Gain Staging 101: The Foundation of a Great Mix

Most mixes I get through RoastYourMix have the same problem, and it has nothing to do with plugin choices, arrangement, or even talent. The levels are wrong before a single fader move happens. Gain staging. The most boring, unsexy, absolutely critical part of mixing that almost nobody does correctly.

I am not going to pretend this is some advanced secret. It is not. But the fact that I keep seeing the same issues in mix after mix tells me that people either skip it, misunderstand it, or think it does not apply to them because they are working "in the box." It applies to everyone. It applies to you.

Let me break this down properly.

---

What Gain Staging Actually Is (And Is Not)

Gain staging is managing the level of your audio signal at every single point in the chain. From the moment sound hits your microphone through every plugin, every bus, every send, all the way to your master output. Every one of those stages has an optimal operating level. When you respect those levels, everything works as intended. When you do not, problems stack up in ways that are incredibly hard to diagnose later.

Here is what gain staging is not: it is not just "turn everything down." I see people interpret gain staging advice as "make everything quiet" and then wonder why their mix sounds thin and lifeless. That is not the point. The point is control. You want every element in your signal flow operating in its sweet spot, not too hot, not too cold.

Think of it like water pressure in a building. If the pressure is too high at the source, every faucet in the building is going to blast water everywhere. If it is too low, nothing works on the upper floors. Gain staging is setting the pressure correctly at the source and at every junction point so that everything downstream behaves predictably.

---

Why This Matters More Than You Think

In my studio, I work in Cubase with an Apollo x6 interface, and I run a lot of plugin processing on most channels. We are talking EQ, compression, saturation, sometimes parallel chains, sometimes four or five plugins deep on a single track. If my gain staging is off at the input, every single one of those plugins is working with a signal it was not designed to handle. The errors compound. By the time that signal hits my mix bus, it is a mess that no amount of fader riding can fix.

Here is what bad gain staging actually sounds like in practice:

• [Clipping](/fix/clipping) you cannot hear immediately — not the obvious digital crackle, but subtle distortion that makes vocals sound harsh and drums sound papery. You keep reaching for a de-esser or a transient shaper when the real problem is that your signal was too hot three plugins ago.

• Muddy low end — accumulated noise and distortion products pile up in the low-mids. You end up cutting 200-400 Hz on every single track trying to clean up a problem that started with levels, not frequency content.

• Compressors acting unpredictably — your compressor threshold is set for one level, but the signal coming in is 6 dB hotter than what you intended. Now your ratio, attack, and release settings are all wrong because the compressor is reacting to a completely different dynamic range than you planned for.

• Master bus that is already peaking before you even start mixing — if your individual tracks are all running hot, the sum on the master bus is going to be way over zero before you have made a single creative decision. Now you are mixing defensively, pulling everything down, instead of building something up.

Warning: If you are constantly pulling your master fader down to avoid clipping, you do not have a master bus problem. You have a gain staging problem on every track feeding into it. The master fader is not a solution. It is a symptom indicator.

---

The -18 dBFS Number: What It Actually Means

You have probably heard that you should aim for -18 dBFS on your channel meters. This number gets thrown around constantly, and while it is a useful guideline, most people do not understand why it exists or when it actually matters.

Here is the history. In the analog world, professional equipment was calibrated so that 0 VU on the meters corresponded to a specific voltage level (+4 dBu in most pro gear). Engineers designed circuits — and later, plugin emulations — to behave optimally around that reference point. When we moved to digital, someone had to map that analog reference to a digital scale. The rough consensus landed on -18 dBFS as the digital equivalent of 0 VU.

So when people say "aim for -18 dBFS," what they really mean is: feed your plugins a signal level that corresponds to where the original analog gear (or its emulation) was designed to operate.

Now here is where it gets more nuanced.

Not all plugins care. A stock Cubase EQ, a FabFilter Pro-Q, or most modern "clean" digital plugins will behave identically whether you feed them a signal at -18 dBFS or -6 dBFS. They are linear. They do not have a sweet spot in the same way analog-modeled plugins do.

Analog-modeled plugins absolutely care. If you are running UAD emulations, Slate Virtual Mix Rack, Plugin Alliance stuff like the Lindell or Black Box models, or Waves API/SSL/Neve emulations — these are designed to respond to input level the same way the original hardware did. Feed them a signal that is way too hot and you are overdriving the virtual circuit in ways the designer did not intend. Feed them a signal that is too quiet and you are not engaging the harmonic character that makes those plugins worth using in the first place.

