|
Questions about EQ & Compression
|
View this Thread in Original format
| ChemTec |
Hey i was just wondering if somebody could explain to me what exactly you are looking for when you use either a compressor or eq? I know how a compressor works and what each control does but what i don't understand is where exactly do you use them other than as a sidechain trigger? What are you trying to accomplish when using a compressor?
I know that it can be used for reducing peaks and balancing out sounds but how do you know when you should use it and when you have either over or under compressed something? How do you listen to something and say for definite "I need to compress that"
My question then about eq is similar. When eqing lets say a mid range bassline to make room for a sub bass. Would you low cut the mid range bass until you loose all the sub frequencies or would you just use a generic setting such as cut everything below 100Hz? The reason i ask is if you eq a mid type bassline when you eq the sound to suit the sub bass you loose a lot of the character of that particular bass when you eq it. Some people would say thats when you should layer your basses but then that is adding more muddiness and frequency masking.
I know these sound like complete "Amateurish" questions but after producing now for 5 years i'm still wondering when you would use a compressor. I find myself using it most of the time just because it is there in my effects library and it is easy to drag in.
So in a nutshell my questions are:
1. When should you use compression & How do you know you have used it properly.
2. How do you know when you have cut the frequencies you don't want or eq something correctly
Any help would be appreciated, Thanks in advance :) :) |
|
|
| evo8 |
I remember years ago doing the exact same thing with compressors, id have one on every channel, compressing away and didnt really understand why i was using them!
These days the only compression i really use (apart from the odd sidechaining) is on the mix bus, as it can tighten up the feel of the mix (between kick and bass usually) - and its something you will definitely notice (and for me sounds much better than without)
Try compression say on your drum bus, it can "gel" your percussion together nicer - but at the end of the day if you cant hear it sounding better or cant hear the compression at all then its better not to use it
rarely i might use a compressor on a snare for example, to bring out the attack part of the snare, it also depends what compressor you are using for what task - i use the UAD 4k a lot as its a nice punchy compressor
I think with softsynths and synths in general, i find i dont use compression on these as its easier to get a particular sound with envelopes
EQ obviously is a lot easier to hear/notice - basses (not sub) i usually CUT at about 150hz to 200hz depending on the bass of course - you need a decent monitoring system (monitors and room) to eq effectively in this region IMO
When to eq??? If you feel something is taking up too much room in the mix, or not standing out enough - i would use quite severe dips (even tho most would say you shouldnt) on pads and lead synths for example, but you have to be careful not to overdo it and kill the character of the sound altogether
So in a nutshell to answer - you know youve used compression/eq properly when it sounds good ;) hope that helps |
|
|
| derail |
| quote: | Originally posted by ChemTec
1. When should you use compression & How do you know you have used it properly.
2. How do you know when you have cut the frequencies you don't want or eq something correctly
|
1) Firstly, be absolutely clear about what a compressor does. A compressor turns down the level when the incoming signal's level exceeds the threshold you've set. In absolute basic terms, a compressor turns down the level, and that's it. Other people may say "it glues things together", "it thickens sounds magically" or a hundred other ways of revealing their confusion.
Having said that, a typical compressor has a number of controls. So here we get from WHAT does it do (turns the volume down) to HOW does it do this:
Threshold: at what incoming volume level should the compressor turn the volume down?
Attack: how quickly should it turn the volume down when the threshold is exceeded?
Release: how quickly should the volume return to it's original level when the incoming signal drops back below the threshold?
Ratio: how much should the compressor turn the volume down when the threshold is exceeded? Only a little bit (for example 1.10:1), meaning the signal can still rise well over the threshold, it'll just be pulled back a fraction when it exceeds the threshold? Or a lot (for example 15:1), where the signal will only rise a little bit above the threshold, since any additional input volume increase will only translate into 1/15th the output volume increase.
Other compressors have a bunch of other controls. And yes, here is where you can get really creative with an effect which "only" turns the volume down - if you set the threshold very low, so the compressor's almost always working on the signal, and the ratio high, the dynamic range of the signal will be massively reduced - the quiet parts will stay quiet, the loud parts will be turned way down so they're not much louder than the quiet parts. Then you can turn the output level of the compressor right up so that the entire signal is consistently loud. On the other hand, you can leave the output level low so the entire signal is consistently quiet - it really depends where you want that part to sit in the mix.
