- Author Profile John Moxey
- John Moxey Website
Noise is an unavoidable part of recording. Every microphone, every cable, every preamp contributes something to the noise floor, and in a multi-track session those contributions stack up. By the time you have twelve or sixteen tracks running, the combined noise can shift from barely noticeable to genuinely distracting – and that’s before you’ve added any processing.
Noise gates and expanders are the tools most commonly used to manage that problem. They work differently, they suit different situations, and understanding the distinction between them is what separates settings that sound natural from settings that make a mix sound like it was recorded inside a cardboard box.
This article covers both tools in full – how they work, what each control does, how to apply them to specific instruments, and the mistakes that are most likely to cause problems.
The Noise Problem
When you record a single track, the noise that comes with it is usually low enough to be of little concern. The trouble with multi-track recording is that noise accumulates. Each track contributes its own noise floor to the mix, and those floors add together. Ten tracks, each with a modest noise floor, can combine to produce a noise level that is clearly audible – sometimes annoyingly so.
The other factor is that noise doesn’t respect the music. During a vocal performance, when the singer takes a breath or leaves a gap between phrases, the track is still running. There’s no signal worth keeping, but there’s plenty of noise worth removing. Multiply that across every instrument in a session and you start to understand why noise management is worth taking seriously.
The key insight is this: once noise is embedded in a signal, removing it cleanly is very difficult. The tools available – noise reduction plugins, spectral repair – can help, but they always involve trade-offs with tone and character. The more effective approach is to prevent unnecessary noise from contributing to the mix in the first place, and that is exactly what gates and expanders do.
How A Noise Gate Works
A noise gate monitors the level of an incoming signal and compares it against a threshold that you set. When the signal is above the threshold – when there’s a performance happening – the gate is open and audio passes through unaffected. When the signal drops below the threshold – during a gap in the performance – the gate closes and the signal is attenuated or cut entirely.
Think of it as a door with an automatic lock. Loud enough signal: door opens. Signal drops away: door closes. The gate doesn’t care about the quality or character of what’s coming through, only its level.
In practice it’s a little more nuanced than that, which is where the controls come in.
Gate Controls
Threshold
Threshold is the central control, and everything else revolves around it. It sets the level at which the gate switches between open and closed. Audio above the threshold passes through; audio below it gets cut.
Setting the threshold correctly is the first task when dialling in a gate. Set it too high and you risk cutting into the signal itself – the starts of notes, quiet passages, decaying tails. Set it too low and the gate is sitting open when it should be closed, letting noise through during gaps.
The practical approach is to find the level of the noise you’re trying to control, then set the threshold just above it – high enough to keep the gate closed during silence, low enough not to interfere with the actual performance.
Attack
Attack controls how quickly the gate opens once the signal has crossed the threshold. It is measured in milliseconds – the time between the signal exceeding the threshold and the gate reaching its fully open state.
Fast attack times are important for sounds with sharp transients: snare drums, plucked strings, staccato playing. If the attack is too slow on a snare hit, the gate is still opening as the transient passes, and you lose the snap that gives the hit its character.
At the same time, very fast attack times – particularly on low-frequency sources – can cause an audible click or distortion. This happens because the gate opens abruptly mid-cycle in the audio waveform, creating a discontinuity. A slightly longer attack, just enough to smooth the transition, usually eliminates the problem without any noticeable impact on the sound.
For most sources, attack times of 1–10ms are a reasonable starting point. Drums and percussion tend to need shorter times; sustained sources like guitars or pads can tolerate longer ones.
Hold
Hold sets a minimum time that the gate will stay open after the signal has dropped below the threshold. Once the signal crosses the threshold going upward and the gate opens, the hold timer prevents the gate from beginning to close until that minimum time has elapsed.
The main purpose of hold is to prevent chatter – the rapid opening and closing that can happen when a signal is hovering close to the threshold. Without hold, a signal that crosses the threshold, dips briefly below it, and comes back up again might cause the gate to start closing and then reopen, creating an audible stutter. Hold keeps the gate locked open long enough that brief dips don’t trigger the release cycle.