Tip: Here is my practical approach. I do not obsess over hitting exactly -18 dBFS on every track. Instead, I use a simple rule: peaks between -12 and -10 dBFS, with the average (RMS) sitting somewhere around -18 to -16 dBFS. This gives me plenty of headroom while keeping signals in the range where my plugin chain behaves the way I expect. For tracks that are going through heavy analog emulation processing, I pay closer attention. For a clean digital EQ followed by a transparent compressor, I am less strict.

Do not let the -18 dBFS thing become a religion. It is a guideline, not a law. The actual law is: know what level your plugins expect and give them that level.

---

How Plugins Misbehave (And Why You Blame the Wrong Thing)

This is the part that I really want to drill into because I see the consequences of this in almost every mix that comes through for a roast.

Someone will send me a mix and say, "I do not know why my vocals sound harsh, I have tried three different de-essers." I pull up the session or listen carefully and the problem is obvious — the vocal was recorded peaking at -3 dBFS, then run through a chain of four plugins with no gain compensation anywhere. By the time the signal hits that de-esser, it is so distorted from being slammed through a compressor and a saturation plugin at the wrong level that no de-esser in the world is going to fix it.

Here is how specific plugin types misbehave with wrong input levels:

[Compressors](/learn/compression) are the most sensitive. A compressor's entire behavior — how much it compresses, how it shapes transients, how it colors the tone — depends on the relationship between the input signal and the threshold. If your input is too hot, the compressor grabs everything too hard. If it is too quiet, the compressor barely engages. Either way, you are not getting the compression you dialed in. You are getting whatever the compressor decides to do with the wrong input level, and then you are fighting it with attack and release adjustments that should not be necessary.

[Saturation](/learn/saturation) and tape emulations respond to input level by design. That is literally how they work — louder signal equals more harmonic distortion. This can be a creative tool when you control it intentionally. But when your signal is already 6 dB hotter than it should be because of poor gain staging upstream, you are getting distortion you did not ask for. And it does not sound like "warm analog character." It sounds like your mix was recorded through a phone.

EQs with analog modeling will add different harmonic content depending on input level. Boosting 3 kHz on a Pultec emulation with a properly staged signal sounds musical. Boosting 3 kHz with a signal that is slamming the input sounds brittle and aggressive. Same plugin, same settings, completely different result.

Reverbs and delays are often the last thing people think about with gain staging, but send levels matter enormously. If you are sending a signal that is too hot into a reverb, the reverb algorithm itself can start producing artifacts, especially with algorithmic reverbs. More importantly, your wet/dry balance is now off because the send level is not where you intended it to be.

Mistake: Stacking plugins without checking levels between each one. I cannot stress this enough. Every time you add a plugin, you should know what level is going in and what level is coming out. Most good plugins have output gain or a mix knob. Use them. An EQ boost of 4 dB at 3 kHz means your signal is now 4 dB hotter at that frequency. If the next plugin in the chain is a compressor, that 4 dB boost just changed how the compressor responds to the entire signal. This is not theoretical — this is the reason your mixes sound "off" and you cannot figure out why.

---

Gain Staging and the Mix Bus

Here is where everything comes together, literally. Your mix bus is the sum of every track in your session. If you have 40 tracks and each one is running 3 dB hotter than it should be, that is a lot of extra energy hitting your mix bus. Even if no individual track is clipping, the combined signal can easily be pushing well above 0 dBFS on the master.

I aim for my mix bus to peak somewhere around -6 to -3 dBFS before any master bus processing. This gives me headroom to work with mix bus compression, EQ, or whatever I am doing on the master chain, without immediately running into the ceiling.

Some people put a limiter on the mix bus early and just let it catch everything. I think that is a terrible habit. You are training your ears to a limited, squashed version of your mix from the very beginning. You are making every decision through a lens of distortion. And when you finally pull that limiter off to send the mix for mastering, you realize the actual balance is nothing like what you thought.

Tip: If you want to monitor louder without ruining your gain staging, turn up your monitor controller or your interface output. Do not push the master fader or slap a limiter on the mix bus to make things louder while you work. Loudness is a monitoring problem, not a mixing problem. My Apollo x6 has a monitor knob for exactly this reason.

The mix bus is also where gain staging problems become impossible to separate. If your low end is rumbling out of control on the master, is it because of your kick? Your bass? Your low-mid buildup on guitars? When the levels are all over the place, every track is contributing to the problem in a different way, and you end up playing whack-a-mole with EQ cuts across twenty tracks instead of fixing the actual source.

---

My Practical Gain Staging Workflow

Here is exactly what I do in Cubase when I start a mix. This is not theoretical. This is what happens every session.

