Gain Staging Without Clipping

Leave headroom, then raise it afterwards. A quiet, clean recording can be brought up losslessly. A clipped one has had its peaks cut off and thrown away, and nothing recovers them.

That asymmetry is the entire reason to set levels conservatively.

A note on this guide: this rests on how digital audio handles levels, rather than on recordings compared here. See how we test.

The problem, as it appears

Your voice is too quiet, so the gain goes up. Now it is loud enough, and there is a harshness on certain words — a crackle or a spitting quality on louder syllables that was not there before.

Or the opposite: everything is clean but you sound distant and thin, and turning it up seems to bring the room up with you.

Why "loud enough" disappoints as a target

Because loud and clean are different goals, and only one of them can be fixed later.

Digital audio has a hard ceiling. Below it, level is just a number that can be scaled up or down. At it, anything louder is simply not represented — the peaks are flattened and that information is gone at the moment of recording.

So aiming for "as loud as possible" spends something irreplaceable to gain something you could have had for free afterwards. Aiming for "comfortably below the ceiling" costs you nothing but a slider move later.

What actually matters

Set it against your loudest, not your normal

The peaks that clip are the ones you did not rehearse — laughing, emphasising a word, reacting to something.

So set the level while speaking the way you actually speak when animated, not while testing politely into the microphone. If your test is quieter than your reality, your headroom is imaginary.

Distance does what gain cannot

If you need a great deal of gain to reach a usable level, more gain is the wrong lever.

Gain amplifies everything arriving at the capsule — your voice, the room, the fans, and the preamp's own noise floor. Moving closer raises your voice relative to all of that, which is the only change that improves the ratio rather than scaling it.

That is the mechanism set out in microphone position and the proximity effect, and it is the free fix that makes gain staging easy.

Every stage has a level, and they multiply

The signal usually passes through at least three: the preamp on the interface or microphone, the operating system's input level, and the application's own gain.

Pushing one hard and cutting another back gets to the same apparent loudness by a worse route. The arrangement that works is a healthy level early — set at the preamp, where the signal is cleanest — and the later stages left near neutral.

Normalisation fixes quiet, not clipped

Worth separating because they get used interchangeably.

Normalisation scales an entire recording so its loudest point reaches a target. On a conservative recording that is exactly what you want. On a clipped one it faithfully scales up a recording whose peaks are already flattened, and the flat tops come with it.

A dynamic microphone makes this easier

A dynamic pattern is less sensitive, which sounds like a disadvantage and is frequently the opposite at a desk: less of the room arrives, so you can run a healthy level without the background coming with it.

The trade is explained in dynamic versus condenser.

The options, and who each suits

This page contains affiliate links. If you buy through them, Ugly Fixes may earn a commission at no extra cost to you. Products are selected independently.

Getting closer and lowering the gain suits everyone and costs nothing. Do this before considering any purchase — most gain problems are distance problems.

An audio interface suits anyone whose preamp is the limit, and its advantage here is a clean gain stage with real range rather than a fixed one buried in a driver.

A dynamic USB microphone suits an untreated room, because rejecting the room is easier than compensating for it at the gain stage.

A boom arm suits holding the closer position you found, which is what makes the lower gain setting stick.

A DAC or interface for monitoring suits hearing what you are actually sending, which is how clipping gets noticed at the time rather than afterwards.

The mistake to avoid

Recording hot on purpose. It is a habit carried over from analogue tape, where pushing the level was genuinely desirable, and it is exactly wrong digitally — the ceiling is hard, there is nothing pleasant above it, and what you lose does not come back. Leave visible space at the top and raise it later.

The second mistake is fixing a noise problem with gain. If the recording is quiet and noisy, adding gain produces a loud, noisy recording — the ratio is unchanged. The variable that improves it is how much of your voice reaches the microphone compared with everything else, and that is decided by distance and by which microphone you chose, not by a slider.