Hold time needs to fit the natural duration of the sound. For a kick drum, a hold of 100–200ms is usually enough to keep the gate open through the body of the sound before release begins. For a sustained vocal phrase, longer hold times may be needed.
Release
Release controls how quickly the gate closes after the hold time has expired and the signal has fallen below the threshold. It is also measured in milliseconds and represents the time it takes for the gate to move from fully open to fully closed (or to the floor setting – more on that below).
Release is arguably the most critical control for natural-sounding gating. A release time that’s too short causes the tail of a sound to be cut off abruptly – the reverb trail of a snare, the natural decay of a guitar note, the end of a vocal phrase. To a listener, sudden silence sounds wrong. It doesn’t match how sounds behave in the real world.
Longer release times let the gate fade closed gradually, which is far less noticeable. The goal is for the closing of the gate to feel like the natural end of the sound rather than an interruption of it. For most material, release times of 100–500ms work well, though the right setting depends entirely on how the source sound naturally decays.
Range (Floor)
Range – sometimes labelled Floor on hardware units – sets how far the gate closes when the signal drops below the threshold. At maximum range, the gate closes completely and the signal is cut entirely. At reduced range settings, the gate closes only partially, allowing a proportion of the signal through even when it’s nominally closed.
This control is more useful than it might initially appear. Full closure is not always desirable. A vocal track recorded in a live room, for example, might have useful ambience during the gaps between phrases – the natural sound of the space. Cutting that completely creates an unnatural dead silence that draws attention to itself. Reducing the range so the gate closes to, say, -20dB rather than fully off lets some of that ambience through, maintaining a sense of the space while still controlling the noise floor significantly.
Range also offers a safety net when threshold settings aren’t quite perfect. A gate that closes to silence when it misjudges is far more obvious than one that closes to a low but non-silent level.
Gate Features
Beyond the core controls, most gates – hardware or software – include additional features that extend what they can do.
Hysteresis
Hysteresis is a way of building two threshold points into a single threshold control. The gate opens when the signal exceeds the threshold, but it only begins to close when the signal drops a set number of dB below that same threshold level. The gap between the opening threshold and the closing threshold is the hysteresis range.
The practical benefit is a more stable gate. Without hysteresis, a signal hovering just at the threshold – rising slightly above it, dropping fractionally below, rising again – can cause the gate to repeatedly open and close. With hysteresis, the signal has to drop significantly below the threshold before the closing cycle begins, which prevents that kind of chatter without needing to set an unusually long hold time.
Not all gates expose hysteresis as a separate control. On many units it is simply built in as part of the gate’s behaviour.
Sidechain
The sidechain input allows the gate to be triggered by a signal other than the one it’s processing. Instead of monitoring the source track to decide when to open and close, the gate listens to the sidechain signal and uses that to make its decisions – while the audio passing through the gate remains the source track.
The classic application is tightening up a rhythm section. Feed a kick drum signal into the sidechain of the bass guitar gate, and the bass guitar track will open and close in response to the kick – the two instruments effectively being locked together rhythmically. Done well, this can add punch and cohesion to a rhythm section that feels slightly loose.
Sidechaining also appears in creative contexts: feeding a synth pad into the sidechain of a gate patched across an audio signal can create rhythmic chopping effects that would be difficult to achieve any other way.
Filters and EQ in the Sidechain
Many gates allow the sidechain signal to be filtered or EQ’d before it reaches the detection circuit. This is particularly useful when the source that’s triggering the gate is competing with other sounds.
The most common example is a snare drum microphone. In a live drum kit recording, the snare mic inevitably picks up spill from other drums – and especially from the hi-hat, which is usually close by and often loud. If the gate is set to trigger from the raw snare mic signal, hi-hat hits can open the gate when they shouldn’t, creating false triggers.