Be clear on what a compressor does, and be clear on what you need/ want to do to each particular sound in your mix. If you have a sound which isn't as "solid"/ consistent as you'd like, potentially reducing the dynamic range with a compressor can help.
In the book "mixing with your mind", the approach to the controls is ARRT - attack, release, ratio, threshold (note - this doesn't work for sidechain mode - this is where the compressor is directly reacting to, and acting on, the sound being compressed):
Set the threshold to a point where it's frequently working on the signal, but not always (this isn't going to be the final threshold value). This will allow you to hear what the attack and release controls are doing to your sound.
Then set the compressor's attack, and listen to the attack/ front parts of the sound. A fast attack will dull the sound - it can turn a satisfying kick transient into damp cardboard and make you wonder why your kick isn't "cutting through the mix" any more. If you're compressing sounds which don't have sharp attack transients (e.g. pads), there's not much harm in setting the attack to minimum. Hey, you can even turn your kicks into damp cardboard as a creative decision - it's your music after all.
Once you're happy with the attack transients of your sound, that they match what you intended to do before you plugged the compressor in, you can proceed to the release control. Here, there are a few more considerations - setting a fast release can mean the volume "rushes up", drawing attention to itself - again, it's your creative decision. Setting it too slow means the compressor will cease to function effectively - by the time the volume should have returned to it's normal level, the incoming signal has already triggered the compressor again. In the "mixing with your mind" book, the approach is more "groove" based - set the release time so that it adds to the groove/ feel/ rhythm of the music - set the release time, then nod your head to the music, feel the groove. Set the release time a bit shorter, or a bit longer, and repeat the nodding/ grooving process. With practice, you should get a feel where the release time most naturally gels within the overall groove/ flow of the song.
Then there's the ratio control - this depends on how much of the life you want to squeeze out of the sound - for something which is already very consistent, and which doesn't need a large dynamic range - say, the main bass foundation - go nuts with a high ratio, make that bass solid and consistent and ensure it never drifts/ wimps out even for a moment. But for an expressive female vocal, or a beautiful melodic lead sound, maybe you want to build from soft, tranquil parts to loud crescendos. Again, be clear about what you want to do to the sound before you plug in a compressor.
Finally, there's the threshold. Currently, it's sitting at a level where it's being triggered frequently - perhaps you want to compress along the entire dynamic range, or maybe you just want to stomp on a couple of instances where the vocalist started screaming, whereas the rest of their performance was reasonably soft and doesn't need adjusting. The threshold and ratio are intertwined - adjustments to one may well require adjustments to the other. If you have a very light/ low ratio which works for a certain threshold, raising the threshold so you only treat the very loudest sounds may require you to raise the ratio so it still lowers the volume of those loudest sounds in a meaningful way.
2) EQ, as with compression, can be used to fix issues, or to be highly creative. Once again, you need to determine what you want to do, be aware of what your tools do, then select the appropriate tool and the appropriate settings.
Parametric EQ can be used for surgical cuts on resonant frequencies - these are frequencies for particular sounds/ instruments which resonate in an unpleasant manner. These will show up on a frequency analyser, and will likely hit your ear as unpleasant (re the frequency analyser, don't confuse a resonant frequency with the dominant frequency - if you're focussing on the intro of the song, and the bassline is sticking to the C note, then obviously you'll see spikes at 55/110/220/440 - it's easier to determine a resonant frequency within the context of a chord progression). With these, you can set a vey tight "Q", a decent amount of cut (from -3dB to 8dB or more - it totally depends on what is needed) and the resonant frequency will disappear, without altering the character of the sound noticeably. You'll just notice that the "pressure" has been releived, and the mix sounds cleaner.
Cuts/ filters can also be applied to remove the parts of the sound which aren't integral to the sound's character within your mix. Yes, a mid bass sound may have nice low end, but does it need that nice low end within your mix? With these broader cuts (whether low pass filters, high pass filters or broad parametric dips), listen to the mix overall as you're determining the position, width and depth of the cut. It's easy to make a sound too dull/ thin, or to cut too much valuable midrange out (side note - if you want a loud mix, the midrange is where it's at).
