Gain Staging Demystified

There is a moment every producer knows: the mix that sounded fine at low volume suddenly turns to mush when you push the master fader. Everything is loud, nothing is clear, and adding EQ or compression seems to make it worse rather than better. The culprit, almost always, is gain staging — or the lack of it.

Gain staging is simply the practice of managing the signal level at each step of your signal chain. Not just the master output, not just the final limiter, but every fader, every plugin, every bus, from the moment audio enters your session to the moment it leaves. The idea is that each stage of processing works best within a certain range, and when you ignore that range, problems compound quietly until they become impossible to fix.

The rule itself is almost embarrassingly simple: aim for an average level of around -18dBFS on each channel before it hits any processing. That number is not arbitrary. It corresponds roughly to 0VU on analogue equipment — the level at which analogue gear was designed to operate, where transformers saturate pleasantly, where compressors respond musically, where the signal has presence without strain. Plugin developers, particularly those modelling vintage hardware, calibrate their internal processing with this reference in mind. Feed a channel strip plugin a signal peaking at -6dBFS and you are essentially overdriving the input stage before it has had a chance to do anything useful.

The practical implication is headroom. At -18dBFS average, your peaks might reach -10dBFS or so, leaving eight to ten decibels of clean space above them. That space matters enormously. Transients — the attack of a snare, the pluck of a guitar string — contain enormous short-term energy that sits far above the average level. If your average is already at -6dBFS, those transients are clipping or forcing limiters into action on a channel that should be completely unprocessed at that point. The transients get crushed, the attack disappears, and the mix loses dimension.

Why Plugins Care

Most DAWs operate at 32-bit or 64-bit floating point internally, which means the signal path itself will not clip no matter how high your levels climb — the maths simply keeps working. This leads many producers to assume gain staging is a relic of the analogue era, irrelevant now. It is not, for one specific reason: plugins do not always operate at floating point. Many — especially emulations of analogue hardware — have fixed internal operating levels baked into their design. A compressor modelled on an SSL channel strip has a sweet spot. A tape emulation starts to saturate at a level that corresponds to the calibration point of the original machine. When you feed these plugins a signal that is 12dB too hot, you are not getting 12dB more of character. You are getting distortion, pumping, and a response curve that bears no relation to what the developer intended.

Even purely digital plugins — modern EQs, linear phase processors, surgical tools — can exhibit increased noise or numerical errors at extreme levels. More practically, a compressor set to a ratio of 4:1 with a threshold at -20dBFS behaves completely differently depending on whether your input signal is hovering around -6dBFS or -18dBFS. Gain staging is what makes your plugin settings meaningful and repeatable across sessions.

How to Do It

The fix is unglamorous. Before you add a single plugin to a channel, pull the fader or trim the clip gain until the signal is averaging around -18dBFS. Most DAWs have a trim or clip gain function that adjusts the level before the plugin chain begins — use that rather than the fader, which you want free to balance the mix later. Set your meters to display RMS or an integrated loudness reading rather than just peak, since peak meters tell you about momentary spikes rather than average energy, and average energy is what you are trying to control.

Do this at every stage. After a bus of drums has been processed and is being fed into the master, check the level again. A compressor on a bus can add 6dB of gain simply through its makeup gain control. That gain needs to be managed before the signal moves downstream. It takes fifteen minutes on a session you have already built, and it is the kind of work that feels tedious right up until you realise your mix is translating clearly to every speaker system you try.

The irony of gain staging is that its effect is most audible in its absence. A well-staged mix does not announce itself — it just sounds like a mix where everything is working. The compression grabs correctly. The reverb sits where you put it. The limiter on the master barely moves, because it does not need to. Somewhere in the chain, you made a series of small, careful decisions about level, and the rest of the signal followed obediently.

That is all it is. Not a secret, not a technique reserved for engineers with decades of experience — just the discipline of feeding each part of your chain what it was designed to receive.

The Resonillator Mixing Guide walks through the key stages of a mix with this approach built in from the start.