By filtering the sidechain to emphasise the frequency range of the snare (typically 150–300Hz for the body, with some presence in the 3–5kHz range) and roll off the high frequencies where the hi-hat lives, the detection circuit becomes much more selective. It responds strongly to snare hits and much less to hi-hat bleed, giving the gate a cleaner trigger.
The same principle applies anywhere you’re trying to gate a source that’s picking up unwanted spill alongside the intended signal.
Ducking
Ducking inverts the gate’s normal behaviour: rather than opening when the signal is loud, the gate closes. When the source signal exceeds the threshold, the output is attenuated rather than passed.
The most familiar context for ducking is broadcast – the presenter’s voice causes the background music to drop in level automatically, then the music returns when the presenter stops speaking. In a music production context, ducking can be used to create space for a lead vocal or solo instrument by automatically reducing the level of other elements when the lead is present.
Used without a sidechain, ducking reverses the dynamics of the source itself – louder parts become quieter, which has occasional creative uses but is rarely what you want for conventional noise management. The interesting applications almost always involve a sidechain: one signal controlling the attenuation of another.
How An Expander Works
An expander is, in broad terms, the opposite of a compressor. Where a compressor reduces the dynamic range of a signal – making the loud parts quieter relative to the quiet parts – a downward expander increases the dynamic range by making quiet parts quieter still.
Like a gate, an expander monitors signal level against a threshold. Above the threshold, the signal passes through unaffected. Below the threshold, the expander applies attenuation – but unlike a gate, which cuts hard (or to a floor level), the expander attenuates gradually and proportionally according to a ratio setting.
This is the critical distinction: a gate is essentially a switch. An expander is a gradual process.
The Ratio
The ratio determines how aggressively the expander attenuates the signal below the threshold. An expander ratio of 1:2 means that for every 1dB the signal falls below the threshold, the output drops by 2dB. A ratio of 1:4 means every 1dB drop below the threshold becomes a 4dB drop at the output.
At lower ratios, the effect is subtle – a gentle push downward that gradually reduces the level of quiet sounds without obviously cutting them. At higher ratios, the expander begins to behave more like a gate, with signals below the threshold dropping away quickly. The difference is that the transition is always smooth, always graduated, rather than switching between open and closed states.
A gate is effectively an expander with a very high ratio – essentially infinity to one, where any signal below the threshold goes to silence.
Expander Controls
Expanders generally have fewer controls than gates, reflecting their simpler mode of operation.
Threshold
Functions identically to the threshold on a gate: the level below which the expander begins to apply attenuation. Set it at the boundary between the wanted signal and the noise floor you’re trying to reduce.
Ratio
Sets the rate at which the signal is attenuated below the threshold, as described above. Most expanders offer ratios from 1:1.5 (very gentle, barely noticeable) up to around 1:5 or 1:10 (more aggressive, approaching gate-like behaviour).
Attack
Controls how quickly the expander begins applying attenuation when the signal drops below the threshold. A fast attack means the expander acts quickly as the signal fades; a slower attack gives a more gradual transition.
Release
Controls how quickly the expander releases its attenuation when the signal rises above the threshold again. A slower release prevents the expander from reacting too sharply as a sound begins, which can otherwise cause a pumping effect at the start of notes.
Gate vs. Expander: When to Use Which
The choice between a gate and an expander usually comes down to one question: do you need silence, or do you just need less noise?
A gate is appropriate when you genuinely want the gaps in a performance to be silent – or close to it. Drum tracks are the obvious case. Between hits, there’s nothing useful happening on a tom mic or a kick mic, and a clean cut between hits is both practical and natural-sounding, because drums are percussive sounds with clear beginnings and ends.
An expander is usually the better choice for sources with more gradual dynamics – vocals, acoustic guitar, piano, strings. These sources don’t have the sharp on/off character of drums, and applying a gate to them often creates an obvious, artificial quality: sounds that cut off too abruptly, gaps that feel unnaturally dead. An expander follows the natural decay of the sound more convincingly, pushing the noise floor down without creating the sense that something has been switched off.
