Two Stage Compression

[vvv]

Mr. Brauer's quoteing of Mr. Caffrey can be found here (search "Caffrey" on the page; it's the second entry, I believe.)

[mcaff]

The last thing to address is the idea of the fast attack/slow attack compression referred to twice in this thread, both one page back and more recently a few posts up. I'm not familiar with Harvey Gerst's example as referred to here, but I'm going to guess it's the technique of using one compressor for peak limiting and one for a more average program compression.

I didn't mean to imply that this technique was limited to the same application as simply chaining two compressors. It seemed that people were having a hard time grasping the routing and execution, so I thought it would help to break it down into blocks that might appear more familiar.

Many thanks for the article. It literally bailed me out on a very frustrating mix that now satisfies the band, even though they still want me to fix even more of their performance f-ups. Argh..... but, yeah, thanks!

Depending on the the specific compressors you use, it could be the same, but the resulting dynamics are very different whn you have the dry singal happening within the first as well.

I'm glad it was helpful. That seems to be the consensus, I've heard similar stories from a lot of people, especially as more time has passed and people have been able try it a little more. I know it's made things a lot better for me.

[mcaff]

Mike, again, my tone was unwarranted, and a bit reactionary. You are right. I'm sorry. I hadn't read this last post yet, and I removed the word in question, because I realized in a message to Larry that as I had misinterpreted your intention of "hack," you misinterpreted what I meant by assimilating somebody's technique without proper credit. I'll remove all references to that word in any of my posts, because you and Larry seem to be upset by it.

I will say that in my defense of why I draw that conclusion, I have to use your own words:
"Mike Caffrey is well versed in Michael Brauer's Multi-Buss compression methods. He's written the definitive Tape Op article on the technique and has demonstrated it with Michael Brauer at the Tape Op conference."

Knowing that, and how the party in question feels about it, I can't help but feel that credit was not given where credit is due. As I told Larry, it's tantamount to me saying that I wrote the definitive article on Pete Townshend's techniques, and that I'm well versed in his methods. I could not in good conscience say that, even though I've performed with him once, and did an interview that was widely acclaimed. I don't even know if Brian Kehew bills himself as an expert in Beatles recording techniques, after writing a book for fifteen years that might be the definitive text on the subject.

BTW, none of this is personal. I'm sorry my initial cantankerous tone made it so. But good friends of mine have publicly given me shit online when I've written something that they had a problem, and they've still been invited to my birthday dinner.

Thank you, that helps quite a bit. I have no problem wihtout discagreeing with me or thinking I have no clue what I'm talking about, any opinions are fine, that's very different than the unsubstantiated charge.

I definitely don't follow how i misinterpreted anything you've written. The principle that I'm trying to explain does not in anyway involve separating instruments so that they don't duck each other.

I also don't follow the words you've got in quotes. If you're saying them, why are they in quotes? If you're quoting me, quotes make sense, exacpt that is not an exact quote, so it's inaccurate to put paraphrasing in quotation marks.


As far as I can tell, the only misinterpretation going on is you misintrpreting noe one or both of the teqhcniues.

If you have a single mic, live or recorded, you can not do multi-buss compression as Brauer intends for it to be done. You could could do the "two tage" technique by sending that one mic through a specific mono daisy chain, multing and summing the singal or using busses to mult and then sum the signal. That should make it clear that there's no overlap and using a buss for a mult is not what Brauer is doing. He would describg that as double bussing which he does sometimes do, but even that is not at all the same as the "two stage" idea. Busses have more functions that Brauer's mix technique.

[JohnDavisNYC]

[lotusstudio]

WOW! Now THAT was a long, slow, excruciating attack with a HARD KNEE.

[Patrick McAnulty]

WOW! Now THAT was a long, slow, excruciating attack with a HARD KNEE.

zing!

[will.record.for.food]

I really am getting great results using this method.

I went from having 2-4 plugs PER DRUM TRACK, to pretty much (besides the sends and bus compression) only 1 or none!

For me, my drum tracks are WAAAY more balanced overall...

I'm also freeing up SO MUCH processing power on my feeble Sempron3000+ system with 512 RAM

[lotusstudio]

I still don't get it. I spent a few hours trying this and my cymbals became way too loud and strident and the overall sound muddier and crowded.

Do you use the same attack and release on both comps?

Do you turn down the faders feeding the comps?

Isn't it better to compress each drum mic separatly to use just what it needs?

What about the overheads? Should they be left out of this chain, or what?

The article says to limit the dynamic range going into comp #2 to @3-6 db. This would mean slamming the first signal really hard, right?

I just haven't been able to figure out how to make it sound good.

[Woodeye]

After enthusiastically reading this article about 5 times, I'm now finally getting around to mixing the 'space-rock' project I've been tracking to which this method will, I'm sure, apply splendidly. Now all I need to do is find a VST compressor plugin with a wet/dry control; anyone out there have any suggestions? I could pony up the $199 for the new UAD Precision Compressor, but I'm not quite comfortable enough with my skills to try and justify the expense...