Step 1: Normalize nothing, clip-gain everything. Before I touch a plugin, I go through every track and use clip gain (or a simple gain plugin at the top of the insert chain) to get each track's peaks sitting around -12 to -10 dBFS. I do not normalize because normalization is a destructive, all-or-nothing process. Clip gain is non-destructive and precise.

Step 2: Check the sum. After clip-gaining every track, I hit play and look at my master bus meter. With all faders at unity, I want the master peaking somewhere around -6 dBFS or lower. If it is hotter than that, I pull individual tracks down proportionally. I do not touch the master fader.

Step 3: Insert plugins with gain awareness. When I start adding processing, I follow a simple rule: the signal level going out of each plugin should be roughly the same as the level going in, unless I am intentionally changing it. This means after an EQ boost, I compensate with the output gain. After a compressor reduces dynamics, I use the makeup gain to bring the level back to where it was. After a saturation plugin adds energy, I trim the output.

Most of the plugins I use — FabFilter Pro-Q, FabFilter Pro-C, UAD emulations, Plugin Alliance stuff, TDR Nova — all have output gain controls. Use them. This is what they are there for.

Step 4: Check the master bus again after processing. Once I have plugins on most tracks, I check the master bus level again. Processing changes things. Compression, especially, can raise the average level significantly even if peaks stay the same. If my master bus is now hotter than I want, I go back and adjust individual output gains rather than pulling the master fader down.

Step 5: Gain stage the buses. If I am using group buses (drums bus, vocals bus, guitars bus — which I always do), I apply the same discipline there. The signal hitting the bus should be at a reasonable level. The processing on the bus should be gain-compensated. The output of the bus feeding the master should be controlled.

---

Things Nobody Tells You About Gain Staging

Gain staging is not a one-time setup task. Your levels change as you mix. You automate volume, you add plugins, you mute and unmute tracks, you change arrangements. Gain staging is something you monitor throughout the entire mixing process. I check my levels multiple times per session, especially after major changes.

Your meters lie (sometimes). Peak meters and RMS meters tell different stories. A track can have modest peak levels but a very high RMS (average) level, which means it is contributing a lot of energy to the mix bus even though it does not "look" hot on the peak meter. I use both. RMS tells me about perceived loudness and energy contribution. Peaks tell me about headroom and transient management.

Gain staging applies to effects returns too. Your reverb returns, delay returns, parallel compression buses — all of these feed into the master bus. All of them need to be gain staged. I have seen mixes where someone has a parallel drum bus running absurdly hot because they never adjusted the return level, and it is adding 6 dB of uncontrolled energy to the master.

Warning: Be very careful with gain staging when you are using sidechain compression or dynamic EQ. These processors react to the level of the input signal (or the sidechain signal). If your gain staging changes upstream, the behavior of every dynamics processor downstream changes with it. This is one of the most common reasons a mix "breaks" when you go back and adjust levels — you have inadvertently changed how every compressor and gate in the session is responding.

---

Checklist: Gain Staging Before You Start Mixing

Before you touch a single plugin, run through this:

• Every track's peaks are sitting between -12 and -10 dBFS (use clip gain or a gain plugin)
• With all faders at unity, the master bus peaks at -6 dBFS or lower
• No individual track is clipping (check every track, including effects returns)
• You know where your metering plugins are and you plan to use them
• Your monitor volume is set for comfortable listening — you are not compensating for quiet tracks by pushing levels

Checklist: Gain Staging During the Mix

• After each plugin insertion, output level roughly matches input level (unless intentional)
• Group buses are not clipping or running excessively hot
• Master bus stays below -3 dBFS peaks during the loudest section
• Parallel processing returns are gain-matched to their contribution, not just blasted in
• Any automation changes have not broken the gain structure downstream

---

The Honest Truth

Gain staging is not exciting. It does not have the instant gratification of slapping a new compressor on a vocal or finding the perfect reverb. There is no YouTube thumbnail that will make gain staging look cool. But after years of working with artists and hearing hundreds of mixes — both my own and through RoastYourMix — I can tell you that bad gain staging is the single most common technical problem in home studio mixes. Not bad EQ decisions. Not wrong compressor choices. Levels.

The good news is that it is also the easiest problem to fix. It costs nothing. It requires no new plugins. It just requires discipline and attention. Spend thirty minutes at the start of every mix getting your levels right, and everything that follows will be easier, sound better, and cause you less frustration.

If you are not sure whether your gain staging is solid, or if your mixes sound "off" and you cannot figure out why, submit your track at RoastYourMix. I will tell you exactly where the problems are — gain staging or otherwise. No sugarcoating. No vague encouragement. Just honest, specific feedback you can act on immediately.