EQ then compress or Compress then EQ?

[Synthzizer]

Oh the moral is, you and no one can get things perfect. And who would want that anyway... Just don't overdo it.

 

[MusicWater]

I do things in parallel, so when it comes to EQ before or after, the answer in my case many times could be neither one. I believe you can do a lot of damage to a signal by engaging extreme fx chains in massive ways. In this way you easily lose contact with what you really try to achieve. The way I'm currently doing it is to for each sound source have:

- One completely dry track, no effects at all
- One completely dry track, which is input to some certain effect, such as reverb
- One completely dry track, which is input to some certain other effect, such as EQ
...

Besides this, some of these tracks might then need multiple effects, but the reason for ALL of this is because the need to create certain characteristics of each sound source, that then gets automated into the mix. So I do tons of analyzis at first, to define what are the original characteristics I have to work with. Then, usually I then add my favorite characteristics on each sound source in parallel, such as warmth, but also less "attractive" characteristics. All of these then sit on volume faders. For instance, the less "attractive" characteristics I might add simply to gradually remove it on a cumulative curvature then at the same time gradually add my more favorite characteristics on a cumulative curvature. The combination is dramatically improving the sound. Why this works is because of knowing exactly what you have on each volume fader. With just a couple of volume faders with a lot of effects in serial, you get lost and the mix is locked into certain characteristics, which creates a mix that does not move the listener in a good way.

In other words, I do a lot of FX application on the dry signal to add characteristics I can blend into the overall. I usually try to narrow down the scope of application to max, so when I combine effects, it's because it's the only way of achieving some certain characteristics. When it comes to EQ then compress, absolutely. I do this for instance on frequency band scoped fx application. Compress then EQ? Yes, that too sometimes, for instance because of some bi-product of some fx application that I want to silence.

In other words, my answer is: Focus on the characteristics you want to create of each sound source and what that tells about what kind of fx chains you need to setup and how. The magic happens when you then balance all of these, so the magic is in balancing the right things, in the right way across the mix.

I hope this helps

 

[BroKen_H]

MusicWater. Hi! and welcome. I'm a huge fan of New York style parallel bus mixing, too. Using Reason with the SSL 9000 desk (digital) that's built in makes it a cinch. Vocals come out sounding fuller and fatter, and you don't have to overblow the fx. Much nicer to automate faders than effects. Parallel compression to bring up the quiet, rather than limit the peak, parallel distortion/destruction, parallel EQ all make things

 

[richinmusic]

The only 2 accurate answers to the question
"EQ then compress or Compress then EQ?" are
"yes" and "no".
There are no right or wrong answers for this, and there are always many paths that bring one to the same end-point. The best advice someone once gave me in a similar situation was "try one way, try the other, and then try another. Make a decision or try yet another".

 

[Greg_L]

The only 2 accurate answers to the question
"EQ then compress or Compress then EQ?" are
"yes" and "no".
There are no right or wrong answers for this, and there are always many paths that bring one to the same end-point. The best advice someone once gave me in a similar situation was "try one way, try the other, and then try another. Make a decision or try yet another".

People don't like to try things anymore. They just want answers spoon fed.

 

[MusicWater]

MusicWater. Hi! and welcome. I'm a huge fan of New York style parallel bus mixing, too. Using Reason with the SSL 9000 desk (digital) that's built in makes it a cinch. Vocals come out sounding fuller and fatter, and you don't have to overblow the fx. Much nicer to automate faders than effects. Parallel compression to bring up the quiet, rather than limit the peak, parallel distortion/destruction, parallel EQ all make things

Hey BroKen, thanks, I feel welcome here! Exactly! I want to use the basics - the volume and pan faders - in ways that make more sense - balancing all aspects about the mix using these faders and being able to automate everything. Over the years, some things seemed to confuse me more than they helped, such as group channels. But with New York style parallel bus mixing it kind of fell into place - you want a single fader because you want to control it in isolation, no matter what goes into it. So in my case, when working with a sound characteristics approach, it's very convenient to then route tracks to group channels that share the same characteristics. Let's say for instance I have an electric piano and one of the tracks contains a pure warmth representation of that sound source, then that track I mix with other tracks that contain that same sound representation of the other sound sources as well (for instance a mellow electric guitar), making it possible to automate sound characteristics spanning across mutliple sound sources within the mix and that is nice to be able to mix like that, especially since you also then always have an option to classic EQing. In every mix I want a fader that contains ALL effects, and a fader that contains NO effects - on master scope. An that is possible by always having one dry track of all sound sources, then routing those into a dedicated DRY FX group channel. With the wet channels the same, they end up on a dedicated WET FX fader. That I learned from Dave Pensado, it makes an awful lot of sense, you can mute the dry or the wet channel on one click to get the information of where most of the noise is coming from and be able to do so repeatedly throughout the process - awesome. So in my case I really don't have any limiter on ALL tracks in the mix, not even in mastering, eventhough I also master the music there is always this perfectly clean unprocessed not even dithered track (although everything is processed at 192 kHz/32-bit). Instead, what I do is to balance the two in 16-bit, so whatever distortions were added or removed from not dithering everything, that gets mixed in to taste by ear anyway, so yeah, these days a certain portion of my mixes are always perfectly untouched and heard just as they were recorded. That also indirectly means that the better the recording is, the better will automatically the mix also be. And that directs the focus towards relying on the early processes - to create the right frequencies to begin with.

