Gain Staging

Setting the volume at each step in your signal's journey so it is loud enough to be clean but never so loud it overloads.

Each stage from mic to amp needs filling near -18 dBFS: too low brings hiss, too high clips at the 0 dBFS wall.

What it is

Setting the right level at every stage of the signal chain so it stays clean: loud above the noise, well below clipping.

Key facts

0 dBFS is the digital ceiling: hit it and the signal clips hard (square-waves) with no warning.
Aim for ~-18 dBFS average on a digital console = roughly 0 VU / +4 dBu = the analogue 'nominal' sweet spot.
Pro line level = +4 dBu (1.228 V RMS); consumer line level = -10 dBV (0.316 V RMS): ~11.8 dB apart.
Standard headroom target: 18-20 dB between your average level and 0 dBFS for transient peaks.
+6 dB = double the voltage (2x); -6 dB = half the voltage. +10 dB ≈ double perceived loudness.
+3 dB = double the POWER (watts); -3 dB = half the power (the 'half-power' point).
dB(voltage) = 20 × log10(V_out / V_in). dB(power) = 10 × log10(P_out / P_in).
Mic level ≈ -60 to -40 dBu (millivolts); preamp gain (typically 0-60 dB) lifts it to line level.
Set gain with the LOUDEST part of the source (soundcheck the chorus, not the verse), leaving ~6 dB spare.
Unity gain (0 dB / fader at the 'U' mark) = signal passes through unchanged, neither boosted nor cut.

How it works

Start with the channel fader and master at unity (0 dB / 'U').
Have the performer play their LOUDEST part during soundcheck.
Raise the preamp GAIN/TRIM until the channel meter peaks around -12 to -6 dBFS.
Leave roughly 6 dB of headroom so transients don't clip.
Check the master bus meter stays out of the red, then set amp/speaker level last.
Watch for the red CLIP/PEAK LED: if it flashes, back the gain off, never the fader.

Real examples

Quiet acoustic guitar DI: gain too low, you crank the fader and hear hiss; add 10 dB preamp gain instead and the hiss vanishes.
Screaming vocalist: gain set on the verse, chorus hits and the red CLIP LED slams on; reset gain on the loud part.
Laptop playback (-10 dBV consumer) into a +4 dBu desk: sounds weak until you boost ~12 dB to match line level.
Kick drum transient: average reads -18 dBFS but the peak still spikes to -3 dBFS, exactly why you keep headroom.
DJ pushing the mixer master into the red feeds a clipped, distorted signal to amps no fader move downstream can fix.

How it helps in live sound

Gain-stage the whole desk BEFORE you touch any EQ or compression: fix levels first.
Target channel meters at -18 dBFS average / -6 dBFS peak on digital consoles (X32, SQ, CL5).
If a channel clips, lower the PREAMP GAIN, not the fader: the fader is post-converter and won't stop clipping.
Boosting EQ adds gain too: a +6 dB low boost can push a clean channel into the red, re-check the meter.
Keep amps/processors at their rated input; don't fix a quiet desk by cranking the amp, you just amplify noise.
Engage the 20-26 dB PAD on hot sources (kick, snare, horns) if gain is near minimum and still clipping.

Everyday analogy

Pouring water down a row of buckets from mic to speakers: each bucket needs filling near the top for a strong stream, but overfill any one and it splashes out as ugly distortion.

Watch out

Myth: 'turn the channel fader down to stop clipping.' Wrong: if the PREAMP overloads the converter, the signal is already distorted before the fader; lower the gain/trim instead.

Fun fact

The -18 dBFS target isn't arbitrary: it lines up 0 VU on old analogue gear with the EBU digital reference (the US SMPTE standard sits a touch lower at -20 dBFS), so a needle hitting 0 leaves ~18 dB of headroom for transients.

Key takeaways

Loud enough to beat the hiss, quiet enough to never clip: that's the whole game.
0 dBFS = hard digital wall; aim for -18 dBFS average with ~18 dB headroom.
Set gain on the LOUDEST part of the source during soundcheck.
Clipping = fix the GAIN/TRIM, hiss = fix the GAIN up; the fader is for mixing, not rescue.
+6 dB doubles voltage, +3 dB doubles power, +10 dB ≈ twice as loud to your ears.