Boosts can also be applied as needed. Some people may say "never boost", but that's not exactly helpful. There aren't many EQs around which don't allow you to boost. When boosting natural sounds (or applying any sort of EQ, cut or boost) such as vocals, or an acoustic guitar, or anything where people may come across it in their day to day lives, it'll be noticeable when you've changed the frequency range of that sound. For unnatural sounds such as synths, there's no expectation in the listener's mind, so they're not going to pick up on a decent boost, if the final sound sounds good within the overall mix.
In terms of knowing whether you've succeeded, mix-wise, my advice is to reference against your favourite mixes. Constantly. You know what a good kick drum sounds like, or a good bass, or a good lead. Open up your music collection and listen, and there'll be plenty of examples of what you would call good sounds. Whether you do it consciously or subconsciously, the music you've listened to in your life dictates what you define as "good sounds", dictates the sounds you choose to create/ put together. It's just much faster to do it consciously, reference your mix against a bunch of great mixes in your style, then go from there in terms of making it your own.
TLDR? |
|
|
| wayfinder |
| quote: | Originally posted by derail
TLDR? |
Not at all, good info for beginners and those who think they're not beginners any more alike. |
|
|
| ChemTec |
| Cheers for the answers lads, really do appreciate it :) |
|
|
| Evolve140 |
I use a compressor to make the sound of something come out more, so that the volume is more consistent. For instance, if a synth sound has a really nice attack and decay, but the volume tends to drop significantly (within relativity) at any point when a note is played, a compressor will keep the entire volume more consistent so each phase of the envelope, and indeed the whole sound, is closer to a unified volume level. The envelope will still be the same, sound the same, but the volume of what is being compressed will be altered. There is a formula. Once you set the "threshold", this is how high the signal must be in dB for the compressor to kick in. If your signal never peaks above -10dB, but your threshold is -5dB, then the compressor will never "kick in". Theoretically, if you were raise the threshold to -11dB, then it will activate the other parameters of the compressor. That will involve your ratio. This determines how much compression will take place after the threshold has been reached. It may takes years to fully grasp how it works and to be honest you can read about what they do all day (heaps of information about this subject is available online) but practice and deployment are really the only way to do it (learn how to use a compressor). I use compression on nearly everything, but since it inherently squashes dynamic range, it must be used aesthetically and pragmatically.
I typically will always do the compression stage after EQing, although there isn't necessarily a rule for it. For EQ, the easiest analogy I can think of is a 3 piece rock band with a singer. You have the drums, bass and guitar. The guitar fills up the high and middle harmonics, and the bass fills the low and low-mid harmonics. You don't ever see a band with like 3 guitars and 2 bass guitars, because it isn't practical and would sound terrible. You want to fill out the frequency spectrum. This is tricky because if you are too surgical it will sound cold and not pleasing to the human ear. If you aren't surgical enough, frequencies will begin to clash creating "mud". You want a balance of frequencies in a harmonically sensible and pleasing way. So, each instrument must be thought of to occupy its own frequency range, but frequencies are allowed to spill and bleed over into the range of other instruments so that they create a harmonic presence. Too much is mud, too little is surgical and unappealing to the ear.
Derail has elegantly described the compressor better than I, but hopefully this helps too.
edit: One final note on EQ. Cutting is generally the rule, since if you are going around boosting heaps of things, you may be choosing the wrong sounds. If you choose sounds correctly, they will already have that presence in certain frequency range. So if you are constantly going around boosting frequencies, you may be in a situation where you are choosing the wrong sounds. Great sound selection and EQ cuts will make your mix solid, not the other way around. |
|
|
| Rodri Santos |
extense responses. My take:
I only use compression on the kick,bass and sometimes i send to a single buss all the instruments and apply some light compression to them. For percussion i only add it if i want it to have a main appearance on the track. |
|
|
| Looney4Clooney |
the concept is easy enough as people have stated. Compression is automated volume control at its most basic level. Things are more complex when the speed of how it changes the volume can affect the actual frequency of a source.
