*IMPORTANT* I can tell already that many are confusing the intentions of this video and I can clarify right now that it's not a video telling you what to do. The eq after compression "rule" is not my thing, it's just something I've read over the years and seen on various mixing videos and seen discussed on various forums and it's more focused on eq'ing heavily and tonally before compression not clean up eq, actual heavy lifts. Either way, this video is to specifically lead on from last week's video and the intentions of these videos is not to state WHAT you should be doing, but WHY others say you should, and WHY others do the opposite. Me personally, I do clean up eq before compression, and then my tonal eq after compression. I have my reasons and my own methodology behind doing what I do but in all honesty you shouldn't give 2 f*cks what I or others do. I'm making these styles of videos to try and give context to those who know WHAT many engineers do, but don't know WHY they do what they do. So they can turn experiment themselves and come up with their own little rules and methods that work better for their mixing approach. I'm just being crystal clear that I don't want to tell people what the "rules" are as they are different to every engineer and that why the Internet has created them. For every rule you hear, there is somebody else saying and doing the opposite. AND also, I decided to put audio examples for my own wee 8k/comp creation as an unlisted video just incase nobody cared for it in this video haha th-cam.com/video/dgtVUlcyaUg/w-d-xo.html
Yeah it's just playing the TH-cam game man. Unfortunately many skip audio examples longer than 30 seconds so it hurts the retention which hurts the video. It's shit I've got to do stuff like that but gotta give the algorithm what it wants
I do the same! Take out unwanted frequencies when in recording to the DAW with my Neve preamp that has EQ’s , after that I almost always go to my 1176 and later to the DAW then I add EQ in the box or on another analog EQ and gain up those nice frecuencies.
Ive been engineering over 30 years and have ALWAYS put EQ before my compressor. The only tonal shaping I do after the compressor is a de-esser on vocals. If you know how to use a compressor, you just make adjustment as you EQ. Seems in this day and age folks are getting too 'in the weeds' with the all technical stuff and not using their ears...turning it into a science project rather than art.
That's the whole point of being an engineer. It's technical. Audio is very much science based and the best engineers I speak to and admire have vast knowledge and understanding of the craft. It's not a one size fits all, it's a "when the source needs it" and knowing when to make certain moves at certain times. Thats all fine and dandy focusing on your art but mixing isn't just throwing stuff at the wall and being super creative all the time. You are a problem solver more than you are an artist. Your an engineer. Your job is to fix things and make it work. Sure many do repeat certain things but they know when somethings not working and the workaround for it
@@PaulThird I think his point is, that you don't need to understand what exactly happens. All you need to understand is: If I do this, this is the result. Now personally - for me - it makes it easier and faster to learn to understand SOME of the "technical stuff". Some people say a producer doesn't need to be an "engineer" - I disagree very much on this. I think the producer is closer to the artist (or is an artist in some cases) than the mixer and the mixer is closer to it than the mastering-engineer. But everyone .. really every kind of engineer and artist needs to know how to get a sound. You don't need to UNDERSTAND why it gets you the sound, but you DO need to know HOW to get it (if you want to be professional that is).
Tbh I agree more with you. Maybe it's the definition of engineer but in my head pushing through a mixing engineer I see it as my job to find out why professionals in the industry do what they do as essentially they are the competition. If they do something I personally wouldn't do then I couldn't care less as its their way of mixing and has nothing to do with me but I still want to undetstand why and how incase there's a time I need it. The biggest problem with "just use your ears" and "if it Sounds good then it sounds good" is that in the big world of mixing it's about being better than the other guy or girl. That's how you succeed. Just being "good" doesn't cut it. You've got to blow the socks off them and undertstand where your competition is and how they do what your mixes lack. The best engineers study other engineers. They learn the technical stuff. They obsorb like a sponge so they understand how to be better. There's one thing taking a trick from somebody that worked on one mix but another thing understanding why that same trick worked on his and not yours. If you make music and that's your main priority then focus on the music but if your a mixer working on other people's music then you better know your shit cause rest assured the other options do and if you run into problems then you better know how to fix them and have the tools to do that. The biggest mixing engineers on the planet are out and out audio geeks and that's where I struggle as my head is so fixed on mixing other peoples music instead of creating my own. I get it but the mixers I aspire to all know the technical aspects of audio. They don't know absolutely everything but they know what they need to know and that's actually the most important to learn in my eyes. Is the stuff you need to know. I don't need to learn the ins and outs of dsp and transfer curves etc I just learn what I need to use tools a certain way.
When I was learning I to followed the advise of EQ after compression. I then went to work along a very experienced producer/mix engineer and noticed how much compression was going on. Eqing before compression, multi bands on guitars, and synths. I mixed some stuff and he pointed out - you are not using enough compression. I learnt how working the compression right, using the right type and Eqing into it began to improve how I mixed, to the point where my mastering requires me to do a lot less. It’s similar to CLA, and as I watched CLA it fell into place for me. It’s very easy to get trapped into “methods” taught. It’s not till you sit alongside these guys you realise how actually there was a lot of bullshit floating around some years ago on how to mix. Don’t be afraid to push into things! Don’t be afraid to compress - it’s a sound some love!
Yeah just do what makes sense to you and let the rest do what they do. You've gotta be your own mixer in the end. Truth is I actually do both in the long run as I clean up then compress then tonally eq after that and then I may compress on a bus as well ie drums. On vocals its clean up, compress, tonally eq and then multiband compression at the very end. I don't always compress now, I used to but I'm a bit more chill now haha I try it out and if it doesn't make it better in the mix then it's off. I just don't give a f*ck anymore about what Davie 2 doors down is doing. If he's getting better results I'll pick it apart and undertstand what he does but still might not do exactly what he does. Always gotta make your own decisions cause its right for you
That is exactly why CLA is CLA....because he does things his way as long as they take him to whatever result he wants. He relies on his ears and years of experience. If it sounds good, it is good no matter what was done to obtain the result. Great video by the way!!!
I usually put it like so: EQ > Comp > EQ -1st Eq to cut off the frequencies I don't want to trigger the comp and other minor adjustments. The rest of the chain benefits from what you said. Everyone happy?
Great video! I'm so stocked on this video!! I was literally just thinking today about whether compressors (different compressor types) have different effects on the frequency range of the audio they're processing. Might be a cool idea for a video to go through different classic comp emulations with pugin doctor and test how/if they treat the freq range differently. Well done Paul, also love that your making your own plugins!
Yeah could be a decent idea 🤓 tbh before I make any more videos in this style I still need to try and get my head round why certain people call themselves engineers but at the same time moan about others trying to discuss the engineering behind what mixing engineers do haha swear to god I'm struggling with the "too much focus on technical aspects and audio science.. Its all about the music bla bla bla...".. I just can't compute shit like that.. Yeah music is important but that's for the recording engineer, artist and producers job to hash out and mixers job to engineer the record and do all the technical stuff to make it work. Yeah obviously still be creative but still fix it and make it work with the tools you have and fully understand. I've completely digressed but it baffles the life out of me how guys can call themselves engineers and not be audio geeks. I'm way too autistically literal for shit like that 🤣🤣
It's probably because for a lot of the main audience of youtube music production tutorials, all these jobs are things they have to do themselves, if you do this as a hobby, you don't exactly have the means to hire a mixing engineer, so you have to learn it yourself, even if you're not an audio nerd.
True but when your hit on youtube every way, left, right, up and down with.. DO THIS.. NO DO THIS... DO BOTH.. DONT DO EITHER, DO THIS INSTEAD It gets pretty convoluted and people end up doing stuff just cause somebody else said it works. That's why i firmly believe in diving deeper, bit by bit, and taking your time so you can make your own decisions. There's tons of stuff I know big engineers do or have done but I don't do as it doesn't work for how I work. I've been there in the past learning lots of things and chucking them into a blender of a mix. Lots of stuff that maybe worked in a certain context but not on that mix or all blended together. You get to a stage where you don't even watch certain videos cause your just not interested and eventually stop getting hung up on why you arent doing what they are doing. TH-cam and especially audio forums are an absolute mindf*ck for beginners as its all DO THIS and not enough LEARN HOW, WHY AND WHEN TO USE IT
@@PaulThird Yeah good point. I suppose this kind of talk can get pretty geeky and boring to most people, and I sort of understand that perspective... Although it does remind me of a common musician argument, which is "If I learn music theory, I'll just become bored with music! I like the mystery in not knowing everything about what im doing!" I find myself accidentally talking my producer friends' ears off when I'm a little too excited over some new mixing tricks i've come across... I've really only gotten into this side of music creation within the last couple of years. ive always just played guitar and wrote songs, the more I learn and deep dive into the technical stuff tho, the more excited i find myself as im able to translate my ideas better in my mixing. But yes, i feel most producers/musicians could give a shit about it haha
@@PaulThird this is why I generaly prefer Tutorials in the style of Dan Worall, where it's more about how sound works,what certain tools do and what effects you can achieve with them, not what exactly you are supposed to do because let's be honest, if you mix Dubstep, Breakcore, Dariacore or whatever other neiche genres a lot of us hobbyists make, most conventional Mixing advice falls appart immediatly. In those cases it's really about improvising, not tried and true "professional" methods.
1:28 when the signal crosses the threshold of a full-band compressor the whole signal is compressed, not just frequencies that triggered the detector. Boosting 4K will make it more likely to trigger the detector but when the compressor kicks in it will compress 300 Hz just as much. Hugh pass in the side chain will reduce the low frequencies' influence on the detector but when the compressor kicks in the low frequencies get compressed too. Sure, you can do emphasis/de-emphasis and tune the nonlinear effects but that's something else.
@@PaulThird I listened to that section a few times to make sure I heard what I thought I heard. It sounded as though you were explaining that certain frequencies will be compressed more or less commensurate to their amplitude in the signal. I mean, I know you know the difference between the detector circuit and the gain reduction circuit, and between full-band and multi-band compression so I was slightly taken aback. It's possible that I missed a nuance in something that you said. It was pretty late and I was pretty tired so I apologize if I misheard. I'm not trying to gaslight anyone!
That's cause it can technically happen when compressing a less complex source that you can abuse eq with. Run a sweep through an la3a compressor or a compressor that predominantly only really makes low order harmonics when compressed and have it set in a way where everything is just under the threshold. No compression. And then crank a high boost say 8k bell so the compressor is only reacting to the frequency range of that 8k bell that you've boosted. Print the sweep and you'll see that there's barely any harmonics in the sweep.. Because the compressor hasn't reacted to the low end information. Yes the compression ducks the overall level but only when that 8k increases the threshold. So when the 8k bell is louder than the rest of the source that's only when the compression kicks in so the rest of the frequency range isn't ever triggering the compression. So set up a certain way, it can be argued that certain frequencies are compressed less as dynamically they remain the same throughout the song UNLESS the 8k bell exceeds the threshold. So it's not like normal compression where all of the frequency spectrum is evenly compressed because in this case the compression only triggers when the 8k bell frequency range you've boosted exceeds the threshold. The compressor is only ducking when when it sees the 8k boost That's what i mean by that and what's why i put text at the bottom of one part discussing why CLA will still slightly let the whole frequency range of the guitar exceed the threshold so he can engage the colour of the la3a compressing the whole frequency range, especially the low end which generates that low harmonic behaviour when triggering the compression and result in that percieved weight people get from an la3. So it's got the colour but really only heavily clamping down when that 8k region is too much which is important as at that particular point the 8k is louder than the rest of the frequency spectrum in that source which means those other frequencies are less perceivable anyway so you don't really notive the gain reduction as much to the rest of the spectrum. What you hear is the sharpness of the 8k being smoothened out That's kind of what I mean by frequencies being compressed less as they aren't being compressed by their own dynamics, they are only compressed when something else is more prominent and perceivable than it
"when the signal crosses the threshold of a full-band compressor the whole signal is compressed". This is true, but, because of the way we hear, low frequencies are louder then the rest, I mean, when we set the loudness of 100hz to be iqual to, say, 3k to our ears, in reality the low frequency is louder than the hi frequency, physically. But what the compressor gets is the physical signal, and not the sonic equivalent signal. That way, it is probable that the low frequency will hit stronger the processor, in general.
you can break 'the rules' once you've learnt them! or more accurately when you've learnt the science behind them and also understand which to break to get the emotional impact you wish to communicate.
