Gain structure

[joel hamilton]

A finer point of recording I have been thinking about.

Gain structure is more important than EQ, or how many plugins, or how many compressors, or how much your vocal mic costs.

Gain structure is not how hard you hit tape, or how much you crank the knob on the pre into your DAW.

Gain Structure is a decision making process that starts with the sound source, and doent stop until the mix is mastered, or at least finished.

Making a decision about how, when, why, and with what device you either apply gain or attenuation is gain structure. At what point do you turn the snare down? with what? do you record it quiet, and leave the fader alone all the way through to the final mix? do you capture the snare really hot, then compress a bit and return it to the console quieter than the original capture? why? What do the two yeild as far as the general qualities you are looking for?

Do you record a vocal really hot, then squeeze it down with one massive compression stage, or do you layer the responses of a few compressors that "hear" different chunks of the spectrum, and key off of different sounds from the source?

Think about your gain structure.

[cgarges]

All of the above.

And I love gear with a hard bypass so I can hear exactly what's going on.

Chris Garges
Charlotte, NC

[I'm Painting Again]

heres a topic with totally underestimated importance..the amplifiers in your signal play a huge role in tone..lately i have been digging my 60's tube spring reverb in the signal path with the reverb turned all the way down..adds its own particular beautiful and organic tube amp stage sound as well as influencing my signal with its unique eq curve..makes guitar sit really well and makes the tone a bit more organic and etherial..every piece of gear will impart its own eq on a signal..words of wisdom from Joel (as usuall)..

[swingdoc]

Joel, so many good questions.....but wheres the answers??

Are there any "pearls" or generalties you can offer regarding this topic?
So often, people wanna hit the tape or mic pre hard to get it "phat". I kinda see the mic pre / mixdown relationship similar to a guitar preamp / amp relationship. ie premp hot, amp less so = choked, distorted.
preamp less hot, amp up = cleaner, more dynamic. (analogy being: "preamp" = mic pre; "amp" = mix buss)
Is this at all fair? ie if all your signals recorded are maxed out at the pre, then trying to squeeze everything through the mixdown buss = "choked, boxy, less dynamic"

This has been my experience anyway. I kinda try to plot the song, and pick which couple channels I want to run hot up front (hard hit pre)...these ones I want less dynamics on and that I want to sound "murkier/darker". Then back off on the others that I want more clarity and ability to move in and out (dynamic).

It kinda seems that typical ideas about gain structure are referenced as a single channel only, without a notion to their relationship to each other and of the entire song or the whole mix.

Obviously the concept of gain structure when combined with all the different gear from mic to ears, has infinite variability, but what are some basics??

Gracias Joel..

[Professor]

Well, since I do lots of acoustic music, whether classical, jazz or rock, I tend to err on the side of getting more dynamic range recorded and manipulating that later. So I tend to to have the first step in amplication taking all of the heat in tracking - with everything else set at -0- all the way to the hard disk. Ideally, I want all the gain to come from a circuit that actually has a 'gain-stage', so I amplify at the preamp "gain" and cut with the "fader" - as nature intended. The preamp is adjusted so the loudest material the performer might deliver does not clip while trusting that my noise floor is sufficiently low enough (and it is) that their softest materials will still be captured. Naturally, I try to maximize my range, so I skirt very close to the edge of clipping.

In mix down, many things happen that could be gain-staging, though are not always taught as such. I think the most important of these steps is the 'strip silence' step.
I hardly ever use the automated 'strip silence' tool in PT as I prefer to go through each track manually and cut the parts where the instrument is not playing, so I know I am actually cutting and preserving what I want. For example, I wouldn't want to lose a ghost note on a snare mic, but I wouldn't want the ambient bleed of a tom hit staying in.
This could all be done automatically by a noise gate, or manually by moving faders, etc. In the end, it is part of gain staging, because it allows what remains within the track to be pulled up without bringing along the unwanted noise. It also cleans the tracks, and makes the combined tracks feel more tight, well-defined and clean - and my favorite description from the techno-phobes, 'natural'.

