Mixdown tips for allowing elements to punch thru

[Undrig]

EZ all,
Just wanted to know what types of methods you all use when it comes toward allowing different elements to punch through your mix in comparison to other elements. There's so many ways to approach this, I thought it would be interesting to get a discussion going. Even people who I hold in extremely high regard seem to approach everything on a very trial & error basis. Especially curious about any clever bussing and compression techniques.

Wouldn't be fair if I didn't start by sharing a couple things I do often.

A couple ways I work with percussion is to bus all percussive elements aside from the kick and snare to one bus, and then assign the kick and snare to another bus. Then I simply sidechain the percussion bus to the kick/snare bus.

I also often toss a compressor on a return channel, give it some pretty harsh settings with the gain turned down to compensate. Then I use the send amounts on each channel being sent to the compressor as a way to dictate how each channel cuts thru the mix in contrast to the other channels going to the same compressor. The technique isn't too different from nyc/parallel compression, except I have more control of how each channel is affected and my arrange window isn't a mess with duplicate tracks being crossfaded with the original dry channel.

[wub]

Generally if I want to let elements come through in the mix I'll sidechain an EQ on whatever needs cutting (for example, pads/synths) around the frequency that the trigger element is hitting (for example, snare @ 200hz)

So, we'd have;

Mixer channel 1 - Snare (Spectral Analyser, Fruity Peak Controller)
Mixer channel 2 - Pads (Parametric EQ2)

Listen to the snare in conjunction with the Spectral to ascertain what frequency it's hitting at, then using the peak controller as the trigger sidechain the relevant section of EQ to 'duck' whenever the snare hits. The pad will then drop out on that specific frequency range to create room for the snare, as opposed to ducking/compressing the entire pad.

Adjust the curves on the EQ/release on the sidechain accordingly so it sounds right.

[mthrfnk]

Volume automation and volume sidechaining on pretty much everything, and EQ sidechaining as wub described on my synth & bass busses - using the kick and snare/claps as triggers.

I find using busses invaluable when mixing, especially for EQ'ing and compressing multiple elements, for example on a synth buss I'd have an EQ straight cutting the low end, and maybe adding the tiniest boosts on the high end then a compressor to help glue all the synths together. On a drum buss I tend to use something like FerricTDS to add a little something to all of my drums and again use an EQ to check the drums aren't overbearing a frequency range. One thing I do when using vocals is to duck my synth bus volume slightly when there's a trigger on the vocal buss, this helps to emphasise the vocals and means you don't have them competing for the listeners attention with synths in a similar frequenct range.

The point I try to keep in mind is that every specific part of a track what do I want the listener to hear? Everything else at that point is irrelevant - so you can essentially diminish everything else by some means (volume duck/eq cut/filtering etc) instead of making the new element really loud. So for example if you have a cool lead riff coming in, you can duck your other synths, perhaps they're pads and piano chords - make these quieter or less prominent and the listener knows they're still there but is now focused on the new lead you've brought in which should effectively stand out in the mix since you've ducked competing elements, rather than just adding more to the mix making everything louder and more incoherent.

[Undrig]

I have not been able to find a way to sidechain eq with Ableton's eq 8 or any other vst I am aware of. Any ideas in that regard? Would much prefer to do it that way instead of using sidechain compression with the eq bit enabled or use multiband compression. Would like to just do it with eq only.

[audiowaves]

I'm pretty sure the compressor got a sidechain function where you're able to select EQ, the frequencie and filter options...

[SunkLo]

Generally if I want to let elements come through in the mix I'll sidechain an EQ on whatever needs cutting (for example, pads/synths) around the frequency that the trigger element is hitting (for example, snare @ 200hz)

So, we'd have;

Mixer channel 1 - Snare (Spectral Analyser, Fruity Peak Controller)
Mixer channel 2 - Pads (Parametric EQ2)

Listen to the snare in conjunction with the Spectral to ascertain what frequency it's hitting at, then using the peak controller as the trigger sidechain the relevant section of EQ to 'duck' whenever the snare hits. The pad will then drop out on that specific frequency range to create room for the snare, as opposed to ducking/compressing the entire pad.