It’s nice to see a video out like this. It’s validating. I’ve actually been doing a little experimenting in this realm. Here’s something I like to do: Once I have my track(guitar, vocal, kazoo, etc) where I want it processing-wise, I set up Spectre from Wavesfactory running into a limiter and on the limiter I set the ceiling to wherever the particular track is peaking. Then I’ll boost fundamentals frequencies on the Spectre into the limiter. My philosophy behind using the limiter set to the peak of the track is that it won’t actually affect the dynamics of the tracks since it’s not actually clamping down on anything that was previously there. It’s just creating density wherever I boost. I first tried this on a bass that I was having trouble getting to cut through. Worked like a charm.
I know jaycen joshua uses spectre a lot instead of eq. Something I've not really experimented with fully. I think it's great when we think out the box 🤓
@@PaulThird, it’s a pretty neat plugin I have to say. I love that it only boosts. Somehow that puts my brain in that frame of mind. “Which frequencies sound good” instead of always in search and destroy mode. 😂
IMHO EQing into the dynamics section (compressor and expander) is an important part of the sound of old SSL desks. Even if you're not using any outboard, you should switch the board dynamics after the EQ ("Channel Out" button). This way you can boost 8k full whack without getting harsh and drive the dynamics section. Good on kicks, snares, toms, guitars, vocals. In plugin-world I would EQ into compression first and then put a little correction EQ afterwards :D
I always thought guys who mixed on SSL consoles always mixed into compression. I know a lot of current engineers EQ after compression now. I thought it was just personal taste. Didn't know many engineers considered this a rule. Excellent video how you explained it!
Very good engineers also use compressors as time based eqs, compressors can be used for way more than just level/dynamic control, and way more can be done than just harmonic or overall tonality differences using compressors. Look into it and you will see. It seems it isn't common knowledge, but plenty of great engineers use compressors for way more. I'm sure if you look up using compressors as eqs you can find some bits on it
I kind of wish I never made it now cause it seems like loads have taken it like I'm saying what you should or shouldnt do when like last week, I'm just showing the facts of what can happen when you do x, y or z. Why people say what they say and do what they do. I think it's shown me that many do eq into compression and hate the modern "rule" of eq'ing after compression as it makes them feel like they are doing something wrong when they aren't. It's just 2 ways to skin a cat. Its reminded me why I despise audio forums haha everybodys gotta be right haha
@@PaulThird You're a man after my own heart Paul, a dislike of forums, incredible knowledge and rational debate through facts and genuine interest. With a DD postcode to boot ye auld boot x
great video.. i think each way has a different identity and have a discrete difference between the methods however as you stated neither are wrong.. be creative and put multiple Eqs and compressors in your chain...
That's weird! I've always heard the opposite as a "rule." 90% of what I've always heard or been told as a general starting point rule of thumb (which can always be broken) is to at least do any major corrective/subtractive EQ *before* compression, so that the frequencies you're trying to attenuate are not potentially getting exacerbated by the compressor, making their attenuation even harder. As a similar idea to getting your recordings right at the source, the idea with that is that you're removing the EQ problems at the source (i.e. before they get made worse by the comp). You can of course always do whatever you want if it sounds good, but generally this thought process makes sense in both theory and in practice. If you're following the method I described, in order for the subtractive EQ to be before any dynamics processing, that EQ is likely done on right the channel instead of on a bus, which means even if you do some more EQ after the compression on the channel, it's still usually going into some more compression on the bus it feeds into, and if not, it absolutely will get some when that bus finally feeds into the Submix, and if for some reason you have absolutely no compression on your submix, (whether it's because you mix into it and then remove it for mastering, or because you just don't use it for whatever reason) the mastering engineer will end up using some. EDIT: I realize you addressed that by saying they may potentially be brought *up* by the makeup gain increase that usually is accompanied after compression, however I find if you attenuate those problem frequencies, and then do your compression, (depending on the compressor) you usually get a more balanced EQ result even if the problem frequency isn't actually 'quieter' than it was before. So you end up with a more balanced frequency range to do more EQ on if you want, leaving you free to either attenuate those area's even m ore if you really hammered your compression, or just to simply have a more balanced canvas to make more creative EQ moves on.
Yeah it makes sense 🤓 that's why I'll do reso a lot before compression but I suppose what's more interesting is that in my era or circles I've been in, eq'ing after compression was more of a "rule" than eq'ing before compression but I think that it's more about what many class as big eq moves instead of what I call clean up eq. I do both. I clean up, compress and then eq tonally Tbh I just go with what makes sense to me and what I prefer. In this video all I've tried to is show people what happens when you do x, y or z and then left it up to the viewer to experiment with but it seems many have taken offense to it and hammered the reasons why you should eq before compression when all I'm trying to do is show why a certain technique from a mixer works the way it does. The "rules" are what kills me in audio. I stopped believing all that shit a long time ago and I just go by my own rule book that makes sense to me. It fascinates me though why people get so consumed about how others mix and it turns into heated debates. In all honesty, I learn the people I aspire to but I still don't do everything they do, I just try to undetstand it and then make my own mind up Honestly this video has been a strange one for me haha
The question of eq before or after comp was simple for me... I've always done it before. I had heard early on some good reasons to do it that way, not many good reasons not to do it, and so I just made that method how I work. I have found recently that, while using the SSL 9000J plugin from BX, for certain tracks I would be compressing and EQing, realize the default was set to comp first, switch it to EQ first, hate the sound, and switch right back. It depends on the tool and the sound, but I'm more careful now to experiment with EQ before and after to get the best result. Hard fast rules for me have sometimes stagnated my growth as a mixer, so I try to be open to new ways to do things in order to get the best result.
Never heard of a rule to not eq before compression, in heavie edm the general advice is eq before AND after the compression. Love bandpassing before an OTT, just ads a weight that's hard to get any other way.
It's the Internet.. There are rules for everything haha especially if you go on forums. That's why I left them all as there were constant "DO's and DONT's" from "professionals". That's why I do what makes sense to me and drown everything else out. If you believe you can't get any better out of how you do things and your clients are happy with the end results then you just crack on with it. Rules in audio is like well behaved kids.. Never gonna last long haha
@@PaulThird Ah, that makes sense ^^ I guess I don't really hear much about such rules is because in dubstep (which is what I make) "professionals tell you not to do this but..." is the introducing sentence before like half of all mixing conventions we have xD
See for me, as a self-taught person, Comp after EQ always made sense in the digital domain because you're shaping what gets compressed. If you're tracking analog then comp/gates before EQ make sense because you're leveling the source prior to tone shaping going to 'tape', but when looking at the audio from a DAW perspective, where you can individually gain-adjust samples of audio, that whole use case breaks down completely and compression becomes largely about style and character (which I don't think is a bad thing)
Paul, will you ever make videos about how you analyze plugins and what you’re looking for so maybe us at home could start to do our own analyses on the plugins we use?
Important to note CLA also adds low end weight the SSL lacks by adding a pultec pair to his mixbus boosting the lowend and mids. Welcome to the CLA rabbit hole.
I’m about to get 2 SSL EQ’s 500 series very soon and know I need to buy an LA3A clone after watching this video! I have 2 1176’s so i’ll try with those first before buying an La3A! Thanks for the video 🔥🙌🏻
It's all by ear for me but I do a lot of hp & LP on my tracks but it's never specific values. It's all about making room for other things in my eyes and it's all done by ear. Sometimes I just leave tracks as they are but I still play about just to see if the source needs it within the context of the mix
@@PaulThird thx bro I think I do a lot of low cuts where I don't necessarily need 10Hz-30Hz. I think I do almost every track. Especially if the recorded source is not a low-cut source. What do you think, Paul?
Think of filters as making room for other instruments. So if you want your kick to dominate the low end maybe 30 odd hz and your bass 50-60. If you want more cymbals in your overheads then hp in the mix until it sounds right. I leave a lot of room for the kick when I mix drums. Guitars, vocals.. Just get that low rumble out there 80-150hz if needed. Just make room for stuff and use your ears to see if its needed and how much
@@닥터최 Bare in mind that some people low cut bass/kicks etc before hitting compressors to create a little more headroom, but use filters which change the phase rotation and therefor actually end up with less headroom than they started with before the low cut as a result. A guy "teaching" hiphop mixing on Warren's channel did this and it was entirely overlooked through not being realised
CLA does this digitally too... he just doesn't think it's an issue and neither do I. I love engineering tech and it's good to know the rules, but none of the renowned mixers everyone loves care a fuck about this. If It sounds bad they move on. Phase shift can be a great way to get unique sounds too.
Remember phase rotation will often reduce your headroom, but also rarely show up on digital meters as happening, as it creates internal errors as a result. It can also result in hitting plugins a few dbs hotter than you're aware is happening and in some cases internally clip them depending on the plugin being used. Something like RX can be used to fix the phase rotation
@@UncleBenjs i've used waves in phase in the past when i've gotten nightmare overheads or something like that. I love that we have tools when we need them, but i personally wouldn't worry about it unless i hear something in an obvious way
With the right combinataion of EQ and compressor you can achieve things. different EQs have different curves and difference compressors different att rls times. Everybody has his chain... The order of these two is in the answer of the next question. Do you want to EQ the compressor or compress the EQ? One makes the sound more airy and maybe relaxed yet dynamically somewhat under control and the other makes the sound much more controlled dynamically, less relaxed but weighted (respectively). And these options might have to do with if you want the sound to sit relaxed in your mix not demanding attention but noticed only when the listener looks for it, the the other is more in your face demanding attention from the very start.
Exactly. If you have the undetstanding of how each affects the other then you make your own decision and then just crack on. Both ways work but it's whether you prefer the difference in sound compared to the other. Plus it's also about what makes more sense in your head a lot. The clearer your vision is the more visable the end destination is
I make EQ before compression exactly bacause of that, the raw sources usually have a lot of low mids information and I don't want that to trigger so much the compressor, and if you cut the same frequencys after the compression which triggered the compressor, what is the point?
Depends on how you see compression. I see it the opposite way to yourself. I want the compressor to focus on the natural dynamics of the source. I want it purely to compress. That way when I eq to taste my eq decisions are not affecting the compression so the eq is purely eq and I'm eq'ing on top of an already dynamically level source. But I see eq in 2 ways. Clean up and tonal. I clean up first, smallish moves. More resonance based and then compress. If clean up and tonally eq into the compressor the eq and compression are fighting each other when I don't think they need to but I still understand why others would disagree Its 2 seperate jobs so in my mind, if I eq heavily (tonally) into the compression then I'm altering how the compressor does its main function which is purely reducing the dynamic range and the eq's job of altering the tonal balance is also affected. In all honesty, both ways have their own validity so its all about what makes the most sense to you
@@PaulThird well yes, I just remembered that I always compress my signal in the way in with hardware, so really that makes it compress before EQ, and also that is why when I mix I make EQ first and then put a last compressor listening to complete mix, in that case I just wanna make clear that the track has the dynamics and also the envelope to sound right in the mix. So yeah I think I do both 😜
Yeah I was thinking about tracking and how many won't actually know that they are eq'ing after compression.. But also eq'ing into another compressor haha I don't think about these things too much. I have my way of doing things and just cloud out the rest. I still remember the last time I was in a big studio with a few guys and this one guy kept on critiquing everyone cause they didn't do things how he did things and as soon as I said "mate who gives a f*ck what you do?" and his response was "me".. And I replied with "exactly, your the only person in this room that gives a f*ck what you do".. He realised to keep his mouth shout and let people do what they feel is best. Personally I think that's the crux of the Internet. Unless you have a few Grammys, worked with a few big artists or have a shit ton of YT subs and views nobody cares about what Jim down the road does so Jim down the road shouts louder at more people so they take notice of what they do. Honestly, it baffles me why people even ask me for advice, I try and not be rude but it still confuses me why they wouldn't experiment themselves or read a few book or interviews online. Just try shit out and make your mind up haha I much prefer debating things with people who do what they do and know why they do it. That's where we learn more from the opposite perspective. As usual.. I needlessly digress 🤣
A raw source with no EQ that is rich in mid or higher frequencies will make the compressor act more on those frequencies using your logic right? ....so EQ'ing before compression in this instance to even out the source so it is evenly rich across the frequency spectrum would allow a consistent level of gain reduction right?
