Rizzle's Guide To Compression And Dynamic Processing

Preach

Well-Known Member
#1
some opening words
there are no stickies about, and quite frankly i can't recall ever seeing any threads about compression in this forum. i thought i'd write up a quick starter's guide to compression. i put "dynamics processing" in the title but we're gonna focus on just compression, which falls in as a subsidiary of the "dynamic processing" subject. the art of compression is the most misunderstood, underrated, most neglected step of the mixing and mastering process, in the history of unprofessional music-making. for the most part, the reason why "underground" producers tend to "forget" to use compression is that they simply don't know what a compressor does. i'm gonna start off by explaining some of the terms you'll encounter if you decide to give it a shot. i have readied two images that i will use to demonstrate, both show a wave and how a compressor affects it. if you don't know what a wave is, you have no business reading this guide in the first place.





-------------------


clipping
when a sound signal is amplified to exceed 0dB, your signal is clipped off, which results in a distorted wave. a good way to see what a majorly distorted digital wave sounds like is loading an audio file up in a wave editor like cool edit, increase the volume by 5-10dB and then decrease it by the same amount and play it.

threshold
the threshold determines how loud a signal must be for the compressor to compress it. if you set the threshold level on your compressor to -10dB, the compressor will apply compression to any and every sound in your wave that exceeds the threshold. that is, any sound louder than -10dB is compressed. digital audio levels are often misunderstood. when you work with digital audio, 0dB is the absolute volume, nothing can exceed 0dB. it's like 100%.

ratio
the ratio determines how much compression should be applied to any signal that exceeds the threshold level. a 2:1 ratio means the output signal is reduced to half the input signal.

attack/release
attack and release (or decay) works just like the envelope controls on a synth/sampler. it determines how the compressor will act when a signal exceeds the threshold you've set. a slow attack will allow for some of the signal that exceeds the threshold level to pass through uncompressed before the compressor kicks in. a slow release means the compressor will keep compressing for a while (normally about 0-50ms) after the offending signal has been compressed.

gain
basically a volume controller. the input gain lets you boost the signal before it's compressed. practically, that means your wave will undergo a harder compression as boosting the volume sub-sequently pushes a larger part of the wave above the threshold. the output gain lets you boost the volume of the compressed signal.

soft/hard knee
compression will alter the sound of a wave, and thus also the way it looks in a wave editor. if you look at image number two above, you'll see that my kickdrum wave has a very soft, round shape. basically, this attribute determines how "soft" your compressed signal is gonna sound/look. it actually makes a big difference.

limiter/gate
this has got nothing to do with this guide, but since i'm already covering compression... a compressor falls in under "dynamic processors." why? because it alters the dynamic range of your mix. a limiter and a noise gate are two other devices that also fall in under "dynamic processors." a noise gate is like a reversed compressor. you set a threshold level, but instead of compressing everything that exceeds the threshold, a noise gate eliminates anything that's below the threshold. if you're recording live music with a band playing all at the same time for instance, letting the vocal mic go through a noise gate will eliminate all the background noises from the other instruments that are being played as the vocalist sings. a limiter uses mild distortion to prevent clipping. once a signal exceeds 0dB, the signal is distorted. the limiter uses a gentlier kind of distortion to cut the signal before it starts to clip. in most cases, the distortion caused by the limiter is barely noticable. this, of course, depends on how drastic your settings are. another dynamic processor is a finalizer. finalizers are kind of like a mastering suite preset. it adds a little reverb, and boosts the highs and lows a little. technically useless. most professionals frown upon them even. i don't think i've ever added reverb to an entire track in my life.


-------------------


rizzle speaks on compression
so what is compression? when you compress, say, a mix, you lower the volume of sounds that stick out of your mix like a sore thumb, so that you can bring up the overall volume of your mix to make it sound more like a professional commercial album. boosting an uncompressed signal would make parts of the wave (most noticably the kick drum and snare) exceed 0dB, and thus distort the entire wave. additionally, compression adds thump. i don't think i've ever made a set of drums without adding compression to the snare and kick. a long attack which lets the climax of your kick and snare go through uncompressed makes for a very thumpy, snappy sound. i vary my release, but i always use a long attack on my drums, regardless.

