Skip to main content

I am curious to know why there has been absolutely no discussion at recording.org on chaining multiple compressors in a row for individual tracks on mixdown. Does this mean that nobody is actually doing it?

It would surprise me if I am the only one who finds themselves pounding 3 or sometimes 4 compressors in a row on lead vocals. Before this sounds like overkill to you, please keep in mind that I am doing little gain reduction (4db) on each compressor. The end result is far more transparent, huge amounts of compression is applied and it doesn't even sound compressed to the slightest.

I would think that with such promising resuts that most professionals here have tried such a technique.

The reason why I ask if anybody uses this method is because in my entire recording career I have never found a better way to make vocals or any other solo instruments sit in the mix without sounding like they've been destroyed so badly. Yeah Micing techniques will take you far if you have Whitney Houston singing in your studio but what about those uncontrollable chicks who scream the loud parts out and whisper the soft parts? There must be more of these chicks in the world than the exceptional singers like Mariah Carey so I assume that I am not the only one who is struggling with over-dynamic singers.

What do you engineers do when you get a chick/dude that has a nice voice but doesn't seem to penetrate through the mix without a ton of compression? I am assuming that if you don't use multiple compressors then you are at least using around 15-20db of gain reduction with at least a 4:1 ratio on mixdown. The problem is, I cannot hear that setting sounding realistic or reasonable on any compressor in the world. I've tried 20db gain reduction on a 1176 and an LA-2A and the problem is you cannot do any more than around 7db gain reduction without these compressors making the signal sound extremely thin.

Sure, gain-riding will help as well but some singers are way too dynamic even within words let alone phrases. I cannot possibly be quick enough to pull down the volume of a take where the singer went really loud on "ver" in the word "over".

Now even the most advanced compressors have their limits. What the heck do you do when you need 20db of gain reduction?

Antonio

Comments

Kev Fri, 06/14/2002 - 14:05

Just like guitar pedals, I swap them around all the time for different tasks.

Comp then Limit, even Limit then Comp and often Comp , EQ , Limit.

Same on mix down. When I'm keen I'll use an eq in the side chain and remove the bottom end so it doesn't catch the compressor.

But Like you I tend to do the comp in smaller bites rather then a big crunch .... but sometimes the crunch is just right. ;)

Antônio Sun, 06/16/2002 - 21:12

Cool, so I'm not alone.

To all the RNC users, you probably know this but "super nice" mode already works in a "chaining" fashion. It chains 3 internal RNC compressors together and this is primarily why the output is as clear and transparent as everyone says.

Bear, you are literally chaining four compressors together in your setup and I think that is way cool. You're doing whatever it takes to make things sound right and that's the way it should be.

I think the main reason why RNC sounds so good is because of the "chaining" mode (super nice). It should be a "well" known published fact that chaining multiple compressors together each with small gain reduction (4db) results in a far more "less obvious" compressed sound. There must be a huge amount of people out there trying to figure out why they cannot achieve 20db of gain reduction without their signals sounding mangled.

It was once impossible to have so many hardware compressors at one's disposal but with the ever-growing computer DAW specs, one can "freely" open up 3 compressor plugins on their sequencer without sacrificing much CPU at all. Plugins usually reveal their poor sound when you start hitting around 8db of gain reduction so by minimizing the gain reduction on each unit to around 4-6db you can get a brilliant sound out of software these days.

Question: How much gain reduction do you guys find yourself applying to your average lead vocal? I know it depends and all but gone are the days where engineers are restricting their one and only vocal compressors to no more than 3db of gain reduction. On todays mixes I hear WAY more. The only thing is that I can't actually hear the compressors "working". This is why I am convinced that they are employing methods such as "multiband" compression or "chained" single-band compressors on lead vocals. So when you guys chain your compressors, how much gain reduction is usually occuring in each compressor? When you guys don't chain compressors, how hard are you pushing your single-band compressor?

This is all interesting stuff.

Antonio

anonymous Tue, 07/16/2002 - 04:04

Yep, I've been using chained compressors. For voices in commercials - improves the subjective loudness without sounding "too compressed". I used a first stage of low Threshold-low Ratio compression, followed by a high threshold-high ratio stage (roughly).
Another approach is the use of feed-forward or look-ahead compressors (in conjunction with delays). There's also quite an un-obvious compression possible.

Matthias

KurtFoster Tue, 07/16/2002 - 07:49

Oh Yeah! That's the ticket.. In my opinion the trick is to squeeze a little out to tape / disk whatever and then a little more at mix, reducing real world dynamics to a more recordable / listenable range in small steps. Sometimes I'll use a comp limiter combo like a LA2a or LA3 and then into the 1178 at 20 to 1. One way to comp/limit a lot without hearing it so much is to do a sidechain eq in the 1K to 3.5k range, to squish out the mids while leaving the lows and highs untouched. As with all of this stuff, try it, if it works it's ok. Just don't plug a speaker out into a amp input (ha ha ha ha ) Sounds like your on the right track. Good luck, Fats

RecorderMan Tue, 07/30/2002 - 08:07

Originally posted by crazy_guitar:
How do you use a compressor/eq in side-chain mode to compress a frequency???
thanks

bus or multi the signal in question through an eq, then into the side-chain input of the compressor in question (if that compressor has that function...most do). Boost the frequency in question/desired. A DeEsser is the simplest/most common version/variation of this technique.