Correct. Well it would mean less of that frequency range was exceeding the threshold and thus not contributing to the detection of the overall gain reduction, thus technically more natural sounding levelling. Could mean that the detection of the compressor isn't being influenced more often by a certain strongly pronounced frequency range and thus allowing certain other frequencies to engage the overall gain reduction. If a certain frequency range is a lot louder than the rest of the spectrum then compression will only bring down that frequency range dependent on the amount of gain reduction. So if say 7-10k is nearly always 3db louder than the rest of the spectrum and its the only thing exceeding the threshold and you only have 2db of GR it will only decrease the level of the 7-10k frequency range as the gain reduction isn't enough to reach the rest of the frequency spectrum. Think about a common trick for a drum bus where you just touch it about a db to catch the snare peaks. In reality what's happening is that you've set a high threshold where it looks for certain frequencies of the snare to exceed the threshold. When they do the overall gain reduces by half db/db which normally is how much the snare is louder than the rest of the drums and if you set a fast enough release the compression will act as snare peak reduction which adds that extra bit of leveling.
I might be wrong, but it doesn't seem to make sense presented like this. The lead guitar sound is obviously taken from a mix but it's played on it's own here. The raw guitar sounds the best for me of all three but none of them sound particulary good soloed like that. But my point is: even with the 8k boost, the 8k freqs aren't the loudest of the sound so the comp will not react differently to the boost. You could as well put the EQ after the comp and it would sound pretty much the same. (Unless LA-3A has a characteristic of being more sensitive at about 8k region.)
That's why i purposely made an unlisted video with lots of variations within the context of the mix and put the link in a pinned comment.. Plus you can hear on the example and also see it in the gain reduction that the la3 compresses an extra 2-3db with the eq on so it is reacting to the boost. You can also hear the top end smooth out with la3 in this video and if you do it yourself, the more GR you apply the more the 8k boost is smoothed out. That's specifically why he eq's into the compressor. He pushes the 8k into the compressor and goes too far to then dial it back with compression for that specific sound. You can't do that the opposite way around. It's a different sound as the compressor is detecting differently pre eq. It just is what it is. Both have a sound and both sound different
@@PaulThird Yeah the idea makes sense, just that in the actual video, this video, with frequency graphics on screen, it didn't seem like. I complained about the presentation of it in this video and didn't look for any comments, pinned or anything else.
Well i put other audio examples in a seperate video as most don't listen to audio examples more than 30 seconds so this video is purely explaining why the technique is used and the other video was made for those who needed more context
Personally, as someone who works exclusively within the Electronica realm, - I will always use corrective eq before compression, and tone shaping afterwards. Hitting a compressor with unwanted frequencies, can really clog the compressor up, causing it not to function as well as it should. Also, working with samplers, drum machines and synthesizers, I personally never use eq boost, other than adding a tiny bass boost on the main bus, if I want high frequency content, I can introduce it at source. I should point out, I work outside of the box, and only use the computer for recording the stereo outputs from my mixer... Worked like this for almost 40yrs, and get solid sounding pieces all the time, don't even track anymore, just record live performances, - I love the "path of least resistance" approach, very liberating once one knows how to get their own sound. Plugins get in the way, keep things simple if one is using a DAW, or an all analogue setup for that matter. - If one can't get a mix or performance sounding, relatively, like a commercial release, with just an eq and compressor on the main inserts, 1 good reverb on the aux sends, some channel eq, channel compression, and a drum buss compressor, - learn how to, before adding all the other bells and whistles, and practice, - good headphones are essential, - HD600 and HD650 sets are my personal favourites. I would suggest just really learning one parametric EQ, and one VCA compressor first, really get to know how they work, and how they behave on the material one is working with. And if one is lazy, there is no shame in using a dual 15 band graphic eq on the main bus, they can really add a little extra glue before the mix hits the compressor. Get sounds correct at source, and mixing is easy. I'm only a hobbyist, but I have fun, and like my results, - and I've spent a small fortune filling my house with lovely audio toys over the years, but one doesn't need to spend a fortune to get a good sound. Influences, - Meat Beat Manifesto, King Tubby, Queen Adreena, Japan, 60s funk, Two Lone Swordsmen, IAMX, Giant Drag, etc love to ALL, feel no hate
@@PaulThird I don't get why they code plugins that generate harmonics above niquist. Why not just put a low pass filter on the "harmonic generator" part of the code so that any harmonics generated above 20khz are just filtered out. Ie. An anti aliasing filter.
The only "rules" I really follow are gain staging. It's been like that since at least 2008. I do like to comply with standards though whether it is TH-cam loudness or CD Redbook. Unfortunately engineers who work for major labels do not follow those guidelines anymore for some reason...
Bit of a backlash and nit picking going on from some but tbh expected when you've pretty much been nailing youtube recently. I finally cracked the audio YT code!! Haha if this hits 10k which I very much expect it to that'll be three 10k+ videos on the trot which for a near 13k sub channel shouldn't really be happening
I cant believe gain staging is a thing in 2022. Of course reasonable levels...but being constrained? Do you know what 32bit maths means for mix engines? And that was 2005....i think 2011 was 64 bit. Analog yes its critical...you cant clip 32bit much less 64...its floating point...try it? Plugin in mfg who have clippable io need to be moved on
Remember analog emulations are calibrated to act like analog vu though. Compressors, eq's, saturation plugins etc. 32 bit float won't stop aliasing or excessive saturation from overloading the emulation past its calibrated 0vu Still gotta have that control when using non linear processes.. In my opinion
The comment section is intense in this video! Great video. I have a question, Paul. In your opinion. Would aliasing stop being a problem in some years? I don't even know how aliasing sounds but I'm worried about it since I've heard about it in your channel some weeks ago. I guess you gotta get to know every plugin you use, but it would be nice not to be worried about it in the future.
Once the average computer is like M1 cpu power most will probably internally oversample to cut it out from any real audible aspect and not even offer an option. I can see it happening. No point to have it in and most analog emulations will want it removed to stay truer to the original so it's really just waiting on cpu power to improve and become more widespread.
If you want to hear very obvious aliasing, take a recording and sample it into an old mpc, akai rack sampler or emu sampler, A/B the original recording to the sample, you'll hear very obvious aliasing
@@PaulThird Oversampling should be an option, I think, cause oversampled signal not always sounds better, in many cases it's not. Or it can sounds worse with 2x oversampling and better with 4x. Oversampling applies filtering and filters changes the sound, that is not want we want to. So we should be able to decide it's good or bad in every case. Almost all the Analog Obsession plugins has oversampling option in it and I rarely use it. But of course there are cases when oversampling can save the sound.
I'm talking about linear phase oversampling. Melda's HQ oversampling is the best I've seen by a country mile. Up to like x16 oversampling and no change to phase. What you are referring to is IIR filter oversampling
@@PaulThird ok, but doesn't it cause other problems like pre-ringing? The plugin Chow Tape comes to mind, which offers to choose between two types of oversampling and up to 16x. Melda’s stuff is crazy, but I just scratched the surface.
The more I’ve been hearing about the importance of the order you use your plug ins like EQ after compression it really makes me question whether or not using these plug ins separately would make a difference. If you took the vocals and only use compression and then bounce the track, now you have this compressed audio track that you run thru the EQ instead of just having this big vocal chain of plug ins or using them all at once.
Nah wouldn't make a difference. Unless you are in protools as their rendering isnt 100% perfect to the original but if your in a daw like studio one there would be no difference
Beautiful. EQing before going into distortion or saturators (or best-case having a beautiful detector-circuit EQ) is very important too - especially for lowend. That's where many producers struggle a lot. Just saying.. distorting, saturating, clipping and EQing a kick or 808 right is a big factor. Like really big. Makes or breaks many hip-hop-productions.
Still find it hilarious that CLA is pretty open about the fact he doesn't use any of the plugins he endorses. not a big fan of his mixes, but i admire his frank honesty in an industry known for a lot of brown nosing and bullshit
That's pretty interesting. Where have you seen that? I always thought he did use at least some. Comps and ssl I knew he wouldn't cause he's got the console and the gear but I thought he did at least use his mix plugin, effects and that reverb plug they released
@@PaulThird He doesn't use Epic, the reverbs and delays are routed into his SSL, basically he doesn't need his plugins because they are made to recreate what he already has in hardware, it makes sense to me haha
It did take a while for you to tell people this is a digital issue not a analog… because I think you had already hit your head against a brink wall enough saying this to be a digital issue already in the last video… lol Thank you again, for making the audio world a bit less “flat”… thank you (again) (lol)… in digital, you can do it, but just kind in mind this issue to not go crazy on it throughout too many tracks Are there anything in the digital world that can undo this annoying artefact? Surely that’ll be the next move for plugin companies making fixing tool plug-ins?
In my own plugin I purposely set the oversampling per compressor so the aliasing was inaudible in each. For whatever reason the vari mu needed x6 OS haha
The frequency labels on the ssl eq’s are deceiving. If you look at it on a scope / plug-in analyzer, 8k is actually 5k - 7k depending on the specific console that was modelled.
And as well bare in mind that ssl's can have pretty wide boosts so the bell covers more than just 8k. 8k is just the centre frequency and its true, many analog frequency dials can be deceiving
I'll say, Paul, I don't see EQ before Comp as a problem. You've basically said that both EQs and Compressors do exactly what they say on the tin. An EQ changes the tone of a source, boosting or cutting specific frequency bands and thus the dynamic content therein as well as changing the dynamic content of the source overall. A compressor reacts to the dynamic content of it's fed source, unless side chain detection EQ is applied. Fairly simple concepts to grasp. But you've failed to illustrate why that is at all a problem. In fact, one could even argue that EQ after compression, which changes the dynamic content of the signal, then "degrades" the compression you've so carefully dialed in. The argument can be applied both ways around and I think that it loses its value because of that. I believe it ultimately comes down to preference, which will be mostly dictated to which way of working you find more predictable. Personally, I like to EQ before compression, because I find it easier to predict how the compressor will react that way.
There's no personal opinion in this video. I'm not saying anything is right or wrong. I'm just taking techniques and showing why they work the way they do and why people say what they say. This video is purely all about the undetstanding. Not about what's a problem. Others create the problems by creating staples and rules where I just do what I do cause it makes sense to me and sounds better. Tbh I genuinely couldn't give 2 flying f*cks about what somebody else does in a mix as its them mixing and not me, but youtube seems to love it 🤷♂️
@@UncleBenjs I'll sometimes grab myself a multiband to shape the tone of a source, yeah. Although that's usually reserved for the high end, when I want to increase brightness while keeping harshness and sibilance under control. I work primarly on speech
@@RadicalRumin Ah its worth pointing out that I'm not referring to multiband compressors when I mention using a compressor as an eq.. tonal changes with a multi band will affect across the entire time domain, I'm talking about tonal changes across the micro time domain, difference between the attack and decay of individual sounds (kick or a snare etc)
Not Eqing before compression sounds ridiculous, if you have unnecessary low end you want to high pass, why would you want that triggering the compression? Eq before seems like it would sound more natural, otherwise the compressor is reacting to a source you aren’t hearing.