The reason why I - just like top cats like Bob Clearmountain - have a sound characteristics approach, is because I primarily want to create a mix of various kinds of pleasant and exciting qualities about the mixes that I can enhance throughout the course of playing the mix. I have switched over from a top down to a bottom up approach, quite recently I did so. I found that cutting the amount of small mistakes I make about my mixes, tend to have an enormous final impact. One example of this was my side chaining habits, they tended to be on the sloppy side. Nowadays I'm much more precise about exactly how, why and on what I use side chaining. Another example is my choice of what volume fader I target. Instead of just jumping to the console and correcting stuff almost on auto pilot, I think through what total impact that fader really has and consider/try the options before I make the final move. Because in my case I have struggled with achieving multi-dimensional mix balance, so that has been an area where I've focused a lot in order to improve my mixes.

In other words, my current mixing style is to create and balance sound dimensions in order to produce mixes that are pleasant and exciting to listen to. I have certain dimensions I relate to pleasant qualities and certain dimensions I relate to exciting qualities. My definition of succeeding with a mix is when I have successfully created and balanced these mix qualities. A lot of attention also goes to analyzing the qualities and relating to them, so that I learn what kind of sound I REALLY like. Because sometimes you logically think you want a certain quality and it turns out your heart desires something completely different. Hence being a mix engineer is also a lot about learning about what qualities you resonate with. Having said that, I don't think it hurts to learn more about A) what dimensions are needed to create more pleasant qualities, and B) what dimensions are needed to create more exciting qualities, and C) how to better balance these across the mix, and D) be able to work like Dave Pensado about describing the qualities you have in the mix as well as precisely knowing what qualities various gear/mixing decisions truly have.

I hope this helps!

 

[Rod Norman]

This is a common question. The best solution for someone who has taken the time to consider the issues involved is to use proper gain staging from the start. (the idea is that a good signal will never need eq.) Once you get all the tracks recorded clean, mix starting with the vocal and bass; add the drums (the heart of the recording) then add the chord instruments. Once you are happy listen for any honking frequencies that stand out and gently cut them back with a parametric equalizer (so you can isolate the offending honk and cut only it). Then; and here's the real answer to your question; apply a multiband compressor. This applies compression to only certain frequencies-compression and eq together) I hope that helps.
Rod Norman
Engineer

I was hoping to make a poll, but I'm not savoy enough to pull it off.

Do you compress an EQ'd signal or EQ a compressed signal (assuming you're both compressing and EQing a signal)?

 

[bouldersoundguy]

Using eq in parallel is exactly how you turn the relatively benign phase effects of filters into audible phase interactions. Even worse would be two tracks with different filter and dynamic processing as that would make phase interactions modulate and be even more audible. There's a reason eq is practically always connected in series: parallel eq just doesn't sound good.

 

[mixsit]

This is a common question. The best solution for someone who has taken the time to consider the issues involved is to use proper gain staging from the start. (the idea is that a good signal will never need eq.) Once you get all the tracks recorded clean, mix starting with the vocal and bass; add the drums (the heart of the recording) then add the chord instruments. Once you are happy listen for any honking frequencies that stand out and gently cut them back with a parametric equalizer (so you can isolate the offending honk and cut only it). Then; and here's the real answer to your question; apply a multiband compressor. This applies compression to only certain frequencies-compression and eq together) I hope that helps.
Rod Norman
Engineer

Hmm. Gain staging? If 'a good signal' means recorded well enough' not to need eq, ok sure. But I would not think 'go with a multi-band comp? is the place to steer newbs' with this question?

The only 2 accurate answers to the question
"EQ then compress or Compress then EQ?" are
"yes" and "no".
There are no right or wrong answers for this, and there are always many paths that bring one to the same end-point. The best advice someone once gave me in a similar situation was "try one way, try the other, and then try another. Make a decision or try yet another".

So my two cents.. Actually.. there's three answers.

There is 'Yes' - for when those reasons are in play, 'No' when they're in play... then a shit load of the cases where, well it really just won't matter all that much either way.

Case in point.

I think also one good thing to remember is that if your are using Eq to fix a problem area then your capture was not pleasing. If you are using Eq to enhance then you are putting a gloss coat on to bring out the best highlights. If you cut with Eq then that actually means you have just boosted each side of the cut. Therefore if you compress after the Eq cut you would most likely bring up the cut level again due to the semi normalizing process of compression. This means it has been a waste of time and will result in a smeared sound and worse than when it when in. If you boost Eq you have just cut each side of the boost so to speak. Then if you compress that you will most likely bring up the cut again, & again, this means it has been a waste of time and will result in a smeared sound and worse than when it when in. Now if you compress a signal that does not need Eq & has a stable & even energy spread then you will lift detail out of the troughs/valleys of its dynamic range. This may result in some eq being needed because you have just exposed some harshness or rumble. Now you think we are going around in circles here and you would be right. So what is the moral of the story? flapped if I know..Hahahha....

It's been said in this thread already, but ..a boiled down ver / recap..

When the eq is needed to clean up a track in ways that will give adverse effect to the action of the comp, dump them, then move on to what ever next step. the unwanted low end, wild out-o balance freq's. I.e, the things the you don't want a detector to see', react to and defining your sound.
This is where some thought about t 'It' can come in handy. 'Hmm, will this matter?' Maybe.

The problem with that in the bold, stated too simply the thought process isn't complete.
If you're cut is to something out of whack enough to affect the detector that much- it can be a flaw that need to get dumped first.
But that.. in general, is not in the realm of the kind of gross detector reaction' you're going to typically see with gentle tone shaping - whether cuts' or boost'!
There in lies the (..a) difference.
And.. 'a thought about it' ..? Maybe a hell of a lot better than 'try it both ways' every time it comes up.



Tune in next for 'When the Side Chain Detector Eq 'steps up the compressor game;
- "For when the tracks are shaped good, but I really need to dial that comp in.. EVEN better."