[roygbiv]

Woodeye Dude -

Check out the Reaper VST-plugin package. It is free, and includes a nice compressor and a multi-band compressor with variable Wet/Dry settings. A pretty nice delay and gate are also included.

Although they're pretty "kick ass", they ain't got no fancy graphics or nuttin. You just get absolute numbers, and thus have to actually work on understanding what you are doing more. I suppose that is a good thing, although it's more fun tweaking virtual knobs.

http://reaper.fm/reaplugs/

[calaverasgrandes]

5 pages and nobody mentions that Sonar already has a two stage compressor? It called VC 64. Its got eq too. I believe the new version that came with 7 has wet dry knobs too. For shame.
for shame

ps. in the computer nerd community I run with, a "hack" is a clever and unconventional way of doing something. Nothing temporary implied. Kludge is more derogatory and definetly implies temporary. Using an HP rack set to mount a Dell server is a Hack. Running cat 5 across a a commonly used walkway while contruction is occuring is a Kludge.
What I dont get is you guys refering to using two of the same comp in 2 stage compression. I always go for a comp followed by a limiter. Or a slow limiter followed by a comp.

[vvv]

FWIW, on vocals I like to use an optical compressor (hardware, or a slow faked software one) followed by a quicker "electro" compressor.

And then I use a limiter.

[mcaff]

After enthusiastically reading this article about 5 times, I'm now finally getting around to mixing the 'space-rock' project I've been tracking to which this method will, I'm sure, apply splendidly. Now all I need to do is find a VST compressor plugin with a wet/dry control; anyone out there have any suggestions? I could pony up the $199 for the new UAD Precision Compressor, but I'm not quite comfortable enough with my skills to try and justify the expense...

Yes, route to two busses and three aux tracks. Rute the drums to busses 1-2.

Feed two auxes from 1-2.

Route the auxes to 3-4

feed and other aux from 3-4.


The last aux is your second stage.

INf those first two auxes, make one your dry buss and th other your compressed buss and you'll have wet/dry control in front of your second stage.

[mcaff]

5 pages and nobody mentions that Sonar already has a two stage compressor? It called VC 64. Its got eq too. I believe the new version that came with 7 has wet dry knobs too. For shame.
for shame

ps. in the computer nerd community I run with, a "hack" is a clever and unconventional way of doing something. Nothing temporary implied. Kludge is more derogatory and definetly implies temporary. Using an HP rack set to mount a Dell server is a Hack. Running cat 5 across a a commonly used walkway while contruction is occuring is a Kludge.
What I dont get is you guys refering to using two of the same comp in 2 stage compression. I always go for a comp followed by a limiter. Or a slow limiter followed by a comp.

They're not daisy chained normally, that's why.

[mcaff]

I still don't get it. I spent a few hours trying this and my cymbals became way too loud and strident and the overall sound muddier and crowded.

Do you use the same attack and release on both comps?

Do you turn down the faders feeding the comps?

Isn't it better to compress each drum mic separatly to use just what it needs?

What about the overheads? Should they be left out of this chain, or what?

The article says to limit the dynamic range going into comp #2 to @3-6 db. This would mean slamming the first signal really hard, right?

I just haven't been able to figure out how to make it sound good.

That's strange because if you're able to make it sound good without doing it this way it should be easy.


It doesn't say to limit the dynamic range to 3-6dB. It says that compressors dont' do as well as the possibly can if they see more. That's a very different meaning.

Set your second compressor they way you normally would, then add the earlier stage with a wet/dry compressor or by using a wet/dry buss.

That will change the gain staging, but when you've corrected it the sound coming out of the second compressor will be far more consistent. You see that there's far less difference between the peak range of meter movement and the average range of meter movement because they compressor is seeing a much less dynamically diverse signal.


I rarely work in the box, but someone brought their G5 in the other day for a mxi consulting session, where they had me adjust what they'd worked on already. They had routed things with three busses ITB, so it was easy for me to convert the routing to work for this technique.

I found I liked the 1176 as the second compressor with the attack set between 1 and 3 (slow to as slow as it goes) and the release set to 7 or a little less (the fastest).

As far as as the first compression stage, on the mix consulting session they had used a TG-1 compressor and I stuck with that in Limit 1. I added EQ in front to clean up the low mids a bit. Limit one is the fastest attack and fastest release. This was on a wet wet and there was a dry buss too. I can twll you that the wet buss faded was pulled down about 1/3rd, but that doesn't tell you much without knowing the volume within the dry buss.

When I submitted my drums for the Katy Perry's "I Kissed a Girl" I used two 1176s set pretty much identically to the way the second one was above and that sounded great.


IF you keep your dry buss hot, the attack time of the parallel wet buss will be nearly irrelevant (like a shutter speed when shooting with strobes). The attack of the parallel wet buss will always be masked by a loud dry buss. So the release time and amount of compression matters more when running that way. And, that's why parallel compression's effect is more to reduce dynamic range than anything else.


For the second compressor, normal rules apply - if you compress too fast, you'd change the envelope of the sound and make it unnatural.