Come on now.. "sounds ridiculous" is extremely OTT, especially as its extremely common. As I said in the video, you can either control what your compressor reacts to more by eq'ing before or allow it to simply compress the natural dynamic range. In a lot of mixing scenarios you'll normally find that the excessive low end has been hipassed already in the recording process but to many, compressors are purely for dynamic range and to level out the source. The main point of a compressor is to reduce the dynamic range and if you have lots of excessive low end then you can easily use a compressor with a hp sidechain detector so it doesn't react to it. However at the same time you could just hipass the lows before the compression so that the low end isn't affecting the detection as much but I was talking more about boosts and cuts outside of hp/LP filters. When I was using sonimus Britson routing I would lowpass and hipass certain sources before the compression but it wasn't a conscious decision regarding the compression as it was more workflow, and I don't really believe in hard rules however it makes more sense in my head to eq after compression for the most part as I want to eq to taste and have my source dynamically level. If you eq to taste prior then you'll have to eq again or eq harder into the compression which can mess with the overall leveling of the dynamics. So I level my source first and then eq to taste. That way I know that both are how I want them as I like to deal with them seperately and be more precise. Each to their own really. As a said, as long as it makes sense to you, block the rest of the shit out Each can get certain results and both will sound different but it's whatever works in your head
@@PaulThird if this “natural” dynamic range is what you’re calling a source without eq, consider the fact that microphones and rooms are imperfect: creating multiple resonances and buildups at different frequencies. A decent portion of eqing aims to fix those resonances, that you yourself at 1:04 demonstrate lead to undesirable effects. As far as I’m concerned, not eqing before compression is exactly letting those resonance bumps and dips mess up your “natural dynamic range” Of course with a synthesized sound in, or straight into a daw this isn’t a case. I’m not saying you shouldn’t eq after compression, there are reasons to do that, but I don’t see clear logic to avoid eqing before compression in most scenarios.
Eq'ing after compression means the 2 aren't fighting each other and getting in the way of each others job. What I've learned about this topic already is that there are very strong opinions regarding this and tbh its all about what makes the most sense to you. I don't like the sound of the 2 fighting together but I'm talking about what I call big tonal moves. I still clean up before compression but it's more focused on resonances. Its smaller moves. I want the compressor to tame the dynamic range so I can eq to taste after without fighting against the compression and then possibly ending up with a source that doesn't sound evenly dynamic. If you eq to taste and feel that you need to go making big boosts or cuts to tonally get where you want to be then you are impacting what goes after the threshold and this is impacting how it dynamically treats the source. So it makes sense in my mind to level the source first and then eq to taste. You can do a mixture of both if you really feel that there is a massive build up somewhere and cut before the compressor to make it react less to whatever and then eq after that compression to tonally get where you want to be then makes sense. I think it depends on how you view your mix. If you view eq as a dynamic and tonal tool then makes sense. If you mostly view them as 2 different jobs then you'll most likely want to seperate the jobs and get your compressor to purely limit the dynamic range so your eq can tonally alter the track to taste without adding more or less compression. But as you can see I view eq in 2 ways. Clean up and tonal. I don't want my sources to sound over compressed on the whole where if you make big boosts into compression then to get the source dynamically even there's a possibility you've had to hit the compression harder to compensate for your eq move thus compressing more when the source only needed 1-2db GR in the first place to sound dynamically even. I do undetstand the build up thing and that's why it's important to understand thresholds and how you can alter what triggers compression. Its a very subjective topic and both work, it's just about the sound you prefer and how you get both tools to specifically work the way you want. This video isn't saying do this or that. It's just showing others how CLA's specific technique works and why he does it. Others online make the 'rules'.. Not me
This is not really the reason CLA does that. The reason why CLA does that is to remove harshness. This is also why he does the same thing with vocals. He has said many times that his compressors are basically his de essers.
But he's the one boosting and adding more harshness in the first place 🤔 I agree with what you are saying but he is essentially making stuff edgier and harsher on purpose for the sound of the compressor smoothing that harshness That's why i included the SOS article. He boosted 8k about 6-8db in order to hit 3-4db of compression
1:38 - 'The compressor would compress less or simply not at all in that specific frequency range'. This statement would be true if you were talking about Multiband compression with the bands mapped similarly to where your EQ notch is. But your explanation is very confusing and not really correct. The compressor is reacting to dynamics, not frequency. By boosting a certain frequency - you're also changing the shape of the envelope of the signal. By boosting a specific frequency, you alter this envelope which would likely push certain parts further over the compressor's threshold. This will compress the whole signal more as the compressor is being pushed harder over its threshold. This affects all the frequencies - by changing the frequency content you supply to the compressor you are changing the dynamics of a signal to which it is reacting. Therefore changing how the compressor reacts. At 4:51 you talk about the 'increasing level of the lower frequencies ranges again'. But actually, it's about the lower level dynamics, whether they're low, mid or high it will bring up everything. I feel like you know what you're talking about but it's a very confusing tutorial, especially for beginners. Sorry for the rant. I'll leave this here! :)
I'm talking about what triggers the threshold. And in reality it is always reacting to frequency. That's what audio is. A collection of frequencies. If a certain frequency exceeds the threshold then it triggers the compression which lowers the level of the whole signal as well as bringing up lower level frequencies, not lower frequencies, lower level frequencies. The lower your threshold, the more frequencies will trigger the compression and dependent on how much certain frequencies exceed the threshold the more the dynamic range will be limited and squeezed. Compression is always reacting to the dynamics of frequencies. For example, try puting a compressor through a sweep and set the threshold to where its on the cusp of advancing the threshold. So no compression. Then boost a wide 8k bell for example and watch the compressor react only to the frequencies of the 8k bell. What occurs is that the rest of the frequency range isn't triggering the compression. The reason I know this is because the sweep shows no lower order harmonics and hardly any audible harmonics whatsoever as the cla3a itself doesn't make any real audible harmonic energy past like 800hz or whatever. However when the lower frequencies trigger the compression it brings in all the lower order harmonics and all of a sudden you have a "weighty sounding guitar" So in reality on a real source if you were to set a sidechain detector to not compress the low end and set the threshold so it isn't compressing and put a crap ton of a certain higher frequency into it then it means that the compression reacts only to the dynamics of that boost. Its not reacting to the dynamics of any other frequency range which means that when that certain high frequency is more perceivable than other frequencies the compressor brings down the level of the whole signal but only really clamps the 8k boost unless you lower the threshold. That's why if you set it up like that and look at the signal on an analyser it's only bringing down that 8k boost or whatever it is. The GR doesn't affect the other frequencies because they don't exceed the threshold. Now you could increase your ratio so that it affects the other stuff more but its possible to set a low ratio with hpf sidechain detection with a big enough eq boost going into the compressor so that it exceeds the other frequencies so much that 2-3db of compression wouldn't really affect the dynamics of the rest of the frequency spectrum. What I mean is the whole signal is being attenuated but if you have a frequency hitting 3db louder than the rest of the spectrum and you have 3db GR then it won't bring down the level of the other frequencies, but it will bring up lower level frequencies across the spectrum. That's what i mean. If compression wasn't dealing with frequencies then it would create harmonic distortion throughout the frequency range of the source but it doesn't.
@@PaulThird 4:46 'Increase the level of the lower frequency ranges' vs increase the level of the lower-level frequencies. It's nuanced and obviously, you understand what's going on but I think that is what makes it confusing. Thanks for taking the time to reply - I like the idea for the video and it's an interesting topic. I feel like actually showing an oscilloscope of the waveform might have made it a bit clearer.
In the case of a guitar I would argue lower frequency ranges are brought up more. That's why I said that guitars are mainly mid forward instruments. The low end is really the quitest parts of that guitar but to be clearer I could have stuck with lower level frequencies. Without being rude, what nearly all viewers dont understand is youtube. This video was originally 15 minutes and I condensed it into 9. I've got to take what I have and piece it all together and it's difficult to include everything. You include everything, people don't watch as long and tend to comment less. You leave stuff out and don't double down on everything you get more engagement and from there you can piece the rest in the comments and increase your engagement. The more comments a video has the more people want their voice out there too. It's a strange world social media but you piece it together if you are hyper focused enough. Long story short.. That's why it's the way it is
@@PaulThird @Paul Third I think we both get what's happening here and I do understand that these videos are incredibly time-consuming to piece together and that you run into all kinds of problems just trying to get a good recording done. I've made videos where I've put 3 hours of content into 30 minutes. You're piecing together a jigsaw and it's a difficult process. TH-cam is a mental place and what makes or breaks content also doesn't make sense half the time. I think your channel is great and I appreciate the huge amount of hard work that has obviously been put in. Anyways thx for the chat ✌
The whole EQ before compression is a bullshit fallacy. It can go infinitely backwards - eq before compression before eq before compression before etc etc etc. What if that natural source was OVER eqed and needs to be eqed to be more normal, etc. The source material means nothing as source material, it just depends what is needed or desired.
I don't make rules. The Internet makes rules.. All I did was show WHY people say what they say and WHY people do the opposite All you have to do is go on enough forums, watch enough YT tutorials and read through a few audio education pages and you'll slowly start to see where the "rules" come from. But if you want to shoot the messenger and say idiotic stuff like I'm just making shit up for no reason I'm then feel free.. But that's why there is "rules" in the title of this video.. Its purposely abbreviated.
This philosophy is sooo odd to me. "Only reacts to the natural dynamics of the source" = "best" ?? ... These rigid thinking statements that where very popular back in the day need to stop, context is everything. Maybe that had a lot of truth in pro studios recording real instruments with a super controlled environment but in todays age rigid statements like that make no sense. If you have a sound that is way too boomy, and you remove that after compression with EQ then your comp is reacting to something that you don't want to be there in the first place, basically it's compressing everything over something that is not there anymore. Why would anyone one want their comp reacting to a 'ghost signal' unless it's for side-chaining or creative sound design?? Cleaning up things before comp is standard practice for most people and its neither right or wrong.
I didn't say anything was best. I just showed what each do. I clean up eq into compression, and then do tonal eq after compression. I couldn't give a f*ck what anybody thinks as I don't do rules. I just do what makes sense to me. The Internet makes rules and I didn't want it to look like I was dictating what rules were so I just dealt with it from either perspective. And in terms of comp pre eq.. Think of tracking. How many records have been or get tracked with compression on the way in? Thousands. You just go with what sounds good and makes sense to you
*IMPORTANT*
I can tell already that many are confusing the intentions of this video and I can clarify right now that it's not a video telling you what to do. The eq after compression "rule" is not my thing, it's just something I've read over the years and seen on various mixing videos and seen discussed on various forums and it's more focused on eq'ing heavily and tonally before compression not clean up eq, actual heavy lifts.
Either way, this video is to specifically lead on from last week's video and the intentions of these videos is not to state WHAT you should be doing, but WHY others say you should, and WHY others do the opposite.
Me personally, I do clean up eq before compression, and then my tonal eq after compression. I have my reasons and my own methodology behind doing what I do but in all honesty you shouldn't give 2 f*cks what I or others do.
I'm making these styles of videos to try and give context to those who know WHAT many engineers do, but don't know WHY they do what they do. So they can turn experiment themselves and come up with their own little rules and methods that work better for their mixing approach.
I'm just being crystal clear that I don't want to tell people what the "rules" are as they are different to every engineer and that why the Internet has created them. For every rule you hear, there is somebody else saying and doing the opposite.
AND also, I decided to put audio examples for my own wee 8k/comp creation as an unlisted video just incase nobody cared for it in this video haha
th-cam.com/video/dgtVUlcyaUg/w-d-xo.html
Yeah it's just playing the TH-cam game man. Unfortunately many skip audio examples longer than 30 seconds so it hurts the retention which hurts the video. It's shit I've got to do stuff like that but gotta give the algorithm what it wants
I do the same! Take out unwanted frequencies when in recording to the DAW with my Neve preamp that has EQ’s , after that I almost always go to my 1176 and later to the DAW then I add EQ in the box or on another analog EQ and gain up those nice frecuencies.
Ive been engineering over 30 years and have ALWAYS put EQ before my compressor. The only tonal shaping I do after the compressor is a de-esser on vocals. If you know how to use a compressor, you just make adjustment as you EQ. Seems in this day and age folks are getting too 'in the weeds' with the all technical stuff and not using their ears...turning it into a science project rather than art.
I have also seen a lot of “top level” engineers often put compression after EQ. I think it depends on what situation you’re in
That's the whole point of being an engineer. It's technical. Audio is very much science based and the best engineers I speak to and admire have vast knowledge and understanding of the craft. It's not a one size fits all, it's a "when the source needs it" and knowing when to make certain moves at certain times.
Thats all fine and dandy focusing on your art but mixing isn't just throwing stuff at the wall and being super creative all the time. You are a problem solver more than you are an artist. Your an engineer. Your job is to fix things and make it work.