OK, so that's my typical gain staging on the way into "tape" and while on "tape".

As for mix down, that will vary from project to project, though I again try only to amplify with gain stages. Since I mix off of PT and through a DM-2000 console, I don't hit another line-amp stage so all of the gain is in the digital world. But even here I use the gain in a compressor plug-in rather than pushing the channel fader higher in the box or on the board. Many times I'll use a compressor like an LA2A or 1176 plug in set so that it is only operating as a gain stage - not holding back any peaks, just bringing up an otherwise quiet performance. When the compressors are used for dynamics control, I prefer lower settings and multiple layers of compression. On a vocal it might be compressor, then EQ, then compressor inside the box and then another compressor then EQ on the console before hitting the mix bus and any mix-bus compression, and mastering-type processing on the way to & through the CDR.
The results I get are generally a very detailed image with dynamic contrast, movement, and space - the sensation of sitting very close to the live performers inside a large space.. At least, that's what I aim for with the primarily acoustic projects.

That's my approach, but I'm curious to read what others do.

-Jeremy

[vernier]

Do you record a vocal really hot, then squeeze it down with one massive compression stage, or do you layer the responses of a few compressors that "hear" different chunks of the spectrum, and key off of different sounds from the source?

No compressor here (when possible) and never multible. For gain structure, I keep it simple, *one* knob or *no* knob.

[joel hamilton]

As I learn (which is every day) I find myself EQ'ing less, and getting tones from the choice of amplification, or attenuation during every stage of the process.

Rather than simply relying on a pre for simply gain, choose for inherent response. My ampex 601, with its slow, rising response times almost "de-ess" a sibilant singer, or mellow out cymbals in a room mic. The Manley tube reference pre brings a glistening presence to an otherwise one dimensional mono acoustic guitar using a simple SM81.... Stacking up these decisions starts to sound really good to me, with no EQ anywhere. I find myself using compression and staging the gain to bring out the thwack of the inside kick mic, or the roar of a fender champ turned all the way up, starting with a BK5 (rca ribbon) on the grill of the little guy, and cranking that to tape to give an even greater illusion of "loud" then a little 2254 compression across the channel off tape, pre fader, returned to the board hot, then attenuated with the fader "as nature intended."

No Eq. Just tons of gain manipulation at every stage, with the right amplifier or attenuator.

I am constantly amazed at the "perceived frequency response" of any given sound relative to the other sounds in a mixing situation. A kick drum that is too quiet at the fader seems to "want" a little 4 or 5k to poke out of the mix for definition, then it seems to scream for some 60hz as well, and you start pulling the silly putty image of your original sound until it no longer looks like a kick drum...

Both the "thwack" and the "thump" of the kick would come up in a perfect, non-headroom eating way WITHOUT these moves if you simply turned it up, or turned everything else down. Amazing. I have been using a 33609 lately for inside kick drum, and compressing it a good amount, and simply returning it really hot, and attenuating at the fader if need be. No EQ, just tons of gain and a correct frequency balance from low to high. The 2mix meters barely move, even when the thing is kicking my ass. Amazing.

Same deal with the bass.

One great trick I have learned is to get the faders happening with the control room monitors CRANKED up, so the kick sounds loud and huge even when the mix bus isnt working very hard at all. That is part of the gain structuring I use to get everything to sound loud. I am not listening loud, I am simply leaving headroom at the mixbus by giving a reasonable control room volume without the need to hit the mix bus hard early in a mix. I couldnt believe how much easier this made my job when I had 87 tracks handed to me, tracked by someone else, and I am being asked to make them ALL work together!

In that case, it wasnt about all the "cutting certain frequencies to leave room for the vocal" type of BS. This is about making things sum together the way they can and will in a great acoustic space.

