Side Chain compression

T

timvracer

Member
Hi again,
As I keep learning new things, much from these forums, I am definitely fascinated and blown away at the depth of capabilities that experienced mixers use. (I am mixing mostly rock/alternative). I have had to do alot of re-mixing/balancing as I discover ways I fundamentally got it wrong to start with.

So my latest learning is side-chaining. I realized that after many passes at my mix trying to "tame" it, I ended up sucking the life out of it. A large part of that was trying to control some of the high end harshness, but also the low end overlaps. Then I discovered what folks here described as "ducking", and have utilized it. I am using it to duck the bass under the kick drum. I am also using it to duck the cymbals under the snare, as I realized a snappy snare needs to exercise the same frequencies on a cymbal hit, and they were clashing harshly. Lastly, I am using it to duck the piano under the violin when it plays enhancing notes in to bring out the violin as the primary.

Is this overdoing it? Also, it seems (by ear) that I need to use very aggressive attack/release (nearly 0 / 0) in order to keep the "duck" directly under the reference material. I tried a gentler attack and/or knee in the compressor, and a bit more gentle release, but I could notice it "pulsing" with a lack of volume in the bass for a very short time. Is this correct? Or am I doing something else wrong?

I am also interested in other ways folks use this technique or forms of it that I might want to consider in a pretty busy rock mix (2 rhythm guitars, 2 lead guitars, piano, violin and of course bass/drums/vocals).

I can already start to see the need for multiband compression with a side chain so I can reduce the frequency in a sub-mix that matches the intermittent back up vocals which compete heavily in frequency. Not sure how to do that in Sonar, I read that C6 supports it, but not for Sonar.
 
timvracer said:
Is this overdoing it?
Click to expand...

Probably. Impossible to say without hearing what you actually have. Sometimes you want side chain comping dead obvious, other times you want it completely invisible. Never do it just to say you did. Always do it with a clear purpose, and usually as a last resort.
 
timvracer said:
Is this overdoing it?
Click to expand...

Maybe? jkuehlin is right that we can't really be sure without hearing it. The most important thing is whether or not it sounds the way you want.
These days most of my mixes have the guitars ducking the kick, vocals, bgvox, and maybe lead guitar. Maybe the bass ducking the kick and vox too. And the snare boosting the OHs. It seems excessive, but it's all a matter of what works for the mix.
 
Everyone is right, if it sounds the way you want, job done.

However, I've found that most of the time, volume automation work better and faster for bringing down one part behind another (guitars getting dialed back when the vocals happen), and side chain compression works better for things like the bass ducking the kick.

This is because you really don't (usually) want the guitar popping back up between syllables in the vocal, and you don't necessarily want the same recovery time after every line. Whereas, with the bass ducking the kick, it makes sense the you want the bass to pop back up consistently. In fact, if you can get it to pop back up in time with the tempo, you can add excitement and drive to the mix.
 
Farview said:
However, I've found that most of the time, volume automation work better and faster for bringing down one part behind another (guitars getting dialed back when the vocals happen), and side chain compression works better for things like the bass ducking the kick.
Click to expand...

Well said, farview. I agree.
 
I suppose it depends on genres, but I'd say that if you really need to do a lot of this in a typical rock mix, then the problem is likely more in arrangement, performance, and tracking than in the mix itself. Why is the piano playing so hard behind the violin? Why are the vocals clashing with another? The kick and the bass are supposed to push together and reinforce each other and if both playing together adds up to too much bass it's probably because one or both have too much bass on their own. The same idea applies to the snare and cymbal.

I agree with farview that automation is better for things like tucking a guitar under a vocal, but again it would probably be even better if the guitarist would just calm the fuck down for a minute.
 
Excellent feedback, I really appreciate the perspectives. Important learning here regarding not overdoing it, but doing whatever I need to do to bring the best out in the mix. I especially agree with and like the notion of "know what your goal is and what you are trying to achieve" versus doing something for the hell of it. I got caught into the latter and sucked the life out of my mixes.

As a noob (and maybe you can remember what it was like), just hearing from experts the sorts of things that are used in various situations helps me to know if I am crazy or not. I had been struggling for months with trying to get the kick, snare, vocals, and other parts to "stick out" in the mix, and just couldn't get it right. I ended up bringing other stuff down so much that the mix lost all life. Getting educated on sidechaining/ducking was a godsend.

