Compressor vs Limiter in Live Sound: Control vs Protection

TL;DR — A compressor reduces dynamic range above a threshold with controllable ratio and timing, shaping a vocal, instrument, bus, or mix. A limiter uses a very high ratio and usually faster behaviour to restrain peaks near a ceiling. Limiting can manage peaks or form part of loudspeaker protection, but a console limiter alone does not guarantee safety. Choose by purpose, calibrate the complete path, verify gain reduction, and document ownership and recheck conditions.

Table of contents

1. Compare compression and limiting
2. Understand the controls
3. Choose processing by purpose
4. Set a compressor for musical control
5. Set a limiter for peak control
6. Treat loudspeaker protection as a system design
7. Avoid common dynamics mistakes
8. Document dynamics in a tech rider
9. FAQ

Compare compression and limiting

Both processors turn down signal after it crosses a threshold. The practical difference is the intended result and severity.

A compressor uses a ratio such as 2:1 or 4:1 to make level changes above the threshold smaller. It may gently control a vocal, add consistency to bass, shape a drum transient, or reduce bus peaks while preserving movement.

A limiter uses a very high ratio, often 10:1 through infinity-to-one depending on the device and convention, to keep output near a defined ceiling. It may catch occasional peaks, constrain a broadcast or recording feed, or operate inside a calibrated loudspeaker-management system.

The boundary is not standardized. Some devices change mode at a particular ratio; others label any high-ratio compressor a limiter. Look at the transfer curve, attack and release behaviour, detection method, lookahead, ceiling and device manual rather than relying on the name alone.

Understand the controls

Threshold and ratio

Threshold defines where gain reduction begins. Ratio describes how much the output rises after the input exceeds that point. At 4:1, 4 dB of input rise above threshold ideally produces 1 dB of output rise before considering knee and detector behaviour. At infinity-to-one, further steady-state input rise ideally produces no output rise.

This arithmetic describes a static curve, not every peak. Attack time, detector response, lookahead, overshoot and release determine what happens in time. A limiter can allow brief peaks above its nominal threshold if its design is not a true brick-wall implementation.

Attack, hold and release

Attack controls how quickly gain reduction develops. A slower attack may preserve a transient but allow a larger peak through. A faster attack catches more of the transient but can reduce punch or distort low-frequency waveforms.

Release controls how quickly gain returns after the signal falls. Too short can create modulation, pumping or distortion; too long can hold the programme down after one event. Some processors add hold, programme-dependent timing or automatic release.

Knee, detector and sidechain

A soft knee introduces compression gradually around threshold; a hard knee follows a sharper transition. Peak and RMS-like detectors respond differently, while modern processors may use proprietary combinations. A sidechain filter changes what the detector reacts to without directly EQing the audible path.

Do not copy settings between processors solely because the control labels match. Time constants, scales, internal headroom and detection behaviour can differ.

Makeup gain and output ceiling

Makeup gain raises level after compression. It can make a processed signal appear better simply because it is louder, and it can drive a downstream limiter harder. Compare bypassed and processed signals at similar perceived loudness.

A limiter's output or ceiling control establishes a target only at that processor's output reference. Downstream digital gain, D/A conversion, analog controls, amplifiers and powered-speaker input sensitivity can invalidate it.

Choose processing by purpose

Start with a sentence that describes the required result.

Processing belongs at the layer where the problem exists. A channel compressor controls one source. A group compressor reacts to the sum of assigned sources. An output limiter sees the destination mix. Stacking processors at several layers without defined ownership can create unexpected gain reduction and hide the cause.

The DCA versus subgroup guide explains why a control group cannot replace a subgroup when combined audio processing is required.

Set a compressor for musical control

1. Establish clean gain staging

Set source, preamp, channel, bus and output levels so the unprocessed path has headroom. Compression cannot repair clipping that occurs before the processor. Follow the complete gain-staging workflow.

2. Define the dynamic problem

Identify whether the issue is occasional peaks, inconsistent phrases, excessive sustain, a weak transient, or competition within the mix. If the source is unstable because of microphone technique, instrument setup or monitor spill, address that first.

3. Begin with moderate controls

Choose a moderate ratio and set threshold so gain reduction occurs on the passages that need control. Adjust attack to retain or restrain the desired transient. Set release so reduction recovers with the programme rather than audibly pumping.

There is no universal vocal, bass or drum preset. Performer dynamics, microphone, arrangement, stage volume and processor design all change the result.

4. Compare in the complete mix

Match output level as closely as practical and bypass the compressor. Listen for intelligibility, envelope, noise, feedback margin and emotional movement. A soloed channel can help expose artifacts, but final decisions belong in the intended FOH, monitor or stream mix.

5. Check every affected destination

If FOH and monitor paths share channel processing, the compressor can change both. Makeup gain may increase spill and reduce gain before feedback. Confirm routing and processing pickoff points before assuming one adjustment is local.

Set a limiter for peak control

1. Name the ceiling and reference point

State what must not be exceeded and where it is measured: console output dBFS, interface input level, processor output voltage, amplifier input, or another calibrated point. Prevent clipping is incomplete without naming the next device and its accepted level.

2. Place the limiter before the owned boundary

Put the limiter where no uncontrolled gain exists between it and the boundary it is meant to protect. For a recording feed, that may be the feed output before the recorder. For loudspeakers, purpose-built processor placement and locked downstream gain are normally required.

3. Set threshold or ceiling from evidence

For creative peak control, use the destination meter and listen at representative programme level. For system protection, calculate and measure according to loudspeaker, amplifier and processor specifications. Do not derive a protection threshold from a generic article or another system's show file.