I like to think of it this way:

If I was handed a tennis ball, and was asked to make it hit something really hard, I wouldnt start changing the tennis ball, I would use a device that is designed to move the original item really fast, in this case a catapult. The tennis ball remains unchanged, but is moved rapidly and hits the wall really hard, with more force than was possible with just your arms.

Take a kick drum, dont modify it, just throw it at the speakers... REALLY HARD!!!

These are just thoughts and observations that I hope inspire your own...

[Disasteradio]

If I was handed a tennis ball...

I'm gonna be thinking about that all goddamn week. best analogy ever.

[b3groover]

Take a kick drum, dont modify it, just throw it at the speakers... REALLY HARD!!!

I don't know... what if you're looking for a really bad-ass, in your face kick sound and the heads suck or the drum is just too small? Then you're going to need a new tennis ball.

[joel hamilton]

I have found that almost regardless of the source, including dead drums, that EQ is not immediately the answer. Use something with a proximity bump for some low end on the front head, and something super close to the beater for the inside, and then impart whatever attack and decay response you are looking for with a compressor, then gain the hell out of it. If you are willing to embrace the unique sound of the heads that you are presented with, you can make a really killer sound with a really fucked up drum. Even something with three pillows in it, and only one head, and rattling hardware can be made to work for you in the context of the song. Of course if you are trying to copy someone elses idea of the "perfect" kick drum sound, and you are starting with drums and a drummer that are "inferior" to the ideal model you are looking to strive for... Of course you will be overwhelmed by the number of operations required to "make" a kick drum sound.

I am simply not into plastic surgery, and this is not a debate over drum tuning or maintenence. If you can capture the wonky, screwy sound of someones beat to shit kick drum, then you can capture anything. You can then urge it into a working world with too many other tecniques to mention here.

With proper gain staging, that shitty kick drum you mention should sound like it is being hit well, and played by a professional, and captured by a professional for the enjoyment of all! Listen to the kick on the neil young song "old man." that kick sound would be laughed at by someone like korn, and I kinda laugh at a lot of things about korn....

In both cases, a professional recorded it and made a deliberate choice as to how that instrument would be represented in the final mix.

I have recorded kick drums that sound like complete shit in the room that wound up getting comments about how cool and big they were on the record. Sometimes I will lay in a front head sample of a great slingerland 24" drum I have, recorded alone in the room with a FET 47 just for some tone in front of a cardboard-y kick.

None of that negates the benefits of proper gain staging throughout the process.

I do hear ya though, B3, it makes life a lot easier when people and their gear sound good in the first place (weird, right?)...

[KennyLusk]

EQ is an awesome tool when mic placement isn't getting youwhat you want or you just don't have the right mic for the job.

But I agree that proper Gain Staging is a huge factor in achieving the desired Tone and capturing the essence of the instrument.

For me, I'm not so much concerned with getting the hottest possible signal as I am with capturing the purest tones the instrument produces with acceptable S/N ratios and useable dB levels ("useable" is a broad term).

I record all acoustic instruments so my goal is always to capture as many those beautiful, swirling harmonics as possible. Gain Staging is a key element in capturing an instruments natural harmonics because unbalanced gain ratios anywhere in the signal chain emphasizes white noise, and white noise will mask the subtle harmonic frequencies in the signal every time. When the instrument (as it's being played) articulates peak sound levels with each note the ears have a tendency to focus on the sound the instrument produces at it's peak sound levels; for a drum, it's the Thump Thump our brain holds onto as we're listening. White noise is unconsciously ignored by the human brain in most acoustic environments and the same thing happens when listening to music. Even as engineers we can miss white noise in the tracking stage (however infrequently). Nonetheless, white noise produced by improper Gain Staging can and does mask harmonic frequencies integral to capturing the true character and life of the object instrument.