That said, since I didn't know about sidechain, I did use the volume automation to achieve this in the large segments (i.e. rhythm under lead, quieting instruments during vocal verses), and my ear was driving that. But for thinks like kick and snare, I didn't have a solve.

I also found that only certain frequencies were clashing with the vocals (High frequencies against all instruments), but I don't want to bring the whole mix down. So using that principal of trying to meet my goal (rather than just doing something for the hell of it), I split my "all instruments" bus into 2 busses, HI and LOW, and blend them equally (one uses a HPF ad 300hz, and the other a LPF at 300hz). I then use side chain to reduce the HI bus based on the vocal. This seems to have worked wonders, as now I don't get those stacked up HF's with the vocals. I try to keep max reduction to 3db.

I also found using "vintage" style compression helps to limit the attenuation to a certain level (I was looking for 3b), so even if the vocal goes louder (using Sonitus compressor as it is the only option for sidechain in Sonar). I am also now learning why people like various DAW's so much, Sonar is great, but does lack the support for some of the better sidechain compressors like C6.
 
If that works for you, then fine. But if the instruments in the mix don't fit together without doing stuff like that, you have picked your sounds or arranged your parts poorly.

I have noticed a lot of newbies do one thing, EQ everything until it sounds the same as everything else. It's as if the person has a 'perfect response curve' in their head and makes every instrument adhere to it. This causes a lot of the problems you are describing.

What you need to do is put the instruments in the mix like puzzle pieces. Everything needs its own space. For example, the highs on the snare can be at 8k and the highs on the cymbals can be at 12k. The crack of the snare can be at 2k, while you can suck out a little 2k from the cymbals.

The midrange of the bass can be at about 800hz, while the slap of the kick can be at 3k.

Now, with things like piano and violin, it would really help if the parts weren't both in the same octave. Also, you can find a frequency range that makes the violin stand out or sound good, and then take that same frequency range and back it off a bit on the piano.
 
ashcat_lt said:
I suppose it depends on genres, but I'd say that if you really need to do a lot of this in a typical rock mix, then the problem is likely more in arrangement, performance, and tracking than in the mix itself.
Click to expand...

This. ^^^^

If you're having to duck and sidechain so much, and now you're looking ahead to multiband compression, your tracking and source sounds are no good. So yeah, you're overdoing it because you screwed up before you even got to the overdoing it part.
 
To be fair, I guess there is an aesthetic scale here. "Overdone" is kind of a popular style of production these days. I always think of Butch Vig from Nevermind on. Mostly because that's about the time I stopped listening to contemporary popular productions. I'm sure there have been many far worse examples since. But the kids seem to dig it.

It still helps to start from a point of capturing sounds and performances that already look forward to fitting properly. With a good foundation, these kinds of fancy tricks can actually work a lot better. When you use it to fix real problems, it's always going to fail to some extent, but if the mix is already good without it, that little extra seasoning can make it even better. If you're going for that sort of thing.
 
ashcat_lt said:
To be fair, I guess there is an aesthetic scale here. "Overdone" is kind of a popular style of production these days.
Click to expand...

I wonder though how much of it is done because the mix needs it, and how much of it is just that...style.
Like the stupid autotune chirp that endlessly found its way into WAY too many mixes for WAY too many years.
Thank god that stupid phase is over.

AFA the overdone compression thing...I fear it's going to keep going because too many newbs who just started recording are already overdoing it with compression in theor mixes...and they still haven't even gotten their tracking chops down.

For me...the less shit I have to use during mixing means the tracking was done right...but that's just me, I'm not applying FX/processing to chase a "style" of sound, though I guess not going nuts with that stuff is style too. :)
 
Great thread (lots of ground covered, good perspectives layed out etc.
All of that aside, this really got me wheel' turning
VomitHatSteve said:
.. And the snare boosting the OHs. ...
Click to expand...
What an intriguing idea. This is something from your ‘kit I assume; Use the snare to bump the kit/ambience up out above it’s normal place in the mix...
I like it !

I could see playing with the attack, release, to have it lay in timely..
Thank you. Very nice. :)
Now, to remember when the time(s) come...
‘Sticky notes on the screen ..? ;)
 
Greg_L said:
This. ^^^^

If you're having to duck and sidechain so much, and now you're looking ahead to multiband compression, your tracking and source sounds are no good. So yeah, you're overdoing it because you screwed up before you even got to the overdoing it part.
Click to expand...

