Taming drum peaks? (¯\(°_o)/¯ content)

[narcossintese]

I know that people usually compress between 2 to 4 db of the mix on the master bus comp, that means that the drums are already ultra compressed, otherwise the master bus comp would only work on the drums peaks instead of glueing all the mix elements together. I´m failing at that.


Ok, let´s make a simple drum loop using Steven Slate Drums 3.5
This is how it sounds raw, straight from Kontakt, and this is how the peaks look like:

Sounds good to me, but too much peaks to hit a master bus compressor.

So, to start butchering it I´m going to apply saturation. IMO it reduces the peaks in a more pleasant way than compressing, so I apply as much as I can without utterly destroying the drum tone. I compensate the saturation added by reducing the input level until it has the same volume it had previously. Now it sounds like this and looks like this.

Still too much peaks, so I´m going to clip it in two stages. I set the first one pretty aggressively. It sounds like this and looks like this

The second one goes to the limit of tolerable. If I set the threshold any lower than this the drums will sound like absolute garbage. It sounds like this and looks like this.

Nothing more that I can do, so I add the bass, add some dbs and the mix sounds like this and looks like this

Note that this master clipper isn´t clipping anything, I´ve set the threshold just to mark where the max peaks are hitting.

Now, if I bypass all the stuff on the drum bus (the saturation and the two clippers), and let only this master clipper (the one that wasn´t doing anything) active, it will sound like this and look like this.

If you compare the last clip with the previous one you will notice that they sound the same. That means that I´m doing all that shit of reducing the drums peaks by steps for nothing. It sounds just like if I hit them with a digital machete in one single shot.

If that´s wasn´t frustrating enough, there´s still a lot of peaks left on the drums. If I hit the master bus like this, it still will only work on the drum peaks. It would be the same as adding the compressor on the drum bus instead of the master bus. It won´t do shit to glue the mix together. In order to make it work only 2-4 dbs on the overall mix, I would have to get ride of at least more 5db of peaks, and you can only imagine how it would sound.

So, yeah, here it is:

I´m clearly doing something very wrong. The clipping and saturation tricks aren´t working and I´m not sure where I´m screwing this up. What tools are you using to tame the drums before the master bus, and how are you doing it? Thanks in advance.

 

[Jordon]

Personally, when mixing, I only like the sound of the master bus compressor taming the drum peaks. IMHO, 2-4db of gain reduction on the master bus would sound a bit too crushed to my ears. I usually go for 2-3 db of gr based on kick and snare hits on the meter.

Sorry if that doesn't make sense, my brain is not functioning at a high level today.

 

[pikachu69]

The clips do sound different to me. The overall tone is there but the transient has lost its attack.

I have always found the transient peaks in slate to be really unpredictable. If you have drummagog or equivalent and gclip open together you can track which samples have the peak causing the most trouble and mute them.
The saturation and drum master buss comp are all fine, try only one gclip taking off the most extreme peaks to glue the kick and snare together then into a comp to help glue close mics to overhang, round 3db reduction.

You will still end up with a result that 'looks' the same, however. I think the issue may lie with the balance of the rest of your mix. rather than trying to bring the drums down to match the mix, try bringing up the rest of the mix to match the drums. this is more like expanding rather than compression and it will have no negative effect on the audio. This will also help you to get a higher average RMS on your master buss, and you will also find that 1 - 2db comp on the master buss is enough.

Cheers.

 

[amarshism]

I've been favoring the Michael brauer technique lately

 

[the decline]

BRAUERIZE

 

[narcossintese]

I've been favoring the Michael brauer technique lately

I´m not familiar with this technique. Is there a thread here explainning it?

try only one gclip taking off the most extreme peaks to glue the kick and snare together then into a comp to help glue close mics to overhang, round 3db reduction.

I usually do that. I use a gclip on the drum bus setting the threshold to match the highest kick peak, so that all snares peak just as much as the kick. But the kick and snare still represents much more peaks than the rest of the drums so, if I apply a compressor on the drum bus, it would still be working exclusively on the snare/kick. The same thing will happen if I put this compressor on the master bus. That is my main problem. The compressor that was supposed to be glueing everything together is only working on the kick and snare. And, to be honest, in a bad way. My master bus fx are basically doing this: leave everything untouched and just crush the kick and snare until it fits the rest of the mix in terms of level.