Tone is definitely another bonus when proper Gain Staging is achieved; I agree. Here's where I also agree that Gain Staging can be used for effect, or desired effect. Sometimes the song requires a super-hot and snappy snare or a very strong and pronounced Pop for hand percussion. And sometimes the snare needs to be a little more warm and full. I've used the -20dB pad on my CAD plenty of times for effect on vocals when the vocal track was intended to sit back in the mix with gobs of reverb, giving the illusion that the vocal is physically far away. It helps for the engineer to know where a song is going and what each instruments purpose will serve as a component of the song. it's not necessary, but it helps.

Proper Gain Staging starts with the microphone and it's principle's apply to every piece of gear in the signal chain. I also agree that it can eliminate some of the need for EQ during mixdown. This is an important topic of discussion. Nice one.

[JASIII]

Great thread, I love topics like this that make me re-evaluate my techniques and processes.

[Devlars]

Allow me to chime in and say that I am by no means a "professional" but as with most things I have very strong views about recording and A perfectionist. I hate the idea of "fixing" things, changing things, making them what they are not after you have recorded something (unless you want to make it sound unusual of course which I do at times) simply because it's not what you had wanted to hear. If thats the case steps should have been taken to make that change while tracking.

For the most part I try to record things in such a way that will allow me to stay as far away from post production processing as possible. I love being able to do a mix that consists soley of panning and fader levels being placed just so nothing more. Think of the phrase, "fix it in the mix" under ideal circumstances (those being trying to create with a perfectionist and complete attitude) then that phrase should not exist, why would you seek to do something "good enough" to then have to fix it? This seems wrong to me. If the source is not adequate or good then go to the source for tweaking not processing knobs!

I agree with Joel (as I find myself doing in response to many of his posts) rather than changing what you have after the fact, get it that way in the first place. Take your time and listen to what you are recording, if it doesn't sound right change it then and there at the source. I feel as though I'm beginning to ramble and perhaps have lost topic. So I'll say that I get my gain stage structure by getting my volume and dynamics from the way things are played or amplified getting the hottest signal possible for everything then simply using the faders to give me a pleasing mix rather than performing EQ surgery and other such things of the like.

Sorry didn't mean this to come across as a rebuke.

[nestle]

More and more I'm trying to ween myself from my analog days as far as the end of the chain. It's almost habit to get things as hot as possible, but with digital I need to revaluate that habit. I've been getting more and more conservative. I'm thinking the headroom you gain especially at the master is more important.
After really working a great mix you'll notice all the faders somewhat at unity, it's cool. In my mind its a sign that you've done something right with all of your gain stages. If everything looks up/down wacky...well you're compensating for something wrong-


I agree with all the previous posts and I think that is the art of engineering in a nutshell. How each stage of the signal flow effects the chain is what it's really all about. These days I only reach for EQ as a last resort, and hardly ever when tracking. But then again there are no rules....

[DavidATX]

Ahhhh, I see how it is....you go away to Italy and come back jivin' like Galileo. Haha, seriously, how was Italy bro?

I tend to try to hit the Pre as hard as I can...this usually means padding the mic in order to do that. I am not as technically savvy as Joel (or as I should be in general)...but it just doesnt give me the "texture" or tones I want to hear if the pre isn't cranked to get the signal. Sometimes I dont have a choice though.

For vocals, I have always used two compressors...usually a tube one (such as the avalon) first...then to a dbx. If it is heavier type of vocals I will squash the hell out of it at both compressors....but if it is more "singy" type of stuff, I will just pull about 5dB at the first and let it tickle the the lights on the second one.

For something like the snare, I want to really hear that crack..so I go pretty hot. Once its in the box, I will slightly compress it, just to kind of help out even some of the hits out.

In general though, I try my best to get the sound I am looking for through mic choice and placement. Sometimes it just doesnt work out and I need to EQ something...but everyone should know by now that the source sound has to be at least close to even consider EQ...you can't polish a turd...unless it is supposed to sound like a mess.