compression is not only recommended, but neccesary, if you want the bottom end of your mix to sound good. you may have noticed that whenever you make a beat, some notes in your bassline stick out because they are way louder than the rest. you may have noticed that some times when you're playing your beat in your car or on your computer's speaker setup, that third note on your triangular bass makes your subwoofer go BOOOOOOM!!!!! (just a funny observation; i love the wu-tang clan but a lot of their old tracks have basslines like this. it ruins the track when you bump it on a bassy stereo system) well this is where you use compression to even it out. however, i must warn you - do NOT apply compression to individual instruments in your mix, nor the entire mix, unless you have a good set of monitors in front of you. headphones, in my personal experience, as well as speakers that lack a good bass element, are useless for mixing and mastering in that compression alters the bass a lot, and you won't be able to tell what it actually sounds like. you need to be able to pick up the slightest change in the bass sound in order to mix it right. i don't have top notch studio monitors but mine are "O.K." for now. compressors have a tendency to make basslines sound very growly. they also have a tendency to kill the bottom end. it's like cutting everything between 0-400hz and boosting everything between 400-1,5khz. it sounds like shit. as long as you have studio monitors or a subwoofer you should be okay. it's an ongoing debate whether one should equalize before or after one applies compression. personally, i prefer to start off with compression, then set the stereo image, then equalize, and then amplify and limit it at the end there. if you equalize your bassline before compressing it, boosting 5dB at 150hz, it's still likely the compressor is gonna rob your bass of it's actual lower frequencies. you can equalize your bass, mix, or whatever before you compress it, but it's likely you'll have to re-equalize it afterwards anyway seeing as how the compressor's gonna change how it (the bassline, the mix, the whatever) sounds entirely.

i use compression both when i mix and when i master, although it's a mastering tool really. when i mix i only compress individual instruments, but when i master i apply very mild compression to the entire mix. when you compress an entire mix, you have to be very careful when setting the threshold level. if you set the threshold lower than the kickdrum's peak, the compressor will compress the entire mix whenever the kick drum hits. in the end it'll sound like someone's sliding the mix's main volume fader up and down like a raging madman.

like everything else that's related to making music, the key, of course, is experimenting. hopefully, now you'll have a little bit more clue as to what it is, exactly, that you're doing. i've merely scraped the surface, but if you actually call yourself "a producer" and have ambition that one day you're gonna sell your beats, you owe it to yourself to read up on, and learn how to use a compressor, because your shit will never sound clean, punchy and professional without it. in fact, you don't need a whole bunch of ambition. whether you produce because it's your passion in life, because it's what you want to do when you grow up, or because you're very bored on afternoons, you should check out and read up on compression anyway cause it's gonna improve your stuff.


-------------------


oh, and if i misunderstood something and am wrong, feel free to correct me ;x
 

PuffnScruff

Well-Known Member
#2
ive been learning about all of this in great detail over the past few weeks in my sound foundations class. i just got out of my mixing lab a few hours ago

good post. your right, im sure alot of people wouldnt have the slightest clue to any of this.
 

Preach

Well-Known Member
#3
PuffnScruff said:
ive been learning about all of this in great detail over the past few weeks in my sound foundations class. i just got out of my mixing lab a few hours ago

good post. your right, im sure alot of people wouldnt have the slightest clue to any of this.
that's awesome that you actually got a sound foundations class. if you don't mind i'd like to hit you up and talk a little bit about shit. i'm assuming you're in uni. i'm thinking about studying something audio production-related and i'm thinking you may be able to answer some questions.

since i'm not really around these days we'll just have to talk sometime whenever.
 

PuffnScruff

Well-Known Member
#4
by all means go ahead, but i dont know a whole lot, i'm still in the learning stages of this audio engineering thing. most of it is still chinese to me.
 

Preach

Well-Known Member
#5
yeah, i was thinking more in terms of the actual study. like what classes there are, what you actually do in class, etc etc. i'll hit you up some time for sure.
 

Bigg Limn

Well-Known Member
Staff member
#6
Alright I got a compressor [actual hardware] and theres a few things on it that you didn't cover and I can't figure out what they are [we haven't started learning about compression yet in my classes], so hopefully U could help me.

Theres a section called "Expander/Gate" with 2 buttons that say "Release" and "Gate" - and a knob labled "Trigger." In the "Compressor/Limiter" section theres a few different buttons I dunno what they R - "SC Mon," "SC Ext," "LO Contour," "Interact Knee" [guessing its basically what U were talking about up there], "De-Esser" [I remember briefly going over this, but don't remember what it is], "I/O Meter"

Think thats it...thx 4 the help oh rizzly 1
 

PuffnScruff

Well-Known Member
#7
Bigg Limn said:
Alright I got a compressor [actual hardware] and theres a few things on it that you didn't cover and I can't figure out what they are [we haven't started learning about compression yet in my classes], so hopefully U could help me.