4. Adjust time behaviour for the application

Fast timing catches peaks but can sound aggressive or distort. Slow timing may allow damaging or clipping peaks through. Protection processors may use separate peak and long-term limiters. Use manufacturer guidance and a qualified system engineer for those settings.

5. Test the complete path safely

Confirm gain reduction, output meter, downstream input and actual destination. Test muting, bypass, scene recall, sample-rate changes and redundant paths where relevant. Never use uncontrolled high-level test signals through connected loudspeakers.

Treat loudspeaker protection as a system design

A limiter does not automatically protect a loudspeaker. Protection depends on the loudspeaker component limits, load impedance, amplifier voltage and gain, DSP reference levels, crossover filters, limiter type and timing, downstream controls, wiring and operating condition.

Peak and long-term limits solve different problems

Short transients can exceed excursion or voltage limits, while sustained energy can overheat voice coils. A single fast limiter may not address both. Manufacturer presets may combine crossover, EQ, delay, peak limiting and long-term or thermal models designed for a specific loudspeaker and amplifier pairing.

Calibration must include downstream gain

If an amplifier input knob or powered-speaker sensitivity control is changed after calibration, the acoustic and electrical result changes. Lock or mark approved controls, document the reference, and verify the path after equipment substitutions.

Preserve venue and manufacturer ownership

Do not replace approved system-processor settings with a console output limiter. The console may provide programme control, while the venue processor owns crossover, alignment and protection. Coordinate with the system technician before changing either layer.

The PA system rider guide shows how to specify coverage and handoff requirements without inventing unsafe processor values.

Avoid common dynamics mistakes

Using a limiter to hide bad gain structure

Constant heavy limiting can mask an overdriven upstream path while adding distortion and reducing impact. Find the first clipping point and correct levels there.

Adding makeup gain until the processor is always busy

Makeup gain raises the processed signal into downstream buses and limiters. Judge the intended control at matched loudness, then set the output for the actual signal path.

Treating infinity-to-one as an absolute brick wall

An analog or digital limiter may overshoot because of attack and detector design. Inter-sample peaks can also exceed a digital sample ceiling after conversion. Use a processor designed for the boundary and leave appropriate margin.

Copying channel settings into monitor outputs

A compressor intended for FOH tone is not a protection limiter. A wedge, IEM, stream and PA output have different transducers, risks and ownership. Configure each destination intentionally.

Assuming more compression creates a safer mix

Compression can raise average energy after makeup gain, increase perceived loudness and reduce feedback margin. A restrained peak meter does not prove that loudness exposure, amplifier load or loudspeaker heating is safe.

Forgetting scene and show-file scope

Processor state, sidechain source, bypass, threshold and output gain may change with scenes. Safes and recall filters can preserve only part of the configuration. Verify critical limiting after loading files and before opening outputs.

Document dynamics in a tech rider

Most artist riders should describe the required source, destination and outcome, then let the responsible engineer adapt processing to the equipment and room. Record exact dynamics values only when they are part of artist-owned equipment, a known show file, a broadcast boundary, or an approved system design.

When exact values are necessary, record device and firmware, path, detector or mode, threshold reference, ratio, knee, attack, hold, release, lookahead, output ceiling, downstream gain, purpose, owner and test condition.

In Techrider.live, add concise processing constraints to the same named inputs and outputs used on the stage plot and input list. Invite the responsible engineers to edit and save the same Rider, inspect history after a setting or ownership change, and export a dated PDF for offline handoff.

Dynamics handoff checklist

• The purpose is defined as dynamic shaping, peak control or calibrated protection.
• The processor path and downstream boundary are named.
• Gain structure is clean before processing.
• Threshold, ratio and timing are understood for the actual device.
• Makeup gain and output ceiling are checked separately.
• FOH, monitor, recording and stream pickoff points are verified.
• Loudspeaker protection includes approved presets, calibration and downstream gain controls.
• Ownership, scene recall and recheck triggers are documented.

FAQ

What is the difference between a compressor and a limiter?

A compressor typically uses a moderate ratio to reduce dynamic range and shape how a source or mix behaves. A limiter uses a high ratio and often faster timing to restrain peaks near a ceiling. The designs overlap, so device behaviour and intended purpose matter more than the label alone.

Is a limiter just a compressor?

Conceptually, a limiter is a high-ratio form of compression. Practically, dedicated limiters may add lookahead, specialized peak detection, output ceilings and timing behaviour that an ordinary compressor does not provide. A compressor set to a high ratio is not automatically a calibrated protection limiter.

Should a limiter go before or after a compressor?

For a common creative chain, compression may shape the source first and a limiter may catch later peaks. That is not a universal rule. Place each processor according to the signal boundary it owns, and account for makeup gain or other processing between stages.

Does a limiter protect speakers?

Only as part of a correctly designed and calibrated system. The limiter must match the loudspeaker, amplifier, impedance, processor reference, crossover and downstream gain. An uncalibrated console limiter cannot guarantee protection against peak excursion, sustained heating, clipping or wiring faults.

Can you use a limiter on live vocals?

Yes, a limiter can catch occasional vocal peaks, but regular heavy limiting may sound harsh, raise noise after makeup gain and reduce feedback margin. Moderate compression, good microphone technique and correct gain staging are usually the better starting points for consistent vocals.

Give every dynamics processor one clear job

Build a reusable Rider with Techrider.live, connect dynamics notes to stable source and destination names, and state who owns each boundary. A verified processor with one defined purpose is safer and easier to hand off than several overlapping compressors and limiters copied from an unknown system.