Adjust the curves on the EQ/release on the sidechain accordingly so it sounds right.

You tried routing the snare to another muted buss, bandpassing at 200hz or whatever and then keying the peak controller with that instead? Then the 200hz of the pads get cut by the 200hz of the snare instead of the whole thing. I love me some sidechain filtering.

Also along the same lines, pre and post eq into compressors or saturators to emphasize what I want to get hit.

[Ada]

I guess I just do the standard stuff. Don't know if others do this but I usually do low and high cutting by cutting very high highs and very low lows with a non-resonant filter (direct cut) and then using a resonant filter to make it sound smoother while still getting rid of _everything_. I used FabFilter before and noticed it's really bad at cutting away the very low lows and very high highs even though it sounds great so now I use Ozone EQ for cutting and the Fab to make it smoother. Sometimes makes a mix sound less muddy and allows me to pump the volume up like 0.2-0.3dB.

Generally if I want to let elements come through in the mix I'll sidechain an EQ on whatever needs cutting (for example, pads/synths) around the frequency that the trigger element is hitting (for example, snare @ 200hz)

So, we'd have;

Mixer channel 1 - Snare (Spectral Analyser, Fruity Peak Controller)
Mixer channel 2 - Pads (Parametric EQ2)

Listen to the snare in conjunction with the Spectral to ascertain what frequency it's hitting at, then using the peak controller as the trigger sidechain the relevant section of EQ to 'duck' whenever the snare hits. The pad will then drop out on that specific frequency range to create room for the snare, as opposed to ducking/compressing the entire pad.

Adjust the curves on the EQ/release on the sidechain accordingly so it sounds right.

When I read stuff like this I feel so bad for using Cubase.
There is no inbuilt multiband SC-compressor and doing it other ways is so much hassle... sounds so easy in FL. FL really has some nice inbuilt plugins that Cubase lack.

[Genevieve]

Generally if I want to let elements come through in the mix I'll sidechain an EQ on whatever needs cutting (for example, pads/synths) around the frequency that the trigger element is hitting (for example, snare @ 200hz)

So, we'd have;

Mixer channel 1 - Snare (Spectral Analyser, Fruity Peak Controller)
Mixer channel 2 - Pads (Parametric EQ2)

Listen to the snare in conjunction with the Spectral to ascertain what frequency it's hitting at, then using the peak controller as the trigger sidechain the relevant section of EQ to 'duck' whenever the snare hits. The pad will then drop out on that specific frequency range to create room for the snare, as opposed to ducking/compressing the entire pad.

Adjust the curves on the EQ/release on the sidechain accordingly so it sounds right.

You tried routing the snare to another muted buss, bandpassing at 200hz or whatever and then keying the peak controller with that instead? Then the 200hz of the pads get cut by the 200hz of the snare instead of the whole thing. I love me some sidechain filtering.

Also along the same lines, pre and post eq into compressors or saturators to emphasize what I want to get hit.

I'll be trying this in Renoise :3

[mthrfnk]

I guess I just do the standard stuff. Don't know if others do this but I usually do low and high cutting by cutting very high highs and very low lows with a non-resonant filter (direct cut) and then using a resonant filter to make it sound smoother while still getting rid of _everything_. I used FabFilter before and noticed it's really bad at cutting away the very low lows and very high highs even though it sounds great so now I use Ozone EQ for cutting and the Fab to make it smoother. Sometimes makes a mix sound less muddy and allows me to pump the volume up like 0.2-0.3dB.

Generally if I want to let elements come through in the mix I'll sidechain an EQ on whatever needs cutting (for example, pads/synths) around the frequency that the trigger element is hitting (for example, snare @ 200hz)

So, we'd have;

Mixer channel 1 - Snare (Spectral Analyser, Fruity Peak Controller)
Mixer channel 2 - Pads (Parametric EQ2)

Listen to the snare in conjunction with the Spectral to ascertain what frequency it's hitting at, then using the peak controller as the trigger sidechain the relevant section of EQ to 'duck' whenever the snare hits. The pad will then drop out on that specific frequency range to create room for the snare, as opposed to ducking/compressing the entire pad.

Adjust the curves on the EQ/release on the sidechain accordingly so it sounds right.

