What level should I record at?

Aim for peaks around -12 dBFS while tracking, set at the interface's input gain with the performer playing their loudest part. It will look too quiet on the meter, but 24-bit recording has plenty of resolution down there, and the headroom protects you from unexpected loud moments that would otherwise clip.

0 dBFS (decibels relative to full scale) is the maximum possible level in a digital system - a hard ceiling with nothing above it. Every level you record is a negative number. Hitting 0 dBFS causes clipping, which in digital is a harsh, unfixable distortion rather than the pleasant saturation of analog tape.

Why does my recording sound thin or noisy?

Almost always a gain-staging problem, not a gear problem. Recording too quietly leaves you near the noise floor, so turning it up later raises the hiss with it. Recording too loud clips. Setting a healthy input level at the source fixes both and makes cheap gear sound far better.

Should I record as hot as possible?

No. That advice comes from the analog tape era, where a hot level pushed the signal above tape hiss into pleasant saturation. Modern 24-bit converters have such a low noise floor that recording hot buys nothing and only risks clipping. Leave headroom instead.

// Recording

Gain Staging: The One Thing Nobody Explains Properly

A HomeRecordingPro guide · Updated 2026-07-19 · 4 min read

Quick answerGain staging means keeping your signal at a healthy level at every point in the chain. While tracking, set your interface's input gain so the loudest peaks land around -12 dBFS, leaving headroom below the 0 dBFS ceiling. Do not "record hot" - that advice was for tape. Cheap gear staged well beats expensive gear staged badly.

Contents

The two numbers that matter
Why 0 dBFS is a wall, not a target
Where to actually set your levels
The "record hot" myth
Staging the whole chain

If your recordings sound thin, noisy, harsh, or just amateur in a way you can't name, the cause is almost never your microphone or your interface. It's gain staging — the level of the signal at every point in the chain. Get it right and cheap gear sounds surprisingly good. Get it wrong and a very expensive microphone sounds like a phone.

The reason nobody explains it well is that it lives at the boundary of two worlds with different rulers: the analog world (your mic and preamp) and the digital world (your converter and DAW, measured in dBFS). Understanding that boundary is the whole game.

The two numbers that matter

Every signal path has a noise floor (the quiet hiss always present underneath) and a ceiling (the loudest level before distortion). The usable space between them is your dynamic range, and gain staging is simply keeping your signal parked in the healthy middle of that space — well above the hiss, comfortably below the ceiling.

Record too quietly and you sit near the noise floor; when you turn it up later, you turn the hiss up with it. Record too loud and you hit the ceiling and clip, which in digital is not a warm sound — it's a hard, ugly, unfixable crackle.

Why 0 dBFS is a wall, not a target

In your DAW, levels are measured in dBFS — decibels relative to full scale. The maximum possible level is 0 dBFS. There is nothing above it; it is a brick wall. Everything you record is a negative number: -18, -12, -6 dBFS.

Newcomers see the meter and instinctively try to get it close to the top, because in every other context "more" is "better." Here it's the opposite. You want your peaks landing around -12 to -6 dBFS and your average level much lower. That empty space at the top isn't wasted — it's headroom, and it's what lets transients (a snare hit, a plucked string) punch through without slamming the wall.

Aim for peaks around -12 dBFS while tracking. It looks "too quiet." It isn't. You have 24 bits of resolution; you are not running out of quality down there, and you're leaving room for the loud moments you can't predict.

Where to actually set your levels

Set gain at the source first — the preamp/interface knob — and get it right there before anything downstream. Have the performer play the loudest part they'll actually play, then set the gain so those peaks hit roughly -12 dBFS. Everything after that (fader, plugins) is for balance and tone, not for fixing a bad input level.

Too low? Turn up the interface gain, not the DAW fader. The fader amplifies the noise floor too; the preamp gets you clean level.
Clipping the input? Turn down the interface gain. If the clip light on the interface flashes, no fader move in the DAW can undo it — the damage happened before the signal was recorded.

If latency makes monitoring your input awkward, that's a separate fix — see our guide to sorting out recording latency.

The "record hot" myth

You'll read old advice to "record as hot as possible." That was correct — for analog tape, where a higher level pushed the signal further above tape hiss and into pleasant saturation. Digital converters don't work that way. Modern 24-bit converters have enormous dynamic range; the noise floor is so low that "recording hot" buys you nothing and risks clipping. The advice survived the technology it was made for. Ignore it.

Staging the whole chain

Gain staging isn't a one-time move at the input — it repeats at every stage. A compressor with too much input gain distorts; an EQ boosting 6 dB raises the level into the next plugin. The discipline is simple: keep a consistent, moderate level flowing through the chain, checking that each plugin is receiving roughly what it expects (many emulated analog plugins are calibrated to sound best around -18 dBFS average). Use gain-trim controls to bring levels back down after anything that boosts them. This same discipline makes mixing vocals far easier, because every plugin behaves when it's fed a sane level.

Do this and your mixes get easier, your plugins behave, and your recordings stop sounding thin. It's unglamorous, it's free, and it matters more than any single piece of gear you'll ever buy.