Yes, but I am just the mixing guy, I did not record the tracks. In all honesty, a big part what I am learning is how to fix/mix poorly tracked songs. This is a pretty much real life thing for most home recording folks anyhow, but especially for me as I don't have a choice. I did initially provide feedback (you should hear the first round of tracks I got... total garbage), so they did redo them, but that was long ago, and the band is essentially not together anymore.

If it was up to me, I'd get them to pay for a studio, real equipment, and track properly, but that's not what I have. :-( I am guessing other readers of this forum may have to deal with similarly recorded material.
 
Farview said:
If that works for you, then fine. But if the instruments in the mix don't fit together without doing stuff like that, you have picked your sounds or arranged your parts poorly.

I have noticed a lot of newbies do one thing, EQ everything until it sounds the same as everything else. It's as if the person has a 'perfect response curve' in their head and makes every instrument adhere to it. This causes a lot of the problems you are describing.

What you need to do is put the instruments in the mix like puzzle pieces. Everything needs its own space. For example, the highs on the snare can be at 8k and the highs on the cymbals can be at 12k. The crack of the snare can be at 2k, while you can suck out a little 2k from the cymbals.

The midrange of the bass can be at about 800hz, while the slap of the kick can be at 3k.

Now, with things like piano and violin, it would really help if the parts weren't both in the same octave. Also, you can find a frequency range that makes the violin stand out or sound good, and then take that same frequency range and back it off a bit on the piano.
Click to expand...

Thankfully, I did learn this so I started from the beginning with separating out the instruments and finding their "place". If you listen to each instrument track separately, they are definitely squeezed into their zones. I use automation as well to craft those zones as appropriate (i.e. open rhythm guitar playing alone utilizes the full frequency spectrum, then I automate it into it's little place when all the other instruments pile in).

However, there is clearly enough overlap with a strong compressed bass and the kick, as well as the mix of 800k->4k instruments with the vocals that some ducking was in order, and it made a great difference.

So suffice it to say that this method was the very last step (it did not sound "bad" without it, but adding it provided a very nice sense of clarity). It gave me the punch I was looking for. However, I fully agree with you that it's important to do all that you describe first, and work hard to make stuff work together without ducking.
 
timvracer said:
Yes, but I am just the mixing guy, I did not record the tracks. In all honesty, a big part what I am learning is how to fix/mix poorly tracked songs. This is a pretty much real life thing for most home recording folks anyhow, but especially for me as I don't have a choice. I did initially provide feedback (you should hear the first round of tracks I got... total garbage), so they did redo them, but that was long ago, and the band is essentially not together anymore.

If it was up to me, I'd get them to pay for a studio, real equipment, and track properly, but that's not what I have. :-( I am guessing other readers of this forum may have to deal with similarly recorded material.
Click to expand...

Ah ok, yeah I hear ya. Having to deal with other people's poorly recorded shit often means you have to throw the kitchen sink at it just to get something usable. In that case, I don't know if you're overdoing it. I suppose not if things are really that bad.

Still, I'd try to keep it as simple as possible before going to these extreme sidechain/ducking/multiband routes.
 
ashcat_lt said:
To be fair, I guess there is an aesthetic scale here. "Overdone" is kind of a popular style of production these days. I always think of Butch Vig from Nevermind on. Mostly because that's about the time I stopped listening to contemporary popular productions.
Click to expand...

I think we're actually on the downslope of a minimalist trend right now. Starting with Lorde, a lot of pop music got really scaled down for a bit (see: Taylor Swift's "Blank Space" and most of the hip hop that's getting radio play). It's starting to clutter back up, but it's not where it was before Lorde yet.

I think these things have always gone in cycles. Rock music got really overblown in the 80s. Then Nirvana reminded everyone of how good simple (if loud) arrangements could be. Then that sound got gradually more and more over-produced until people gave up on it. Etc.

Still, even if you're doing over-blown production and arrangement, giving everything its own space in the spectrum is pretty key - as others have said again and again. A dozen tracks that all sit in that 600-2000 Hz range is going to sound cluttered; 60 tracks that all have their own frequency niche can actually sound bombastic but coherent.
 
I assume you're kinda overdoing it, but I assume that it will be useful for understanding what you'll have to do in the next mixes. Sidechaining a multicomp to get clarity in the mix seems a little too much for me, but YMMV. I assume, that some of the 2000ish pop punk sound did stuff like that, but always be careful that your Sound doesn't get too fatiguing... I often discovered that I had to redo everything after listening the next day...
 
You must log in or register to reply here.
Back
Top