When I read stuff like this I feel so bad for using Cubase.
There is no inbuilt multiband SC-compressor and doing it other ways is so much hassle... sounds so easy in FL. FL really has some nice inbuilt plugins that Cubase lack.

You don't need a multiband comp. The way I do it (and I think wub does too) is to use Peak Controller and a standard EQ - PC is literally the simplest thing ever, if you input a sound it can generate a variety of triggers, so when you place it on a snare everytime the snare hits it generates an automation signal - you can map this to an EQ notch on another channel, so say you have a 200Hz notch on a synth channel, you just link the EQ slider to the PC's automation and play with the settings till it ducks the signal as you want (conversely you could boost the EQ notch too). You could do this manually with any EQ and simple automation curves, with a standard drum beat you'd only have to draw the automation a handful of times for your kick/snare etc.

[Ada]

I guess I just do the standard stuff. Don't know if others do this but I usually do low and high cutting by cutting very high highs and very low lows with a non-resonant filter (direct cut) and then using a resonant filter to make it sound smoother while still getting rid of _everything_. I used FabFilter before and noticed it's really bad at cutting away the very low lows and very high highs even though it sounds great so now I use Ozone EQ for cutting and the Fab to make it smoother. Sometimes makes a mix sound less muddy and allows me to pump the volume up like 0.2-0.3dB.

Generally if I want to let elements come through in the mix I'll sidechain an EQ on whatever needs cutting (for example, pads/synths) around the frequency that the trigger element is hitting (for example, snare @ 200hz)

So, we'd have;

Mixer channel 1 - Snare (Spectral Analyser, Fruity Peak Controller)
Mixer channel 2 - Pads (Parametric EQ2)

Listen to the snare in conjunction with the Spectral to ascertain what frequency it's hitting at, then using the peak controller as the trigger sidechain the relevant section of EQ to 'duck' whenever the snare hits. The pad will then drop out on that specific frequency range to create room for the snare, as opposed to ducking/compressing the entire pad.

Adjust the curves on the EQ/release on the sidechain accordingly so it sounds right.

When I read stuff like this I feel so bad for using Cubase.
There is no inbuilt multiband SC-compressor and doing it other ways is so much hassle... sounds so easy in FL. FL really has some nice inbuilt plugins that Cubase lack.

You don't need a multiband comp. The way I do it (and I think wub does too) is to use Peak Controller and a standard EQ - PC is literally the simplest thing ever, if you input a sound it can generate a variety of triggers, so when you place it on a snare everytime the snare hits it generates an automation signal - you can map this to an EQ notch on another channel, so say you have a 200Hz notch on a synth channel, you just link the EQ slider to the PC's automation and play with the settings till it ducks the signal as you want (conversely you could boost the EQ notch too). You could do this manually with any EQ and simple automation curves, with a standard drum beat you'd only have to draw the automation a handful of times for your kick/snare etc.

Yea I'd have to do it manually in Cubase. Problem with that is that I often change my mind and when the automation is manually laid out it takes alot of time to change stuff if I decide to do so. This is the way I do it now but it is really akward compared to the stuff that FL or Reaper has.

[SunkLo]

I did a search for 3rd party plugins that do the same thing and found a couple:

Blue Cat's DP Meter Pro which also serves as a pretty nice metering plugin. Seems pretty versatile.

Piz AudioToCC Freeware, seems to do what we're talking about.

[Libra]

The Simplest and most effective thing is to choose instruments/synths that simply dont clash in the frequency ranges.
Its not always easy to do, but theres ways of compensating one instrument for another and vice versa to get a cleaner sound.

[Add9]

Generally if I want to let elements come through in the mix I'll sidechain an EQ on whatever needs cutting (for example, pads/synths) around the frequency that the trigger element is hitting (for example, snare @ 200hz)

I've seen this done in FL but has anyone done this in Logic? I don't think the Logic EQ can do frequency-specific side chaining unless you were to bus the signal to like 4 different buses with different EQ settings on each and sidechain compress only one of those buses... not exactly an elegant approach. Or I suppose I could invest in a plugin that has that capability, I heard good things about fabfilter but I don't know if it can do that.