The implementation is usually aesthetic or technical in nature which i would say there are 3 types. Creating space, levelling and glueing elements together . Recordings have always had a limited dynamic range and rather sparse set of sources that to recreate real sound that could work without any compression in the real world impossible for a recording. I won't go into the history of it all but basically, compression is used in a way that takes into account how we perceive sound and exploits those perceptions to get away with more than what you normally could if you didn't.
For example. A percussion might naturally have a longer tail, live snare drums lets say. Our brains notice the transient and the tail if it is solod but when you have it in a mix, you can scrap alot of the tail and still not change the perceived sound of the snare. So a long attack and short release, will make your snare take less room and do what it was doing just as well.
There are also the sort of levelling type compression which is sort of redundant now with automation. For example, an arrangement might have an element when the mix is sparse. Once the vocals come in, there is no need to keep say a hihat as loud as it once was. The brain cannot pay attention to everything, in fact the brain can only really pay attention to one thing at a time in detail, 2 things in a somewhat computer like cpu cycling but your brain can't really process more. Mine can but that is besides the point.
So as your point of focus changes, you lower certain elements. They will still be there in your mind and might notice if you muted them but they are barely audible. The same technique is used for post production sound editing in film. When you establish say a jungle scene, you will notice the sound fx are loud at the beginning but once you believe you are in the jungle , you can lower them drastically and with the dialogue, they are probably not noticeable but you still feel like they are.
The aesthetic part is more subjective. I like the sound of a really compressed drum group with quick release that gives you that pumping effect overlaid over a version that is less or not compressed.
I think if you just think about why you are doing it, you will just come to understand it and why you would use it. Everyone is guilty of using things for no reason. Once you feel you have a grasp of the tool and the controls, start to approach your mix in a methodical way and have a list of things you need to fix and what you can do to fix it.
Creating a list will force you to listen and that is really the hardest part. Knowing what is wrong. This takes years of ear training. Even longer if you just mix without doing any actual drills that are focused to train your ear on noticing small db changes, eq changes ..... |
|
|
| Beatflux |
| quote: | Originally posted by Looney4Clooney
So as your point of focus changes, you lower certain elements. They will still be there in your mind and might notice if you muted them but they are barely audible. The same technique is used for post production sound editing in film. When you establish say a jungle scene, you will notice the sound fx are loud at the beginning but once you believe you are in the jungle , you can lower them drastically and with the dialogue, they are probably not noticeable but you still feel like they are.
|
Maybe if you could go a little bit more in depth in how this applies to mixing, that would be cool. |
|
|
| Looney4Clooney |
| you don't need to hear the entire snare for your brain to fill in the details. You can EQ something once other things come in and the brain won't notice that you've made accommodations. The point is to do only as much as you can without people noticing something is sounds weird. Compression is not needed for acoustic performances. But as soon as you start reducing your space either recording or going into a PA, you need to find room and this is why dynamic manipulation is used. I think its better to think of it as that rather than compression as compression is not the whole story. |
|
|
| wayfinder |
This thread has a lot of great points, the only thing I'd like to add (which some people have touched on but which I think hasn't been made explicit enough) is this:
Compression as a mixing tool is something you do to stuff over which you have no deeper control (deeper in this context means earlier in the chain), or where you want to affect many things at once.
When recording a guitar or a vocalist, the universe probably won't allow you to set the singer's sibilant loudness or the guitar's release, or make people perform at a more or less constant level of volume. A compressor after the fact will.
For a synth part, that's often not necessary, though: unless the synth itself restricts your options, you can probably adjust the volume of the noise oscillator that produces that unpleasant sharp click at the beginning of the sound, or decrease the release value in the amp envelope. This is advantageous because it will have no side-effects on other aspects of the sound, whereas compression that served the same purpose may. |
|
|
| Teezdalien |
There's some good info here, you need to be careful about how you order them in the chain which isn't always going to be the same.
You can end up trapped in a never ending cycle of EQing, then compressing then EQing and compressing and so on due to how each process affects the sound. |
|
|
|
|