What I´m doing to consistently reach between -10 and -9 RMS is clipping a lot of the snare peaks, using the masterbus to take like 4db of drums and then slamming the maximizer really hard. Like 5 or more db reduction until it reaches the desired RMS. The single reason that I can get away doing this is because I use TLs Maximizer, that is ridiculously good to inject insane ammounts of loudness without completely destroying the mix.

From where I stand, the only viable way I see to tame the drum peaks before they hit the master bus compressor, would be to put a clipper, a 2bus compressor and a fuckin brickwall limiter on the drum bus... and that doesn´t sound right It is like importing a finished song (that happens to be only drums) on my project and putting it together with the rest of the mix.

 

[Fragle]

"If that´s wasn´t frustrating enough, there´s still a lot of peaks left on the drums. If I hit the master bus like this, it still will only work on the drum peaks. It would be the same as adding the compressor on the drum bus instead of the master bus. "

the peaks you're hearing (sounds like you are more like watching them in gclip though...trust your ears a little more!) come from the transient punch of the drums. that's actually something DESIREABLE. drums are supposed to punch through the mix. i guess the reason why you think otherwise is that the loudness war effectively reduces transient punch in favor of a louder mix, and you're probably used to smashed to hell recordings.

but to answer your master bus compression question, no, a few dbs of gain reduction on the master bus is NOT the same as on the drum bus alone. on the master bus, every time the compressor detects a (drum) transient it pulls down the WHOLE mix, not just the drums. that's part of the thing you described as glue, and the reason why you should use it in moderation unless you know your shit. overcompressing leads to a pumpy mix unless you know your way around it.

if you have a transient that really fucks up the master bus compressor (most of the time it will be the snare), you can try to clip the snare (not the whole drum bus) to chop off some of the highest peaks without noticeably reducing the transient punch. if it's not fucking up the 2bus comp, don't do it though.

 

[amarshism]

I´m not familiar with this technique. Is there a thread here explainning it?

Rather than having one master bus he splits them up over different bandwidths. I don't do it exactly that way, but have been experimenting with putting comps across different combined busses. You can read about it in detail on his websites FAQ.

 

[s34nsm411]

Rather than having one master bus he splits them up over different bandwidths. I don't do it exactly that way, but have been experimenting with putting comps across different combined busses. You can read about it in detail on his websites FAQ.

ive read him talk about how this is impossible to implement ITB due to imperfect latency compensation. an analog signal chain is apparently the only way to do this right

 

[xFkx]

gcliping drums is pointless. When the mix is just right everything falls into its place and glues nicely under the 2bus compression.

A little transient ehancement on the other hand can be quite nice for the little bit of extra punch.


What is is with you guys and making flat as a pancake mixes ?

 

[Ermz]

fitz... have you hijacked xFKx's account???

 

[Emdprodukt]

very interesting thread. I use gclip less and less because I didn't like what it did to my snare and kickdrum.

 

[crillemannen]

I dont have any problems with drum transient peaking. It should peak a little bit above the mix (the snare) before i send it to mastering. I don't use gclip on snare just a fair amount (to say the least haha) of compression. I think that when you are getting better and better at mixing those "problems" will eliminate themselfs. So just strife to get better at mixing

 

[crillemannen]

and Michael Brauer link: http://www.mbrauer.com/qna2.asp .

 

[drew_drummer]

This should be renamed to "Killing dynamics! (¯\(°_o)/¯ content)"

Sorry.. not into it at all.

 

[Ermz]

Transients ≠ dynamics.

 

[drew_drummer]

Squashing =/= taming

 

[Pedro Teixeira]

Gclip ≠ Good

Seriously... Gclip sucks balls, dont use that shit, I never need to use it to bring up level on anything, dont understand why should anyone need to use it. It sounds like rubbish

 

[NSGUITAR]

Gclip ≠ Good

Seriously... Gclip sucks balls, dont use that shit, I never need to use it to bring up level on anything, dont understand why should anyone need to use it. It sounds like rubbish

Weird.. I've been using it my past two mixes on individual drums, and it works great. What's wrong with it?

 

[Pedro Teixeira]

There's nothing wrong with it apparently, I just don't like what it does, I've never had trouble bringing up levels without Gclip and when I used it I just didnt like it