Sure many do repeat certain things but they know when somethings not working and the workaround for it
Music is why I got into making music, not the technical stuff, so I gotta agree with the sentiment here
@@PaulThird I think his point is, that you don't need to understand what exactly happens. All you need to understand is: If I do this, this is the result. Now personally - for me - it makes it easier and faster to learn to understand SOME of the "technical stuff". Some people say a producer doesn't need to be an "engineer" - I disagree very much on this. I think the producer is closer to the artist (or is an artist in some cases) than the mixer and the mixer is closer to it than the mastering-engineer. But everyone .. really every kind of engineer and artist needs to know how to get a sound. You don't need to UNDERSTAND why it gets you the sound, but you DO need to know HOW to get it (if you want to be professional that is).
Tbh I agree more with you. Maybe it's the definition of engineer but in my head pushing through a mixing engineer I see it as my job to find out why professionals in the industry do what they do as essentially they are the competition. If they do something I personally wouldn't do then I couldn't care less as its their way of mixing and has nothing to do with me but I still want to undetstand why and how incase there's a time I need it.
The biggest problem with "just use your ears" and "if it Sounds good then it sounds good" is that in the big world of mixing it's about being better than the other guy or girl. That's how you succeed.
Just being "good" doesn't cut it. You've got to blow the socks off them and undertstand where your competition is and how they do what your mixes lack.
The best engineers study other engineers. They learn the technical stuff. They obsorb like a sponge so they understand how to be better. There's one thing taking a trick from somebody that worked on one mix but another thing understanding why that same trick worked on his and not yours.
If you make music and that's your main priority then focus on the music but if your a mixer working on other people's music then you better know your shit cause rest assured the other options do and if you run into problems then you better know how to fix them and have the tools to do that.
The biggest mixing engineers on the planet are out and out audio geeks and that's where I struggle as my head is so fixed on mixing other peoples music instead of creating my own.
I get it but the mixers I aspire to all know the technical aspects of audio. They don't know absolutely everything but they know what they need to know and that's actually the most important to learn in my eyes. Is the stuff you need to know. I don't need to learn the ins and outs of dsp and transfer curves etc I just learn what I need to use tools a certain way.
When I was learning I to followed the advise of EQ after compression. I then went to work along a very experienced producer/mix engineer and noticed how much compression was going on. Eqing before compression, multi bands on guitars, and synths. I mixed some stuff and he pointed out - you are not using enough compression. I learnt how working the compression right, using the right type and Eqing into it began to improve how I mixed, to the point where my mastering requires me to do a lot less. It’s similar to CLA, and as I watched CLA it fell into place for me. It’s very easy to get trapped into “methods” taught. It’s not till you sit alongside these guys you realise how actually there was a lot of bullshit floating around some years ago on how to mix. Don’t be afraid to push into things! Don’t be afraid to compress - it’s a sound some love!
Yeah just do what makes sense to you and let the rest do what they do. You've gotta be your own mixer in the end. Truth is I actually do both in the long run as I clean up then compress then tonally eq after that and then I may compress on a bus as well ie drums.
On vocals its clean up, compress, tonally eq and then multiband compression at the very end.
I don't always compress now, I used to but I'm a bit more chill now haha I try it out and if it doesn't make it better in the mix then it's off.
I just don't give a f*ck anymore about what Davie 2 doors down is doing. If he's getting better results I'll pick it apart and undertstand what he does but still might not do exactly what he does. Always gotta make your own decisions cause its right for you
That is exactly why CLA is CLA....because he does things his way as long as they take him to whatever result he wants. He relies on his ears and years of experience. If it sounds good, it is good no matter what was done to obtain the result. Great video by the way!!!
🤓🤓
Thanks Paul, videos looking slick
🤜🤛
Great video as always Paul! Liked button smashed 👍🏼
🤜🤛
I usually put it like so: EQ > Comp > EQ
-1st Eq to cut off the frequencies I don't want to trigger the comp and other minor adjustments. The rest of the chain benefits from what you said.
Everyone happy?
Very good compromise 🤓
That’s definitely how I use EQ too
Great video! I'm so stocked on this video!! I was literally just thinking today about whether compressors (different compressor types) have different effects on the frequency range of the audio they're processing. Might be a cool idea for a video to go through different classic comp emulations with pugin doctor and test how/if they treat the freq range differently. Well done Paul, also love that your making your own plugins!
Yeah could be a decent idea 🤓 tbh before I make any more videos in this style I still need to try and get my head round why certain people call themselves engineers but at the same time moan about others trying to discuss the engineering behind what mixing engineers do haha swear to god I'm struggling with the "too much focus on technical aspects and audio science.. Its all about the music bla bla bla...".. I just can't compute shit like that.. Yeah music is important but that's for the recording engineer, artist and producers job to hash out and mixers job to engineer the record and do all the technical stuff to make it work. Yeah obviously still be creative but still fix it and make it work with the tools you have and fully understand.
I've completely digressed but it baffles the life out of me how guys can call themselves engineers and not be audio geeks. I'm way too autistically literal for shit like that 🤣🤣
It's probably because for a lot of the main audience of youtube music production tutorials, all these jobs are things they have to do themselves, if you do this as a hobby, you don't exactly have the means to hire a mixing engineer, so you have to learn it yourself, even if you're not an audio nerd.
True but when your hit on youtube every way, left, right, up and down with.. DO THIS.. NO DO THIS... DO BOTH.. DONT DO EITHER, DO THIS INSTEAD
It gets pretty convoluted and people end up doing stuff just cause somebody else said it works.
That's why i firmly believe in diving deeper, bit by bit, and taking your time so you can make your own decisions. There's tons of stuff I know big engineers do or have done but I don't do as it doesn't work for how I work.
I've been there in the past learning lots of things and chucking them into a blender of a mix. Lots of stuff that maybe worked in a certain context but not on that mix or all blended together.
You get to a stage where you don't even watch certain videos cause your just not interested and eventually stop getting hung up on why you arent doing what they are doing.
TH-cam and especially audio forums are an absolute mindf*ck for beginners as its all DO THIS and not enough LEARN HOW, WHY AND WHEN TO USE IT
@@PaulThird Yeah good point. I suppose this kind of talk can get pretty geeky and boring to most people, and I sort of understand that perspective... Although it does remind me of a common musician argument, which is "If I learn music theory, I'll just become bored with music! I like the mystery in not knowing everything about what im doing!" I find myself accidentally talking my producer friends' ears off when I'm a little too excited over some new mixing tricks i've come across... I've really only gotten into this side of music creation within the last couple of years. ive always just played guitar and wrote songs, the more I learn and deep dive into the technical stuff tho, the more excited i find myself as im able to translate my ideas better in my mixing. But yes, i feel most producers/musicians could give a shit about it haha
@@PaulThird this is why I generaly prefer Tutorials in the style of Dan Worall, where it's more about how sound works,what certain tools do and what effects you can achieve with them, not what exactly you are supposed to do because let's be honest, if you mix Dubstep, Breakcore, Dariacore or whatever other neiche genres a lot of us hobbyists make, most conventional Mixing advice falls appart immediatly. In those cases it's really about improvising, not tried and true "professional" methods.
Great video, man!
🤜🤛
Hey Paul, any news regarding the podcast? Been waiting for a new episode but unfortunately, there is none :(
We're in discussions right now about content so I'm not entirely sure when we'll be recording again but I'd like to say 2 weeks
@@PaulThird oh alright, amazing, right on time for my travel to berlin :D
btw. Would love to hear some opinions on loudness in music from you both 😁
1:28 when the signal crosses the threshold of a full-band compressor the whole signal is compressed, not just frequencies that triggered the detector. Boosting 4K will make it more likely to trigger the detector but when the compressor kicks in it will compress 300 Hz just as much.
Hugh pass in the side chain will reduce the low frequencies' influence on the detector but when the compressor kicks in the low frequencies get compressed too.
Sure, you can do emphasis/de-emphasis and tune the nonlinear effects but that's something else.
Did I not say that in the video?
@@PaulThird I listened to that section a few times to make sure I heard what I thought I heard. It sounded as though you were explaining that certain frequencies will be compressed more or less commensurate to their amplitude in the signal.
I mean, I know you know the difference between the detector circuit and the gain reduction circuit, and between full-band and multi-band compression so I was slightly taken aback.
It's possible that I missed a nuance in something that you said. It was pretty late and I was pretty tired so I apologize if I misheard. I'm not trying to gaslight anyone!
That's cause it can technically happen when compressing a less complex source that you can abuse eq with. Run a sweep through an la3a compressor or a compressor that predominantly only really makes low order harmonics when compressed and have it set in a way where everything is just under the threshold. No compression. And then crank a high boost say 8k bell so the compressor is only reacting to the frequency range of that 8k bell that you've boosted.
Print the sweep and you'll see that there's barely any harmonics in the sweep.. Because the compressor hasn't reacted to the low end information. Yes the compression ducks the overall level but only when that 8k increases the threshold. So when the 8k bell is louder than the rest of the source that's only when the compression kicks in so the rest of the frequency range isn't ever triggering the compression.
So set up a certain way, it can be argued that certain frequencies are compressed less as dynamically they remain the same throughout the song UNLESS the 8k bell exceeds the threshold. So it's not like normal compression where all of the frequency spectrum is evenly compressed because in this case the compression only triggers when the 8k bell frequency range you've boosted exceeds the threshold. The compressor is only ducking when when it sees the 8k boost
That's what i mean by that and what's why i put text at the bottom of one part discussing why CLA will still slightly let the whole frequency range of the guitar exceed the threshold so he can engage the colour of the la3a compressing the whole frequency range, especially the low end which generates that low harmonic behaviour when triggering the compression and result in that percieved weight people get from an la3.
So it's got the colour but really only heavily clamping down when that 8k region is too much which is important as at that particular point the 8k is louder than the rest of the frequency spectrum in that source which means those other frequencies are less perceivable anyway so you don't really notive the gain reduction as much to the rest of the spectrum. What you hear is the sharpness of the 8k being smoothened out
That's kind of what I mean by frequencies being compressed less as they aren't being compressed by their own dynamics, they are only compressed when something else is more prominent and perceivable than it
"when the signal crosses the threshold of a full-band compressor the whole signal is compressed". This is true, but, because of the way we hear, low frequencies are louder then the rest, I mean, when we set the loudness of 100hz to be iqual to, say, 3k to our ears, in reality the low frequency is louder than the hi frequency, physically. But what the compressor gets is the physical signal, and not the sonic equivalent signal. That way, it is probable that the low frequency will hit stronger the processor, in general.
you can break 'the rules' once you've learnt them! or more accurately when you've learnt the science behind them and also understand which to break to get the emotional impact you wish to communicate.
🤓
It’s nice to see a video out like this. It’s validating. I’ve actually been doing a little experimenting in this realm. Here’s something I like to do: Once I have my track(guitar, vocal, kazoo, etc) where I want it processing-wise, I set up Spectre from Wavesfactory running into a limiter and on the limiter I set the ceiling to wherever the particular track is peaking. Then I’ll boost fundamentals frequencies on the Spectre into the limiter. My philosophy behind using the limiter set to the peak of the track is that it won’t actually affect the dynamics of the tracks since it’s not actually clamping down on anything that was previously there. It’s just creating density wherever I boost. I first tried this on a bass that I was having trouble getting to cut through. Worked like a charm.
I know jaycen joshua uses spectre a lot instead of eq. Something I've not really experimented with fully. I think it's great when we think out the box 🤓
@@PaulThird, it’s a pretty neat plugin I have to say. I love that it only boosts. Somehow that puts my brain in that frame of mind. “Which frequencies sound good” instead of always in search and destroy mode. 😂
IMHO EQing into the dynamics section (compressor and expander) is an important part of the sound of old SSL desks. Even if you're not using any outboard, you should switch the board dynamics after the EQ ("Channel Out" button). This way you can boost 8k full whack without getting harsh and drive the dynamics section. Good on kicks, snares, toms, guitars, vocals. In plugin-world I would EQ into compression first and then put a little correction EQ afterwards :D
🤓🤓
Paul, you in general do great videos but this one was a sweet spank for reality check. Thanks man!!!