Theres a section called "Expander/Gate" with 2 buttons that say "Release" and "Gate" - and a knob labled "Trigger." In the "Compressor/Limiter" section theres a few different buttons I dunno what they R - "SC Mon," "SC Ext," "LO Contour," "Interact Knee" [guessing its basically what U were talking about up there], "De-Esser" [I remember briefly going over this, but don't remember what it is], "I/O Meter"

Think thats it...thx 4 the help oh rizzly 1
a de-esser is good to be used for a vocalist, it compresses the sibilant freq like the "s" sounds, if the singer sounds like they might have a lisp or slurs a little you would use a de-esser in that case.

the noise gate weakens the unwanted low level information that comes from the signal. i was told to think of it like a gate on a fence. like if you want to use a snare through the noise gate, the gate will open when it gets the information of the snare coming through it and then close when it passes.

the trigger could be either your threshold, range or attack time, im not sure with out looking at it.

the expander basically makes soft sounds softer and loud sounders louder, or the opposite of compression, it amplifies the signal above the threshold and weakens(attenuates) the signal below the threshold.

the release determins the time the gate will open. a long release time will cause a fading out effect and a short release time sounds like the signal has gone below the threshold. you'll need to use the range control to get these sounds too. it makes more sense when you play with it.

not sure about the stuff in the compressor/limiter with out looking at it

im going to take a wild guess and say the I/O meter just tells you the levels of your inputs and outputs.


where are you studying at limn? are you studying audio engineering too?
 

Preach

Well-Known Member
#8
puffnscruff basically covered it. my actual experience with hardware compressors is vague. it's there but it's vague. i've been thinking about getting a really high-tech compressor for the fuck of it lol
 

Bigg Limn

Well-Known Member
Staff member
#9
PuffnScruff said:
a de-esser is good to be used for a vocalist, it compresses the sibilant freq like the "s" sounds, if the singer sounds like they might have a lisp or slurs a little you would use a de-esser in that case.

the noise gate weakens the unwanted low level information that comes from the signal. i was told to think of it like a gate on a fence. like if you want to use a snare through the noise gate, the gate will open when it gets the information of the snare coming through it and then close when it passes.

the trigger could be either your threshold, range or attack time, im not sure with out looking at it.

the expander basically makes soft sounds softer and loud sounders louder, or the opposite of compression, it amplifies the signal above the threshold and weakens(attenuates) the signal below the threshold.

the release determins the time the gate will open. a long release time will cause a fading out effect and a short release time sounds like the signal has gone below the threshold. you'll need to use the range control to get these sounds too. it makes more sense when you play with it.

not sure about the stuff in the compressor/limiter with out looking at it

im going to take a wild guess and say the I/O meter just tells you the levels of your inputs and outputs.


where are you studying at limn? are you studying audio engineering too?
Good lookin out...yeh I thought thats what the de-esser was but I wasnt sure. Im jus studying at the local community college right now, but they have a Digital Audio Producer degree so Im studying for that right now - if I keep on with school Ima look in2 the SAE schools, or maybe something around here for an Audio Engineering degree.

Ill post up a pic of my compressor in a bit so U can C what Im talking about 4 some of these.

pz
 

PuffnScruff

Well-Known Member
#10
i've heard SAE is a good school. i was looking into going there before i decided to attend Full Sail. I checked out their nashville campus a few years ago. It wasnt bad.

There is a good audio engineering school in AZ but i cant remember the name.
 

Bigg Limn

Well-Known Member
Staff member
#11
yeah i was lookin at some nice ones here...but Id rather get the fuck out of here. And my digi camera is fux0rd right now so nvm on the pic.
 

Dante

Meyer & Dante Best Friends4eva
#12
rizzle asked me once how i have "brighter" vocals in my mixes than a lot of others that may use the same acapella. the answer: i have created several compression presets for myself and i run the vocals of a mix through it before the final mixdown.

what does this accomplish? well vocals are touch and go a lot of times, meaning there are volume inconsistencies from the performer trailing off, a poorly sensitive mic, poor mic placement, or just different takes. running vocals through a limiter will not only produce a much more compressed (less spikey) waveform, but it adds an element of vibrance into the signal. the end result is a rectangular looking wave that is not clipped, and loses a lot of spikes from hard sounds (puh sounds from the letter p for example). as needed i also de-es, which again i have created presets for myself based on generic male vocals, generic female vocals, and of course 2pac vocals. de-essers tend to dampen the signal a touch, so they should be used very sparingly. as with all recording, it's best to do it right the first time (the act of recording your vocals) rather than applying massive amounts of technology to fix things.

