attack and release times become rhythmically important and obvious in relation to the tempo of the music.
Other useful controls on a compressor include:
Soft or hard knee: A soft knee makes the compressor gradually increase the compression to the ratio set by the user, making the compression more transparent sounding. A hard knee makes the transition from compressed to uncompressed much faster and harder.
RMS vs. Peak: Peak level compression deals with the peaks of audio, the meters you see in your DAW at their top level. Peak compression is great for fast transient material like drums, but it doesn’t take into account how humans really hear. RMS, or Root Means Square compression, uses the RMS loudness function applied to the signal before thinking about compression, creating a more relaxed compression that better mimics how humans hear audio.
Frequency dependant sidechain: some compressors in hardware and software feature a built-in frequency-dependent sidechain to filter out certain frequencies before the compressor looks at the signal. A great example of this is the Focusrite Compounder hardware compressor, which featured a circuit that let the bass bypass the compression circuit. Since bass frequencies have more power than higher frequencies, it meant the compressor wouldn’t hit its threshold based on these powerful low frequencies and the compression wouldn’t be centered around the bass, which is common when compressing full mixes. It means a cleaner, punchier full mix compression that preserves the basses. Trentemoller used to swear by it.
Make-up gain: This is on by default in the Ableton compressor. It will start to increase the output volume based on reducing the threshold and the compression the processor does. Since compressors are designed to reduce overall dynamic range, one of the most common uses is to boost the overall gain of a signal. Sometimes this is good, sometimes it isn’t. It’s really difficult to hear what compression is doing when you pair it with a big jump in perceived volume. Always check to make sure you’re not destroying your sound with too much compression.
Look-ahead: Available in the Ableton limiter amongst other places. Look-ahead splits the signal into a delayed and non-delayed copy. The non-delayed signal is used to drive the compressor or limiting of the delayed signal, giving you the ability to use smoother attack values to catch transients, at the expense of a delayed signal.
Wet vs. Dry: aka parallel compression. It’s the ability to mix wet vs. dry signals on a compressor means you can mix different styles of the same signal. The most used example of this is probably New York Style Compression, where a drum signal that’s highly compressed and eq’d is mixed with a non-compressed version. You get the sharp transient punch of the uncompressed signal mixed with a colorful compressed signal, increasing perceived loudness without increasing transient spikes to match.
WHERE
Compressors, like EQs, have become a ubiquitous inclusion with modern DAWs. You’ll find at least one flavor of compressor bundled with your software. Ableton has four: a traditional compressor, a “Glue” compressor, a brickwall limiter, and a multiband compressor.
Because there are so many types of compressors, many software companies have created their own versions. Some of these are meant to emulate classic compressors, limiters, and leveling amplifiers of history. Some notables free ones include:
Analog Obsession FETISH
ADHD Levelling Tool
Tokyo Dawn Labs TDR Kotelnikov
Klanghelm MJUC jr.
Xfer OTT
Whatever you have in your DAW
WHY
Why compress?
How much time do you have?
There are a ton of situations in which compression can help your mix, and almost as many on how it can mess it up.
As a tool, as opposed to an effect, compression can be used to level out dynamic performances like vocals, so that they sit in a pocket throughout the mix. This means that even if the singer gets extra loud or extra quiet, they’ll still occupy relatively the same loudness level in the mix.
I use compression on every vocal I work with. Sometimes multiple layers. An initial relatively fast and hard compressor to balance out quick changes in the voice, and then a slower, more musical compressor to transparently even out longer changes.
I’ll still need to automate gain if the mix gets bigger, like in the chorus, but a nice, right compressed vocal is always something I’m going to go for.
If a performance is consistent in volume, you may not need to use a compressor to even out the performance. You may instead reach for a compressor to shape the sound, especially the transients and sustain.
Drums are a perfect candidate for this. Acoustic drums have a sharp spike, called a transient, followed by some form of resonating decay. How the decay sounds like hugely variable, influenced by everything from drum size, skin type, astrological sign, and microphone type and placement.
Compressors let you reach in and shape the attack and decay of drums, getting a consistent punchy transient or a longer, beefier sustain.
Limiters are compressors with a high ratio and quick attack. These are good for shaving off peaks of audio, either as a utility, or as an effect. Sometimes I’ll use a limiter on a particularly spikey synth sound to take the peak so it can sit up higher in the mix.