At first I only seen up to the "but this one.. " and thought great hear we go.... But very glad I clicked to see the rest haha 🤜🤛
the informal rule that I have used is EQ in to Comp for 'real music' and EQ after Comp for programmed tracks.
🤓
I always thought guys who mixed on SSL consoles always mixed into compression. I know a lot of current engineers EQ after compression now. I thought it was just personal taste. Didn't know many engineers considered this a rule. Excellent video how you explained it!
Very good engineers also use compressors as time based eqs, compressors can be used for way more than just level/dynamic control, and way more can be done than just harmonic or overall tonality differences using compressors. Look into it and you will see. It seems it isn't common knowledge, but plenty of great engineers use compressors for way more. I'm sure if you look up using compressors as eqs you can find some bits on it
I kind of wish I never made it now cause it seems like loads have taken it like I'm saying what you should or shouldnt do when like last week, I'm just showing the facts of what can happen when you do x, y or z. Why people say what they say and do what they do.
I think it's shown me that many do eq into compression and hate the modern "rule" of eq'ing after compression as it makes them feel like they are doing something wrong when they aren't. It's just 2 ways to skin a cat.
Its reminded me why I despise audio forums haha everybodys gotta be right haha
@@PaulThird absolutely! I agree 💯 percent!
@@PaulThird You're a man after my own heart Paul, a dislike of forums, incredible knowledge and rational debate through facts and genuine interest. With a DD postcode to boot ye auld boot x
PH now haha
great video.. i think each way has a different identity and have a discrete difference between the methods however as you stated neither are wrong.. be creative and put multiple Eqs and compressors in your chain...
🤓🤓
That's weird! I've always heard the opposite as a "rule." 90% of what I've always heard or been told as a general starting point rule of thumb (which can always be broken) is to at least do any major corrective/subtractive EQ *before* compression, so that the frequencies you're trying to attenuate are not potentially getting exacerbated by the compressor, making their attenuation even harder. As a similar idea to getting your recordings right at the source, the idea with that is that you're removing the EQ problems at the source (i.e. before they get made worse by the comp). You can of course always do whatever you want if it sounds good, but generally this thought process makes sense in both theory and in practice.
If you're following the method I described, in order for the subtractive EQ to be before any dynamics processing, that EQ is likely done on right the channel instead of on a bus, which means even if you do some more EQ after the compression on the channel, it's still usually going into some more compression on the bus it feeds into, and if not, it absolutely will get some when that bus finally feeds into the Submix, and if for some reason you have absolutely no compression on your submix, (whether it's because you mix into it and then remove it for mastering, or because you just don't use it for whatever reason) the mastering engineer will end up using some.
EDIT: I realize you addressed that by saying they may potentially be brought *up* by the makeup gain increase that usually is accompanied after compression, however I find if you attenuate those problem frequencies, and then do your compression, (depending on the compressor) you usually get a more balanced EQ result even if the problem frequency isn't actually 'quieter' than it was before. So you end up with a more balanced frequency range to do more EQ on if you want, leaving you free to either attenuate those area's even m ore if you really hammered your compression, or just to simply have a more balanced canvas to make more creative EQ moves on.
Yeah it makes sense 🤓 that's why I'll do reso a lot before compression but I suppose what's more interesting is that in my era or circles I've been in, eq'ing after compression was more of a "rule" than eq'ing before compression but I think that it's more about what many class as big eq moves instead of what I call clean up eq. I do both. I clean up, compress and then eq tonally
Tbh I just go with what makes sense to me and what I prefer.
In this video all I've tried to is show people what happens when you do x, y or z and then left it up to the viewer to experiment with but it seems many have taken offense to it and hammered the reasons why you should eq before compression when all I'm trying to do is show why a certain technique from a mixer works the way it does.
The "rules" are what kills me in audio. I stopped believing all that shit a long time ago and I just go by my own rule book that makes sense to me. It fascinates me though why people get so consumed about how others mix and it turns into heated debates. In all honesty, I learn the people I aspire to but I still don't do everything they do, I just try to undetstand it and then make my own mind up
Honestly this video has been a strange one for me haha
The question of eq before or after comp was simple for me... I've always done it before. I had heard early on some good reasons to do it that way, not many good reasons not to do it, and so I just made that method how I work. I have found recently that, while using the SSL 9000J plugin from BX, for certain tracks I would be compressing and EQing, realize the default was set to comp first, switch it to EQ first, hate the sound, and switch right back. It depends on the tool and the sound, but I'm more careful now to experiment with EQ before and after to get the best result. Hard fast rules for me have sometimes stagnated my growth as a mixer, so I try to be open to new ways to do things in order to get the best result.
🤓
Never heard of a rule to not eq before compression, in heavie edm the general advice is eq before AND after the compression. Love bandpassing before an OTT, just ads a weight that's hard to get any other way.
It's the Internet.. There are rules for everything haha especially if you go on forums. That's why I left them all as there were constant "DO's and DONT's" from "professionals". That's why I do what makes sense to me and drown everything else out. If you believe you can't get any better out of how you do things and your clients are happy with the end results then you just crack on with it.
Rules in audio is like well behaved kids.. Never gonna last long haha
@@PaulThird Ah, that makes sense ^^ I guess I don't really hear much about such rules is because in dubstep (which is what I make) "professionals tell you not to do this but..." is the introducing sentence before like half of all mixing conventions we have xD
See for me, as a self-taught person, Comp after EQ always made sense in the digital domain because you're shaping what gets compressed. If you're tracking analog then comp/gates before EQ make sense because you're leveling the source prior to tone shaping going to 'tape', but when looking at the audio from a DAW perspective, where you can individually gain-adjust samples of audio, that whole use case breaks down completely and compression becomes largely about style and character (which I don't think is a bad thing)
Nearly all of my compression is character and tone based. Bar vocals and bass. They are set a very specific way for dynamics
Paul, will you ever make videos about how you analyze plugins and what you’re looking for so maybe us at home could start to do our own analyses on the plugins we use?
Try these 3
th-cam.com/video/qIugQOPDLpA/w-d-xo.html
th-cam.com/video/-igqxB44EnM/w-d-xo.html
th-cam.com/video/ohicgKSmC2w/w-d-xo.html
Thanks for the Queen solo! 🙂
Thank produce like a pro haha 🤓
Great Paul, Thanks again !✊
🤜🤛
Important to note CLA also adds low end weight the SSL lacks by adding a pultec pair to his mixbus boosting the lowend and mids. Welcome to the CLA rabbit hole.
True.. Forgot about the pultec on the mixbus. I actually stopped using a pultec on the mixbus a few months ago
Oh no
Why?
I’m about to get 2 SSL EQ’s 500 series very soon and know I need to buy an LA3A clone after watching this video! I have 2 1176’s so i’ll try with those first before buying an La3A!
Thanks for the video 🔥🙌🏻
🤓🤓
@ 6:50 how many of you went to touch the overload light on the LA3A meter ?
I'd like to hear your thoughts on 10hz-30hz low cut on each track.
It's all by ear for me but I do a lot of hp & LP on my tracks but it's never specific values. It's all about making room for other things in my eyes and it's all done by ear. Sometimes I just leave tracks as they are but I still play about just to see if the source needs it within the context of the mix
@@PaulThird thx bro I think I do a lot of low cuts where I don't necessarily need 10Hz-30Hz. I think I do almost every track. Especially if the recorded source is not a low-cut source. What do you think, Paul?
Think of filters as making room for other instruments. So if you want your kick to dominate the low end maybe 30 odd hz and your bass 50-60. If you want more cymbals in your overheads then hp in the mix until it sounds right. I leave a lot of room for the kick when I mix drums.
Guitars, vocals.. Just get that low rumble out there 80-150hz if needed. Just make room for stuff and use your ears to see if its needed and how much
@@닥터최 Bare in mind that some people low cut bass/kicks etc before hitting compressors to create a little more headroom, but use filters which change the phase rotation and therefor actually end up with less headroom than they started with before the low cut as a result. A guy "teaching" hiphop mixing on Warren's channel did this and it was entirely overlooked through not being realised
So if we become a member, do we get the full Paul Third (Melda hosted) plugin bundle? 😁😎
They are all on mxxx server now
CLA does this digitally too... he just doesn't think it's an issue and neither do I. I love engineering tech and it's good to know the rules, but none of the renowned mixers everyone loves care a fuck about this. If It sounds bad they move on. Phase shift can be a great way to get unique sounds too.
👍
Remember phase rotation will often reduce your headroom, but also rarely show up on digital meters as happening, as it creates internal errors as a result. It can also result in hitting plugins a few dbs hotter than you're aware is happening and in some cases internally clip them depending on the plugin being used. Something like RX can be used to fix the phase rotation
@@UncleBenjs i've used waves in phase in the past when i've gotten nightmare overheads or something like that. I love that we have tools when we need them, but i personally wouldn't worry about it unless i hear something in an obvious way
@@shanelarue8162 You've misunderstood what I said completely
@@UncleBenjs guess so. I'm agreeing its important but disagree that it matters in the long run . Lots of bad sounding records are loved
Can't wait till we get some Paul Third edition plugins. That has to be the natural progression of all of your hard work and expertise in this area 🤌🏾
You can if you have melda mxxx. I wouldnt use any other dev to make plugins with
I like my vocals bright and smooth so I compress first then eq. But I mix rap vocals so idk it might be different. Sounds better to me personally.
🤓🤓
It sounds amazing in the digital domain because of all the natural, beautiful aliasing. 🥴
Seriously though. Nice explanation.
🤓🤓
With the right combinataion of EQ and compressor you can achieve things. different EQs have different curves and difference compressors different att rls times. Everybody has his chain... The order of these two is in the answer of the next question. Do you want to EQ the compressor or compress the EQ? One makes the sound more airy and maybe relaxed yet dynamically somewhat under control and the other makes the sound much more controlled dynamically, less relaxed but weighted (respectively). And these options might have to do with if you want the sound to sit relaxed in your mix not demanding attention but noticed only when the listener looks for it, the the other is more in your face demanding attention from the very start.
Exactly. If you have the undetstanding of how each affects the other then you make your own decision and then just crack on.
Both ways work but it's whether you prefer the difference in sound compared to the other.
Plus it's also about what makes more sense in your head a lot. The clearer your vision is the more visable the end destination is
This happens in all walks of life with skill ..doing things without knowing why is a form of insanity
🤓🤓
great vid!
🤜🤛
I make EQ before compression exactly bacause of that, the raw sources usually have a lot of low mids information and I don't want that to trigger so much the compressor, and if you cut the same frequencys after the compression which triggered the compressor, what is the point?
Depends on how you see compression. I see it the opposite way to yourself. I want the compressor to focus on the natural dynamics of the source. I want it purely to compress. That way when I eq to taste my eq decisions are not affecting the compression so the eq is purely eq and I'm eq'ing on top of an already dynamically level source.
But I see eq in 2 ways. Clean up and tonal. I clean up first, smallish moves. More resonance based and then compress.
If clean up and tonally eq into the compressor the eq and compression are fighting each other when I don't think they need to but I still understand why others would disagree
Its 2 seperate jobs so in my mind, if I eq heavily (tonally) into the compression then I'm altering how the compressor does its main function which is purely reducing the dynamic range and the eq's job of altering the tonal balance is also affected.
In all honesty, both ways have their own validity so its all about what makes the most sense to you
@@PaulThird well yes, I just remembered that I always compress my signal in the way in with hardware, so really that makes it compress before EQ, and also that is why when I mix I make EQ first and then put a last compressor listening to complete mix, in that case I just wanna make clear that the track has the dynamics and also the envelope to sound right in the mix. So yeah I think I do both 😜
Yeah I was thinking about tracking and how many won't actually know that they are eq'ing after compression.. But also eq'ing into another compressor haha
I don't think about these things too much. I have my way of doing things and just cloud out the rest.
I still remember the last time I was in a big studio with a few guys and this one guy kept on critiquing everyone cause they didn't do things how he did things and as soon as I said "mate who gives a f*ck what you do?" and his response was "me".. And I replied with "exactly, your the only person in this room that gives a f*ck what you do".. He realised to keep his mouth shout and let people do what they feel is best.