are there limits to the effectiveness of using compression? the answer is yes. if i were to take a chunk of audio and run it through a limiter or dynamic processor in a loop, the chunk would begin to sound more and more clippy, and eventually sound like a radio signal. you can only really alter the slopes and peaks of a waveform a couple times before audible consequences begin to surface. it is therefore noteworthy to consider the source of one's audio when mixing. here's why...

the best dj's and producers in the industry in the 80's and 90's (when sampling was intensive and at the heart of beat creation) would use different layers. stretch's production is a prime example. snares from stevie wonder, a melody from prince, and a bassline from zapp. ok, so we have our pieces now, do we mix then compress or do we compress each piece first? the answer is to master each piece first (master refers to more processes than just compression). why? quite simply, compression brings a spikey waveform into a managable rectangle while boosting overall mixability. quite frankly, your pieces will fit together with less masking and more accurate audible prediction. awesome, but here's the catch. when mixing a cd or a digitally processed chunk of audio, you're dealing with a segment that has already undergone compression and mastering. sure it's possible to mix vocals into an instrumental from cd, then compress, however it's also true that you will lose amounts of quality. mastering the vocals first helps the final master of your mixed tracks to be less processed, therefore minimizing the changes made to the already compressed instrumental. make sense?

dante's audio pet peeve....

you mother fuckers. learn how to use an eq. the layperson's concept of an equalizer is that it is to "boost" frequencies that they favor. i like bass, so crank that summa bitch up! wrong. unless you gain down your audio, any eqing usually results in either clipping or distortion. ever wonder why fan made 2pac "remasters" you download off the internet lack audio fidelity and generally either sound flat with too much bass or too much treble? the person who made it likely jacked up the eq to compensate for the effects of applying massive noise reduction to a track (i can speak for days on how to properly use noise and hiss reduction). don't jack "up" an eq - jack it down! studio eq's were designed as a subtraction tool, not an additive one (however this image was destroyed by the four little sliders on your average boombox). therefore, decrease gain instead of boosting. you will find that your highs and lows stay pure and you didn't go and muddy up another song.
 

Preach

Well-Known Member
#13
nice to see you shed some thoughts and facts on this. i started out making tupac remixes, but as a "producer" - yeah i feel like a moron when i say that but fuck you - my ambitions have changed. at first, i just made tupac remixes. right now i'd love to get to work with some artists. not because i'm sporting the cocky i'm-gonna-break-through-i-know-it-attitude, but because sitting alone at home mixing the same old tupac acapellas over your beats can only bring so much fun to the table before you get bored and want to kill yourself. your (dante's) thing, as a mixer, is the og vibes mixes and that's actually a dope concept and you actually create mixes that you yourself will keep bumping for a while. i don't play my own remixes a lot at all, and none of the people i hang out with are pac heads so if anything, my remixes are for my own pleasure. i guess what i'm trying to say is that since i've moved over to focusing on producing for people as opposed to making remixes, i haven't worked a lot with the mixing and mastering of vocals lately, and it's only lately that i really feel i've gotten into the technical aspects of production, mixing, and mastering. if anyone wants to add a few words about compressing vocals, what to do and what not to do, little tips and tricks, typical n00bie mistakes, whatever, please do.

Dante said:
rizzle asked me once how i have "brighter" vocals in my mixes than a lot of others that may use the same acapella.
that was a very long time ago ;p

Dante said:
studio eq's were designed as a subtraction tool, not an additive one (however this image was destroyed by the four little sliders on your average boombox).
:thumb:

although mine only has four.

Dante said:
(i can speak for days on how to properly use noise and hiss reduction)
in fact, if you wouldn't mind i'd love if you'd write up a thread about this. i've been looking into noise reduction and all the guides and faqs i've found were either too detailed or too vague. i know how to reduce noises using the built-in adobe audition noise reduction tools. i know that as a general rule of thumb you should never apply more than 30% noise reduction. here ends my knowledge on the subject. i'm gonna look into interacting with musicians and recording live instruments in a while so i know i sure would appreciate it.
 

Latest posts

Donate

Any donations will be used to help pay for the site costs, and anything donated above will be donated to C-Dub's son on behalf of this community.

Members online

No members online now.
Top