I guess I'm just wondering for those of you who do use this strategy for separating sounds, does it really improve your sound significantly when compared to, say, using regular sidechain compression? I know some professional producers do this (I saw Rogue do a tutorial once where he used this technique) but I'm not sure how important it really is since I've never tried it myself.

[Libra]

Generally if I want to let elements come through in the mix I'll sidechain an EQ on whatever needs cutting (for example, pads/synths) around the frequency that the trigger element is hitting (for example, snare @ 200hz)

I've seen this done in FL but has anyone done this in Logic? I don't think the Logic EQ can do frequency-specific side chaining unless you were to bus the signal to like 4 different buses with different EQ settings on each and sidechain compress only one of those buses... not exactly an elegant approach. Or I suppose I could invest in a plugin that has that capability, I heard good things about fabfilter but I don't know if it can do that.

I guess I'm just wondering for those of you who do use this strategy for separating sounds, does it really improve your sound significantly when compared to, say, using regular sidechain compression? I know some professional producers do this (I saw Rogue do a tutorial once where he used this technique) but I'm not sure how important it really is since I've never tried it myself.

For that I use Melda Dynamic EQ. its not only an amazing EQ, but you can sidechain eq bands/peaks whatever

[SunkLo]

Yeah dynamic EQs are purpose built for this but if you look at my post above, there's a freeware amplitude to cc plugin that you could use for this purpose.

[Coolschmid]

If I am trying to get drums to cut through a heavy brosteppy bass and its ok for the track to sound aggressive, using a transient shaper on the kick and snare is like a 2 second fix that has worked for me so far. Maybe there's something wrong with just doing that but idk. It causes the drums to take over the mix for like a half a second each hit.

[sburton84]

Yea I'd have to do it manually in Cubase. Problem with that is that I often change my mind and when the automation is manually laid out it takes alot of time to change stuff if I decide to do so. This is the way I do it now but it is really akward compared to the stuff that FL or Reaper has.

Are you using a really old version of Cubase? They've included sidechain support in some of the built-in plugins since Cubase 4.1, and the latest Cubase is one of the few DAWs that supports VST3 which is designed precisely for this sort of thing.

[Sharmaji]

for a non cart-before-the-horse answer:

mix those bits louder than the rest of the elements.

sometimes it really is just that easy.

[fragments]

for a non cart-before-the-horse answer:

mix those bits louder than the rest of the elements.

sometimes it really is just that easy.

+1

Also, begin my choosing/designing sounds that already work together. My mixdowns are far from perfect, but I still feel like people start to over-complicate things with tons of EQs, filters and all this sidechainning business (to be fair, I use these and they do have their place). I feel like sometimes producers do things because they can, not because it sounds best.

[extremesociety]

for a non cart-before-the-horse answer:

mix those bits louder than the rest of the elements.

sometimes it really is just that easy.

+2

Gain staging is so important. Make sure there's headroom in your mix so that if you want to push something louder you can without clipping. If something is clipping but it's still not loud enough in the mix, just turn everything down, and tun your monitoring volume up until there's sufficient headroom and it's not clipping. Don't start compressing and limiting all over the place when there's no need/purpose. Don't start doing any fancy multi-band side-chaining until there's a need/a specific purpose for it.

Secondly, smart EQing. If you know you want you're snare to punch at 250 and 1500k points (for example), make sure it's more or less sitting there by itself, or at the very least make sure it's the dominant sound at those frequencies. There are no "magical" EQs for this, but I do love the sound of the SSL and API plug-ins, and I love the sound of the same modules in analog (SSL board EQ and API 500 series EQ). But those are just personal preferences based on my experience.

I think compression in EDM (and really anything) is most useful as a glue if you're not talking master fader. Only use it where it's appropriate and helpful. Running all of your drums through an 1176 with a ratio of 4 may help glue the whole thing together. Sometimes it won't. Sometimes agressive parallel compression with the classic 1176 ALL button setup will help your drums punch. Sometimes it won't. EDM is a funny business, since when you're talking samples you're not really going to run into problems where a live drummer may have played a passage really well, but some of the dynamics need to be tamed with compression. When you have complete control over each track, you can make it do whatever you want without having to resort to classic mixing fixes, which is nice.