The expander is also the right tool when you’re dealing with a signal that has already been over-compressed. Heavy compression can make a recording feel flat and unnatural by reducing the dynamic contrast. Applying a downward expander below the threshold can restore some of the original dynamic feel, giving quiet passages room to breathe again without dramatically affecting the louder sections.
A useful rule of thumb: if the difference between the signal and the noise is large and the source has clear transients, use a gate. If the transition between signal and silence is gradual or the source is continuous, use an expander.
Instrument-Specific Use Cases
Kick Drum
The kick drum is one of the cleanest applications for a gate. The sound has a sharp attack and relatively short sustain, and the gaps between hits should be silent. A gate with a fast attack (1–5ms), a hold long enough to cover the body and tail of the kick (100–200ms is a common starting point), and a moderate release (50–150ms) will typically give clean results without cutting into the sound itself.
The threshold needs to be set carefully. The kick mic will often pick up significant spill from the rest of the kit, particularly the bass drum beater mechanism and snare sympathetic resonance. The threshold needs to be high enough to reject that spill without catching the quieter hits in a dynamic performance.
Snare Drum
Snare gating follows similar principles to kick, but the snare mic is usually dealing with more bleed – hi-hat spill is a persistent problem. A filtered sidechain that emphasises the snare’s frequency range and reduces sensitivity to high-frequency content from the hi-hat is often essential for clean triggering.
Attack on a snare gate can typically be very fast – 1ms or less – since the crack at the top of a snare hit is exactly what you want to preserve. Release needs to be long enough to capture the snare’s characteristic ring and reverb tail; cut it too short and you lose the character of the drum.
Toms
Tom gating is straightforward in principle but can be fiddly in practice. The tom attack is sharp, the sustain moderate, and a gate works well to isolate individual hits. The challenge is that toms are often hit at varying velocities, so the threshold needs to be low enough to capture the quieter hits without letting bleed through from louder nearby sources.
A sidechain filter that emphasises the fundamental frequency of each tom can help with selective triggering. Attack times can be fairly fast (2–10ms); release should be long enough to capture the natural ring of the drum, which on a well-tuned tom can be considerable.
Electric Guitar
Between phrases, an electric guitar through an amplifier is producing noise – amp hiss, interference, pick-up buzz. Gating between phrases is a standard part of guitar recording, particularly for high-gain sounds where the noise floor of the amp and pedal chain can be significant.
The challenge with guitar is that the signal level can vary quite a bit, especially in a dynamic performance. A gate needs to be set to handle the quietest notes without triggering on the amp noise between them. An expander is often more forgiving here – it’ll push the noise floor down during quieter moments without hard-cutting between phrases, which can sound more natural on held chords that decay gradually.
Attack on a guitar gate should generally be fast enough not to miss the pick attack; release needs to be long enough for naturally decaying notes to ring out. If you’re cutting off the end of sustained chords, the release is too short.
Vocals
Gating vocals requires particular care, because the voice is a continuous and highly dynamic source. Hard gating between phrases can work, but it’s easy to make mistakes – cutting breath sounds that are actually part of the performance, clipping the end of sustained notes, or creating an unnaturally dead silence between lines that the listener subconsciously notices.
An expander is often the better choice for vocals. It pushes the noise floor and room sound down during quiet moments without cutting them to zero, maintaining the sense of a real acoustic space while still reducing the noise contribution to the mix. This is particularly true for vocals recorded in a treated but not dead room, where some ambience is desirable.
If a gate is used, release time needs special attention. The end of a vocal phrase usually has some decay and breath – cutting that abruptly is one of the most noticeable artefacts of over-aggressive gating.
Bass
Bass guitar is a case where the gate/expander decision depends heavily on style. A played or picked bass with clear notes and gaps between phrases can be gated similarly to electric guitar. A more continuous playing style – busy runs, walking bass lines – may be better served by an expander.
The sidechain/kick drum technique mentioned earlier is worth exploring for bass: triggering the bass gate from the kick drum signal so the two instruments open and close together. This can significantly tighten a rhythm section and help both instruments feel more locked in. The effect ranges from subtle to very pronounced depending on how aggressive the gate settings are.