Personally I think that's the crux of the Internet. Unless you have a few Grammys, worked with a few big artists or have a shit ton of YT subs and views nobody cares about what Jim down the road does so Jim down the road shouts louder at more people so they take notice of what they do. Honestly, it baffles me why people even ask me for advice, I try and not be rude but it still confuses me why they wouldn't experiment themselves or read a few book or interviews online. Just try shit out and make your mind up haha
I much prefer debating things with people who do what they do and know why they do it. That's where we learn more from the opposite perspective.
As usual.. I needlessly digress 🤣
A raw source with no EQ that is rich in mid or higher frequencies will make the compressor act more on those frequencies using your logic right? ....so EQ'ing before compression in this instance to even out the source so it is evenly rich across the frequency spectrum would allow a consistent level of gain reduction right?
Correct. Well it would mean less of that frequency range was exceeding the threshold and thus not contributing to the detection of the overall gain reduction, thus technically more natural sounding levelling.
Could mean that the detection of the compressor isn't being influenced more often by a certain strongly pronounced frequency range and thus allowing certain other frequencies to engage the overall gain reduction.
If a certain frequency range is a lot louder than the rest of the spectrum then compression will only bring down that frequency range dependent on the amount of gain reduction.
So if say 7-10k is nearly always 3db louder than the rest of the spectrum and its the only thing exceeding the threshold and you only have 2db of GR it will only decrease the level of the 7-10k frequency range as the gain reduction isn't enough to reach the rest of the frequency spectrum.
Think about a common trick for a drum bus where you just touch it about a db to catch the snare peaks. In reality what's happening is that you've set a high threshold where it looks for certain frequencies of the snare to exceed the threshold. When they do the overall gain reduces by half db/db which normally is how much the snare is louder than the rest of the drums and if you set a fast enough release the compression will act as snare peak reduction which adds that extra bit of leveling.
"My hardware is like freshly squeezed orange juice, and plugins are like, Tang"
-Chris Lord Alge
I might be wrong, but it doesn't seem to make sense presented like this. The lead guitar sound is obviously taken from a mix but it's played on it's own here. The raw guitar sounds the best for me of all three but none of them sound particulary good soloed like that. But my point is: even with the 8k boost, the 8k freqs aren't the loudest of the sound so the comp will not react differently to the boost. You could as well put the EQ after the comp and it would sound pretty much the same. (Unless LA-3A has a characteristic of being more sensitive at about 8k region.)
That's why i purposely made an unlisted video with lots of variations within the context of the mix and put the link in a pinned comment..
Plus you can hear on the example and also see it in the gain reduction that the la3 compresses an extra 2-3db with the eq on so it is reacting to the boost. You can also hear the top end smooth out with la3 in this video and if you do it yourself, the more GR you apply the more the 8k boost is smoothed out.
That's specifically why he eq's into the compressor. He pushes the 8k into the compressor and goes too far to then dial it back with compression for that specific sound. You can't do that the opposite way around. It's a different sound as the compressor is detecting differently pre eq.
It just is what it is. Both have a sound and both sound different
@@PaulThird Yeah the idea makes sense, just that in the actual video, this video, with frequency graphics on screen, it didn't seem like. I complained about the presentation of it in this video and didn't look for any comments, pinned or anything else.
Well i put other audio examples in a seperate video as most don't listen to audio examples more than 30 seconds so this video is purely explaining why the technique is used and the other video was made for those who needed more context
@@PaulThird Understood, keep up the good work!
Personally, as someone who works exclusively within the Electronica realm, - I will always use corrective eq before compression, and tone shaping afterwards.
Hitting a compressor with unwanted frequencies, can really clog the compressor up, causing it not to function as well as it should.
Also, working with samplers, drum machines and synthesizers, I personally never use eq boost, other than adding a tiny bass boost on the main bus, if I want high frequency content, I can introduce it at source.
I should point out, I work outside of the box, and only use the computer for recording the stereo outputs from my mixer...
Worked like this for almost 40yrs, and get solid sounding pieces all the time, don't even track anymore, just record live performances, - I love the "path of least resistance" approach, very liberating once one knows how to get their own sound.
Plugins get in the way, keep things simple if one is using a DAW, or an all analogue setup for that matter.
- If one can't get a mix or performance sounding, relatively, like a commercial release, with just an eq and compressor on the main inserts, 1 good reverb on the aux sends, some channel eq, channel compression, and a drum buss compressor,
- learn how to, before adding all the other bells and whistles, and practice, - good headphones are essential,
- HD600 and HD650 sets are my personal favourites.
I would suggest just really learning one parametric EQ, and one VCA compressor first, really get to know how they work, and how they behave on the material one is working with.
And if one is lazy, there is no shame in using a dual 15 band graphic eq on the main bus, they can really add a little extra glue before the mix hits the compressor.
Get sounds correct at source, and mixing is easy.
I'm only a hobbyist, but I have fun, and like my results, - and I've spent a small fortune filling my house with lovely audio toys over the years, but one doesn't need to spend a fortune to get a good sound.
Influences, - Meat Beat Manifesto, King Tubby, Queen Adreena, Japan, 60s funk, Two Lone Swordsmen, IAMX, Giant Drag, etc
love to ALL, feel no hate
Wow, this is interesting!
🤓🤓
The guitars of "Nimrod" Green Day album mixed by CLA sound like this haha
🤓🤓
Dumb question probably. But why do antialiasing filters not work?
How do you mean? Like if you were to just low pass your audio at 20k?
@@PaulThird I don't get why they code plugins that generate harmonics above niquist. Why not just put a low pass filter on the "harmonic generator" part of the code so that any harmonics generated above 20khz are just filtered out. Ie. An anti aliasing filter.
The only "rules" I really follow are gain staging. It's been like that since at least 2008.
I do like to comply with standards though whether it is TH-cam loudness or CD Redbook. Unfortunately engineers who work for major labels do not follow those guidelines anymore for some reason...
I only follow the "rules" that make sense to me haha gain staging being one of them
@@PaulThird PS this was a great video!
Bit of a backlash and nit picking going on from some but tbh expected when you've pretty much been nailing youtube recently.
I finally cracked the audio YT code!! Haha if this hits 10k which I very much expect it to that'll be three 10k+ videos on the trot which for a near 13k sub channel shouldn't really be happening
I cant believe gain staging is a thing in 2022. Of course reasonable levels...but being constrained? Do you know what 32bit maths means for mix engines? And that was 2005....i think 2011 was 64 bit. Analog yes its critical...you cant clip 32bit much less 64...its floating point...try it? Plugin in mfg who have clippable io need to be moved on
Remember analog emulations are calibrated to act like analog vu though. Compressors, eq's, saturation plugins etc. 32 bit float won't stop aliasing or excessive saturation from overloading the emulation past its calibrated 0vu
Still gotta have that control when using non linear processes.. In my opinion
The comment section is intense in this video! Great video. I have a question, Paul. In your opinion. Would aliasing stop being a problem in some years? I don't even know how aliasing sounds but I'm worried about it since I've heard about it in your channel some weeks ago. I guess you gotta get to know every plugin you use, but it would be nice not to be worried about it in the future.
Once the average computer is like M1 cpu power most will probably internally oversample to cut it out from any real audible aspect and not even offer an option. I can see it happening. No point to have it in and most analog emulations will want it removed to stay truer to the original so it's really just waiting on cpu power to improve and become more widespread.
If you want to hear very obvious aliasing, take a recording and sample it into an old mpc, akai rack sampler or emu sampler, A/B the original recording to the sample, you'll hear very obvious aliasing
@@PaulThird Oversampling should be an option, I think, cause oversampled signal not always sounds better, in many cases it's not. Or it can sounds worse with 2x oversampling and better with 4x. Oversampling applies filtering and filters changes the sound, that is not want we want to. So we should be able to decide it's good or bad in every case. Almost all the Analog Obsession plugins has oversampling option in it and I rarely use it. But of course there are cases when oversampling can save the sound.
I'm talking about linear phase oversampling. Melda's HQ oversampling is the best I've seen by a country mile. Up to like x16 oversampling and no change to phase.
What you are referring to is IIR filter oversampling
@@PaulThird ok, but doesn't it cause other problems like pre-ringing? The plugin Chow Tape comes to mind, which offers to choose between two types of oversampling and up to 16x. Melda’s stuff is crazy, but I just scratched the surface.
The more I’ve been hearing about the importance of the order you use your plug ins like EQ after compression it really makes me question whether or not using these plug ins separately would make a difference. If you took the vocals and only use compression and then bounce the track, now you have this compressed audio track that you run thru the EQ instead of just having this big vocal chain of plug ins or using them all at once.
Nah wouldn't make a difference. Unless you are in protools as their rendering isnt 100% perfect to the original but if your in a daw like studio one there would be no difference
Beautiful. EQing before going into distortion or saturators (or best-case having a beautiful detector-circuit EQ) is very important too - especially for lowend. That's where many producers struggle a lot. Just saying.. distorting, saturating, clipping and EQing a kick or 808 right is a big factor. Like really big. Makes or breaks many hip-hop-productions.
🤓🤓
You always aim to keep you low end intact.
🤓
There are rules???
I thought the only rule was - “if it sounds good, it IS good”.
Until your client says is shite 😂
Still find it hilarious that CLA is pretty open about the fact he doesn't use any of the plugins he endorses. not a big fan of his mixes, but i admire his frank honesty in an industry known for a lot of brown nosing and bullshit
That's pretty interesting. Where have you seen that? I always thought he did use at least some. Comps and ssl I knew he wouldn't cause he's got the console and the gear but I thought he did at least use his mix plugin, effects and that reverb plug they released
@@PaulThird He doesn't use Epic, the reverbs and delays are routed into his SSL, basically he doesn't need his plugins because they are made to recreate what he already has in hardware, it makes sense to me haha
haha, i make my own plugins in reason to😅the combi is great 4 that
🤓
It did take a while for you to tell people this is a digital issue not a analog… because I think you had already hit your head against a brink wall enough saying this to be a digital issue already in the last video… lol
Thank you again, for making the audio world a bit less “flat”… thank you (again) (lol)… in digital, you can do it, but just kind in mind this issue to not go crazy on it throughout too many tracks
Are there anything in the digital world that can undo this annoying artefact?
Surely that’ll be the next move for plugin companies making fixing tool plug-ins?
In my own plugin I purposely set the oversampling per compressor so the aliasing was inaudible in each. For whatever reason the vari mu needed x6 OS haha
There's no energy at 8k, it will never be triggering gain reduction
But it triggers gain reduction in this video... Cause its been boosted into the compressor..
@@PaulThird he must be crazy boosting the 8k to hit more than the fundamental guitar note! All there is up there is cone breakup and amp hiss!
The frequency labels on the ssl eq’s are deceiving. If you look at it on a scope / plug-in analyzer, 8k is actually 5k - 7k depending on the specific console that was modelled.
And as well bare in mind that ssl's can have pretty wide boosts so the bell covers more than just 8k. 8k is just the centre frequency and its true, many analog frequency dials can be deceiving
I'll say, Paul, I don't see EQ before Comp as a problem. You've basically said that both EQs and Compressors do exactly what they say on the tin. An EQ changes the tone of a source, boosting or cutting specific frequency bands and thus the dynamic content therein as well as changing the dynamic content of the source overall. A compressor reacts to the dynamic content of it's fed source, unless side chain detection EQ is applied. Fairly simple concepts to grasp. But you've failed to illustrate why that is at all a problem. In fact, one could even argue that EQ after compression, which changes the dynamic content of the signal, then "degrades" the compression you've so carefully dialed in. The argument can be applied both ways around and I think that it loses its value because of that.
I believe it ultimately comes down to preference, which will be mostly dictated to which way of working you find more predictable.
Personally, I like to EQ before compression, because I find it easier to predict how the compressor will react that way.
There's no personal opinion in this video. I'm not saying anything is right or wrong. I'm just taking techniques and showing why they work the way they do and why people say what they say.
This video is purely all about the undetstanding. Not about what's a problem. Others create the problems by creating staples and rules where I just do what I do cause it makes sense to me and sounds better.
Tbh I genuinely couldn't give 2 flying f*cks about what somebody else does in a mix as its them mixing and not me, but youtube seems to love it 🤷♂️
@@PaulThird That's totally fair! Thanks for clarifying.
I like to use compressors as an EQ, but that is something else. Look into it as it seems to have been missed here
@@UncleBenjs I'll sometimes grab myself a multiband to shape the tone of a source, yeah. Although that's usually reserved for the high end, when I want to increase brightness while keeping harshness and sibilance under control.
I work primarly on speech
@@RadicalRumin Ah its worth pointing out that I'm not referring to multiband compressors when I mention using a compressor as an eq.. tonal changes with a multi band will affect across the entire time domain, I'm talking about tonal changes across the micro time domain, difference between the attack and decay of individual sounds (kick or a snare etc)
Not Eqing before compression sounds ridiculous, if you have unnecessary low end you want to high pass, why would you want that triggering the compression? Eq before seems like it would sound more natural, otherwise the compressor is reacting to a source you aren’t hearing.
Come on now.. "sounds ridiculous" is extremely OTT, especially as its extremely common. As I said in the video, you can either control what your compressor reacts to more by eq'ing before or allow it to simply compress the natural dynamic range.
In a lot of mixing scenarios you'll normally find that the excessive low end has been hipassed already in the recording process but to many, compressors are purely for dynamic range and to level out the source. The main point of a compressor is to reduce the dynamic range and if you have lots of excessive low end then you can easily use a compressor with a hp sidechain detector so it doesn't react to it.
However at the same time you could just hipass the lows before the compression so that the low end isn't affecting the detection as much but I was talking more about boosts and cuts outside of hp/LP filters.
When I was using sonimus Britson routing I would lowpass and hipass certain sources before the compression but it wasn't a conscious decision regarding the compression as it was more workflow, and I don't really believe in hard rules however it makes more sense in my head to eq after compression for the most part as I want to eq to taste and have my source dynamically level.
If you eq to taste prior then you'll have to eq again or eq harder into the compression which can mess with the overall leveling of the dynamics. So I level my source first and then eq to taste. That way I know that both are how I want them as I like to deal with them seperately and be more precise.
Each to their own really. As a said, as long as it makes sense to you, block the rest of the shit out
Each can get certain results and both will sound different but it's whatever works in your head
@@PaulThird if this “natural” dynamic range is what you’re calling a source without eq, consider the fact that microphones and rooms are imperfect: creating multiple resonances and buildups at different frequencies.
A decent portion of eqing aims to fix those resonances, that you yourself at 1:04 demonstrate lead to undesirable effects. As far as I’m concerned, not eqing before compression is exactly letting those resonance bumps and dips mess up your “natural dynamic range”
Of course with a synthesized sound in, or straight into a daw this isn’t a case.
I’m not saying you shouldn’t eq after compression, there are reasons to do that, but I don’t see clear logic to avoid eqing before compression in most scenarios.
Eq'ing after compression means the 2 aren't fighting each other and getting in the way of each others job.
What I've learned about this topic already is that there are very strong opinions regarding this and tbh its all about what makes the most sense to you.
I don't like the sound of the 2 fighting together but I'm talking about what I call big tonal moves. I still clean up before compression but it's more focused on resonances. Its smaller moves. I want the compressor to tame the dynamic range so I can eq to taste after without fighting against the compression and then possibly ending up with a source that doesn't sound evenly dynamic.
If you eq to taste and feel that you need to go making big boosts or cuts to tonally get where you want to be then you are impacting what goes after the threshold and this is impacting how it dynamically treats the source.
So it makes sense in my mind to level the source first and then eq to taste. You can do a mixture of both if you really feel that there is a massive build up somewhere and cut before the compressor to make it react less to whatever and then eq after that compression to tonally get where you want to be then makes sense.
I think it depends on how you view your mix. If you view eq as a dynamic and tonal tool then makes sense. If you mostly view them as 2 different jobs then you'll most likely want to seperate the jobs and get your compressor to purely limit the dynamic range so your eq can tonally alter the track to taste without adding more or less compression. But as you can see I view eq in 2 ways. Clean up and tonal.
I don't want my sources to sound over compressed on the whole where if you make big boosts into compression then to get the source dynamically even there's a possibility you've had to hit the compression harder to compensate for your eq move thus compressing more when the source only needed 1-2db GR in the first place to sound dynamically even.
I do undetstand the build up thing and that's why it's important to understand thresholds and how you can alter what triggers compression.
Its a very subjective topic and both work, it's just about the sound you prefer and how you get both tools to specifically work the way you want.
This video isn't saying do this or that. It's just showing others how CLA's specific technique works and why he does it.
Others online make the 'rules'.. Not me
This is not really the reason CLA does that. The reason why CLA does that is to remove harshness. This is also why he does the same thing with vocals. He has said many times that his compressors are basically his de essers.
But he's the one boosting and adding more harshness in the first place 🤔 I agree with what you are saying but he is essentially making stuff edgier and harsher on purpose for the sound of the compressor smoothing that harshness
That's why i included the SOS article. He boosted 8k about 6-8db in order to hit 3-4db of compression
Autogain? someone tell White Sea Studio
🤣
🤓🤓🤓
Remember to check out my autism channel if you want to learn more about my life 🤓🤓
Too many newbies focused on phasing and numbers and equations. CLA focuses on what sounds good.
So he didn't learn anything about engineering? Just jumped on a console and winged it?
I will try this with my guitar parts. However I have to say that was a crappy guitar sound just made shite in a different way.
Did you listen to the unlisted video in the context of the mix?
I didn't. I will do so tomorrow. I enjoyed the video. The guitar parts not so much.
1:38 - 'The compressor would compress less or simply not at all in that specific frequency range'. This statement would be true if you were talking about Multiband compression with the bands mapped similarly to where your EQ notch is. But your explanation is very confusing and not really correct. The compressor is reacting to dynamics, not frequency. By boosting a certain frequency - you're also changing the shape of the envelope of the signal. By boosting a specific frequency, you alter this envelope which would likely push certain parts further over the compressor's threshold. This will compress the whole signal more as the compressor is being pushed harder over its threshold. This affects all the frequencies - by changing the frequency content you supply to the compressor you are changing the dynamics of a signal to which it is reacting. Therefore changing how the compressor reacts.
At 4:51 you talk about the 'increasing level of the lower frequencies ranges again'. But actually, it's about the lower level dynamics, whether they're low, mid or high it will bring up everything. I feel like you know what you're talking about but it's a very confusing tutorial, especially for beginners. Sorry for the rant. I'll leave this here! :)
I'm talking about what triggers the threshold. And in reality it is always reacting to frequency. That's what audio is. A collection of frequencies.
If a certain frequency exceeds the threshold then it triggers the compression which lowers the level of the whole signal as well as bringing up lower level frequencies, not lower frequencies, lower level frequencies. The lower your threshold, the more frequencies will trigger the compression and dependent on how much certain frequencies exceed the threshold the more the dynamic range will be limited and squeezed.
Compression is always reacting to the dynamics of frequencies.
For example, try puting a compressor through a sweep and set the threshold to where its on the cusp of advancing the threshold. So no compression. Then boost a wide 8k bell for example and watch the compressor react only to the frequencies of the 8k bell. What occurs is that the rest of the frequency range isn't triggering the compression.
The reason I know this is because the sweep shows no lower order harmonics and hardly any audible harmonics whatsoever as the cla3a itself doesn't make any real audible harmonic energy past like 800hz or whatever.
However when the lower frequencies trigger the compression it brings in all the lower order harmonics and all of a sudden you have a "weighty sounding guitar"
So in reality on a real source if you were to set a sidechain detector to not compress the low end and set the threshold so it isn't compressing and put a crap ton of a certain higher frequency into it then it means that the compression reacts only to the dynamics of that boost. Its not reacting to the dynamics of any other frequency range which means that when that certain high frequency is more perceivable than other frequencies the compressor brings down the level of the whole signal but only really clamps the 8k boost unless you lower the threshold. That's why if you set it up like that and look at the signal on an analyser it's only bringing down that 8k boost or whatever it is.
The GR doesn't affect the other frequencies because they don't exceed the threshold.
Now you could increase your ratio so that it affects the other stuff more but its possible to set a low ratio with hpf sidechain detection with a big enough eq boost going into the compressor so that it exceeds the other frequencies so much that 2-3db of compression wouldn't really affect the dynamics of the rest of the frequency spectrum.
What I mean is the whole signal is being attenuated but if you have a frequency hitting 3db louder than the rest of the spectrum and you have 3db GR then it won't bring down the level of the other frequencies, but it will bring up lower level frequencies across the spectrum.
That's what i mean. If compression wasn't dealing with frequencies then it would create harmonic distortion throughout the frequency range of the source but it doesn't.
@@PaulThird 4:46 'Increase the level of the lower frequency ranges' vs increase the level of the lower-level frequencies. It's nuanced and obviously, you understand what's going on but I think that is what makes it confusing. Thanks for taking the time to reply - I like the idea for the video and it's an interesting topic. I feel like actually showing an oscilloscope of the waveform might have made it a bit clearer.
In the case of a guitar I would argue lower frequency ranges are brought up more. That's why I said that guitars are mainly mid forward instruments. The low end is really the quitest parts of that guitar but to be clearer I could have stuck with lower level frequencies.
Without being rude, what nearly all viewers dont understand is youtube. This video was originally 15 minutes and I condensed it into 9. I've got to take what I have and piece it all together and it's difficult to include everything.
You include everything, people don't watch as long and tend to comment less. You leave stuff out and don't double down on everything you get more engagement and from there you can piece the rest in the comments and increase your engagement.
The more comments a video has the more people want their voice out there too. It's a strange world social media but you piece it together if you are hyper focused enough.
Long story short.. That's why it's the way it is
@@PaulThird @Paul Third I think we both get what's happening here and I do understand that these videos are incredibly time-consuming to piece together and that you run into all kinds of problems just trying to get a good recording done. I've made videos where I've put 3 hours of content into 30 minutes. You're piecing together a jigsaw and it's a difficult process. TH-cam is a mental place and what makes or breaks content also doesn't make sense half the time. I think your channel is great and I appreciate the huge amount of hard work that has obviously been put in. Anyways thx for the chat ✌
The whole EQ before compression is a bullshit fallacy. It can go infinitely backwards - eq before compression before eq before compression before etc etc etc. What if that natural source was OVER eqed and needs to be eqed to be more normal, etc. The source material means nothing as source material, it just depends what is needed or desired.
What rules? There aren't any rules. You're just making shit up. Show me the rules. I'll wait.
I don't make rules. The Internet makes rules.. All I did was show WHY people say what they say and WHY people do the opposite
All you have to do is go on enough forums, watch enough YT tutorials and read through a few audio education pages and you'll slowly start to see where the "rules" come from.
But if you want to shoot the messenger and say idiotic stuff like I'm just making shit up for no reason I'm then feel free..
But that's why there is "rules" in the title of this video.. Its purposely abbreviated.
This philosophy is sooo odd to me. "Only reacts to the natural dynamics of the source" = "best" ?? ... These rigid thinking statements that where very popular back in the day need to stop, context is everything. Maybe that had a lot of truth in pro studios recording real instruments with a super controlled environment but in todays age rigid statements like that make no sense.
If you have a sound that is way too boomy, and you remove that after compression with EQ then your comp is reacting to something that you don't want to be there in the first place, basically it's compressing everything over something that is not there anymore. Why would anyone one want their comp reacting to a 'ghost signal' unless it's for side-chaining or creative sound design?? Cleaning up things before comp is standard practice for most people and its neither right or wrong.
I didn't say anything was best. I just showed what each do.
I clean up eq into compression, and then do tonal eq after compression.
I couldn't give a f*ck what anybody thinks as I don't do rules. I just do what makes sense to me.
The Internet makes rules and I didn't want it to look like I was dictating what rules were so I just dealt with it from either perspective.
And in terms of comp pre eq.. Think of tracking. How many records have been or get tracked with compression on the way in?
Thousands. You just go with what sounds good and makes sense to you
L take