Generally I EQ before compression, to filter out the frequencies I don’t want from the signal so the compressor threshold doesn’t key on frequencies that I don’t want to hear. Remember that any sound in any frequency band contributes to the overall db or RMS of a signal. I prefer to get that balance right before compression, but sometimes I’ll have a series of compressors and eqs all doing small things in a chain.
Reverb before compression can be a really wonderful effect. Because compression will kick in on the loudest parts and raise the volume of quiet parts, a heavy compressor following a reverb will create a massive sucking experience that no-one will be able to resist.
The same goes for bringing out the background noise of any signal. Heavy compression reduces the differences between loud and soft sounds, so if you’re not cleaning your signal with EQ, gate, or noise reduction beforehand, you’ll hear more noise.
Sidechain compression is something you should know how to do, but not overuse. When we use part of our drums to trigger a compressor, we create a dynamic rhythm from the compression we impose on other elements. If you sidechain everything, you will lose any sense of sustain in your parts. Sometimes it’s best to leave pads, keys, strings, or other sustained parts free of sidechain compression so they float over a mix and create the texture they were meant to.
When sidechaining, these are the settings I use. I pay special attention to the release, because that’s the control that creates the rhythmic relationship with the tempo. Too fast, it won’t pump. Too slow and it will never return to 0 and just be mush. I always try to make sure my gain reduction is returning to 0.
Multiband compression is a very powerful tool that I use in every mix and master. It allows me to shape dynamics based on different sections of the frequency. I can, for instance, shave off the transients of the 3-6k range, or apply compression to the mids for a guitar solo. I can also use this tool to listen to my individual bands and hear if there’s any particular buildup of frequencies or instruments fighting each other for dominance. UwU.
Because music is sound waves changing over time in volume and frequency, too much compression will rob a track of its dynamics and of its life. If you’re going to use compression in a way to help your mix, use it on individual instruments first, then busses for those instruments for glue. Be careful especially when using it during mastering, because too much compression will cause the louder elements to overtake the quieter ones and the whole mix will be spoiled.
Compression may seem like a complex tool, but if you take time to understand the individual components of how it works, it will become much easier to use in your mixes. Do yourself a favor and do some research on your favorite artists, albums, or songs when it comes to compression, especially older material and how they used compression as a creative tool. There’s a whole universe of fun ideas out there, and all you have to do is be curious.
Squish that cat.
Comments
Jeremy Blake
I'm really glad it helped!!
spacerwoman
Thanks so much for this! I knew some bits of this but this was a better explanation in practice than I'd seen before. Particularly appreciate you explaining just what the hell that "EQ" option in Ableton's Compressor does (I had been confused and thought it was supposed to be some pseudo-multiband thing; my ear didn't pick up that it was affecting what triggers it and not doing some filtering on what actually gets compressed).
Lera
Useful!
John McNeil
This was so helpful! Thanks!
monotonehell
J: "Expansion is the opposite of compression." Woah. Slow down with all the technobabble.
Josh Dick
This was great! As a mixing novice who is scared to incorrectly use compression, this was informative. I’d love a followup at some point in the future with an example showing how to use compression to “de-mud” a multitrack mix/using an example song.
David Cocciante
This is really a good refresh on what compression is and what it can do to sound. Also you are very good at explaining things. I'll be looking out for these videos.
Lonnon Foster
So I *thought* I understood compression, but you corrected some of my bad assumptions and straight-up wrong ideas with this video. And filled in some gaps where I was taking something on faith, like "don't compress the master track," which I knew resulted in crap most of the time I'd tried it, but I didn't know *why*. It's also great to have some more understanding of what the different controls of a compressor do. Sure, I can wiggle them around and come out with something that sounds better, but knowing ahead of time what a particular knob's effect will be saves me a lot of time. Some of this stuff I'd already developed an intuition for, but making it concrete really helps me reach for the right tool and spend less time flailing. Super useful, information-dense presentation. Thank you! I'm looking forward to applying what I've learned!
Braden Liatris
This is fantastic stuff, Jeremy! Thank you for making it!
Mario Godoy
This was great. Most of the info was review for me, but still learned some invaluable things. I’d love to see a part two that digs into parallel compression and its use cases. I’d also love to see how you’d use compression on groups. Lastly, I’ve always been curious what people mean by “use compression to glue your mix”. Great work! Thank you!