Live Recording
Live recording presents the noise management challenge in its most acute form. Multiple microphones are open simultaneously, every mic is picking up bleed from surrounding sources, and the performances are happening in real time with no opportunity for overdubbing.
Gates and expanders are widely used in live sound for exactly this reason – to prevent microphones that aren’t currently in use from contributing ambient noise and spill to the mix. The approaches are the same as in the studio, but the margin for error is smaller: there’s no opportunity to go back and fix a gate that’s chopping off a performer’s quiet phrases.
Filtered sidechains are particularly valuable in live recording, where the acoustic relationships between sources are often complex. A sidechain filter tailored to each source’s frequency range makes a gate more selective and less prone to false triggering.
Common Mistakes
Threshold Set Too High
The most frequent problem. Setting the threshold too high – trying to be aggressive with noise reduction – means the gate is cutting into the signal itself. Quiet parts of a performance, the natural tail of sustained notes, vocals that drop in level at the end of phrases: all of these become casualties of a threshold that isn’t quite low enough. Reduce the threshold until the signal passes cleanly, then adjust the other controls to handle the noise below it.
Attack Too Fast on Low Frequency Sources
Very fast attack times on kick drum, bass, or other low-frequency sources can introduce a click at the moment the gate opens. This is an electrical discontinuity – the gate is switching in mid-waveform. Extending the attack time slightly, typically to 2–5ms, usually removes the click entirely without any audible effect on the transient.
Release Too Short
Short release times are responsible for most of the obviously unnatural gating people notice in recordings – that sense of sound being switched off rather than decaying. If you can hear the gate closing, the release is probably too short. As a starting point, set it longer than you think you need, then bring it back until it sounds natural rather than until it becomes inaudible.
No Hold Time, Causing Chatter
A signal hovering near the threshold will cause a gate without adequate hold time to rapidly open and close – the chatter problem mentioned earlier. The symptom is a kind of stuttering that’s particularly obvious on sustained sounds. Increasing the hold time until the gate stays open through the natural duration of each sound usually resolves it. Hysteresis, if available, is an alternative approach.
Over-Gating the Mix
Applying aggressive gating to many tracks simultaneously can create a mix that sounds unnaturally clean – too clean. Real recordings have ambience and incidental noise between events, and when all of that is removed, the result is a kind of sterile silence during the gaps between sounds that draws attention to itself. Selective gating, applied to the tracks that need it most, usually sounds more natural than gating everything. Where ambience is desirable, reducing the range control rather than cutting to complete silence often helps.
Using a Gate When an Expander Would Sound More Natural
For any source with continuous or gradually changing dynamics, a hard gate will almost always sound less natural than an expander. If you find yourself constantly fighting with release times trying to make a gate sound smooth, it may simply be the wrong tool. Try an expander instead – lower ratios especially can achieve significant noise reduction while being virtually inaudible in operation.
A Note On Subtlety
The best use of a noise gate or expander is usually the one nobody notices. When gating is working well, the listener isn’t aware of it – they simply experience a clean recording with well-defined dynamics and no distracting background noise.
The trouble is that the instinct when learning these tools is often to be aggressive: set the threshold high, set the release short, make sure every trace of noise is gone. That instinct tends to produce the opposite of the intended result – a mix full of audible switching artefacts and unnaturally truncated sounds.
The practical discipline is to set everything conservatively first, then tighten gradually while listening critically. If at any point you can hear the gate or expander working, that is the signal to back off rather than push further. Noise that’s slightly audible in isolation often disappears entirely in the context of a full mix; gating artefacts almost never do.
Knowing when to leave things alone is as much a part of using these tools as knowing how to set them.
Discuss this article in our Music Forum.
Related Articles
Do you want to read more gear reviews? If so, you can find articles and tutorials on our Music Product Reviews page.
You might find the following selection of recording and music production articles by John Moxey useful:
