A MusicProductionWiki Publication Sound Better →
The Producer's Bible
All entries →
Beginner
Understand first: Compression Gain Staging Headroom

Limiting

noun / signal processing tool
The limiter is the last wall between your mix and the void — the difference between controlled power and catastrophic digital distortion.
Quick Answer

A limiter is an extreme form of compression with a very high ratio (typically 10:1 or greater, often described as ∞:1) that prevents a signal from exceeding a defined ceiling, known as the threshold or output ceiling. Unlike standard compression, a limiter acts as a hard stop — no signal should pass above the set level, making it the final safeguard against clipping in the signal chain. Limiting is used at virtually every stage of modern music production, from controlling individual track transients to maximizing loudness on a master bus.

New to Limiting? Start here
Parameters Before / After Quick Reference Common Mistakes
Common Misconception

Most producers believe that pushing a limiter harder makes their track louder and more competitive on streaming platforms.

Streaming platforms normalize all content to a target LUFS level, meaning an over-limited track will simply be turned down to match quieter content — you gain no loudness advantage and only lose dynamic range and sonic quality. A transparently mastered track at the platform's target LUFS will sound as loud as any heavily limited track, with none of the distortion, pumping, or listener fatigue that results from over-limiting.

What Is Limiting?

The limiter is the last wall between your mix and the void — the difference between controlled power and catastrophic digital distortion.

A limiter is an extreme form of compression operating at ratios of 10:1 or higher — often described as ∞:1 in the context of brickwall limiting — that enforces an absolute ceiling on a signal's output level. Where a standard compressor nudges peaks down and shapes dynamic contour, a limiter plants a flag and says nothing passes this line. That distinction is not semantic. It is the difference between a tool that shapes dynamics and a tool that eliminates them at a defined point. Every signal that strikes the threshold gets gain reduction applied fast enough that, in theory, nothing escapes above the ceiling. In practice, how well that theory holds depends entirely on the limiter's design, its look-ahead buffer, its release characteristics, and how hard you push it.

Limiting operates at virtually every stage of a modern production. On individual tracks, a fast-attack limiter catches rogue transients from a snare hit that overloads a bus, from a vocal that jumps an octave unexpectedly, or from a synthesizer whose envelope briefly spikes before settling. On the mix bus, a limiter acts as a safety net — preventing any combination of summed signals from hitting digital zero and clipping the output stage. At the mastering stage, the brickwall limiter becomes the primary loudness maximization tool, raising the overall level of the mix to hit streaming targets while preventing inter-sample peaks from causing distortion after digital-to-analog conversion. Three completely different contexts, three completely different approaches — but the same underlying mechanism.

The confusion most producers carry into their first serious mastering session is treating the limiter as a loudness machine. It is not. A limiter is a ceiling enforcer. Loudness is the byproduct of raising your mix's level until it hits that ceiling. The louder you push, the more gain reduction the limiter applies, and the more dynamic information you permanently remove from the signal. There is a point — and experienced mastering engineers find it by ear in minutes — where the gain reduction crosses from transparent to audible, where the drums stop hitting and start smearing, where the low end loses its punch and turns into a sustained blur. That point defines the loudest you can go before the limiter starts costing you more than it delivers.

What separates a limiter from a compressor at a functional level is attack speed and intent. A well-designed brickwall limiter uses look-ahead buffering — typically 1–10ms of delay introduced into the signal path — so the gain reduction algorithm can see the incoming peak before it arrives and begin attenuating before the ceiling is breached. This is fundamentally different from a compressor reacting to a signal after it has already appeared. The look-ahead is what makes a limiter capable of true brickwall behavior rather than approximate peak control. Without it, even a ∞:1 ratio compressor will let transient peaks slip through because no circuit or algorithm can apply gain reduction faster than the physics of the signal itself. The look-ahead buffer solves this by creating a controlled delay — the signal is held briefly while the gain reduction catches up, then released with the peak already attenuated.

Understanding true peak limiting versus standard peak limiting is non-negotiable for anyone delivering audio to streaming platforms. Standard peak limiting measures the digital sample values in your DAW — the numbers in the audio file. True peak limiting accounts for inter-sample peaks: the values that appear between samples when the digital signal is reconstructed by a DAC or converted during encoding. A signal that reads -0.1 dBFS in your DAW can produce peaks of +0.3 dBTP or higher after reconstruction. On streaming platforms that apply codec encoding on top of your file, those inter-sample peaks become audible clipping that you never heard in the session. True peak limiting catches and attenuates these values before they cause downstream distortion.

"True peak limiting is not optional in 2024. Inter-sample peaks cause clipping after encoding that you never hear in your DAW. It's an invisible problem that destroys streams."

— Patchwork (Manon Grandjean), Mastering Engineer (Christine and the Queens, FKJ) — Attack Magazine — Mastering for Streaming Platforms, 2022

A limiter is the absolute ceiling enforcer at the end of your chain — its job is not to make things loud but to ensure nothing passes above a defined level, and everything below that level is your loudness decision made upstream.

How Limiting Works

The core mechanism of a limiter is gain reduction applied at extreme ratios the moment a signal crosses the threshold. In a standard compressor, the ratio describes a gradual relationship: 4:1 means for every 4dB the signal exceeds the threshold, only 1dB exits. At ∞:1, the math collapses — for any amount the signal exceeds the threshold, zero additional level exits. The signal is clamped flat at the ceiling. In reality, no analog circuit achieves true infinity-to-one, but modern digital limiters in look-ahead mode come close enough that the output ceiling is treated as absolute. The gain reduction element — whether a VCA, FET, optical cell, or digital algorithm — receives a control voltage or coefficient derived from the level detector and attenuates the signal in real time to enforce that ceiling.

Look-ahead buffering is the design feature that separates a true brickwall limiter from an aggressive compressor. The signal is split into two paths: a detection path that runs slightly ahead of the audio path, and the audio path itself, which is artificially delayed by the same amount. When the detection path identifies an incoming peak that will exceed the ceiling, it calculates the necessary gain reduction and sends it to the attenuator before the peak in the audio path arrives. The audio path arrives at the attenuator already reduced. The look-ahead window is typically 1–10ms, chosen as a tradeoff: shorter look-ahead means faster response to very brief transients; longer look-ahead gives the algorithm more time to apply smooth gain reduction, reducing artifacts on complex material. The cost is latency — every ms of look-ahead is 1ms of delay added to the signal chain, which is why limiters are almost always the last processor in the chain rather than inline during tracking.

Release behavior is where most limiter artifacts originate. Once a peak is caught and gain reduction is applied, the limiter must return to unity gain — but how quickly? Release too fast and you hear a click or pop as the gain snaps back; release too slow and the sustained gain reduction after the peak creates pumping, where the level seems to duck and recover audibly after each transient. The best modern limiters use program-dependent release: an algorithm that analyzes the signal content after the peak and adjusts the release time based on whether loud content continues or the signal drops away. Short peaks in otherwise quiet passages get fast release to avoid pumping; dense, sustained passages get slower release to avoid clicking. This is why a well-designed limiter on a loud rock mix sounds transparent while a poorly designed one sounds like someone is riding a fader erratically behind every snare hit.

A limiter's intelligence lives in its look-ahead detection and release algorithm — the threshold sets the ceiling, but the timing characteristics determine whether the ceiling sounds like a wall or a cliff edge.

Limiting — Key Parameters

Every limiter presents a subset of the same core parameters, and knowing the difference between what each one controls versus what it sounds like in practice is the gap between setting a limiter by numbers and setting it by ear. The following parameters appear across virtually every limiter in the hardware and plugin world — from the Neve 33609 to FabFilter Pro-L 2 — and they interact in ways that make each one's setting meaningless without knowing the others.

Threshold / Input Gain Typically -∞ to 0 dBFS; practical range -20 to 0 dBFS

The level at which the limiter begins applying gain reduction. On many mastering limiters, the threshold is expressed as an input gain knob instead — you raise the input into a fixed ceiling rather than lower the ceiling to the input. Either way, the functional result is the same: more threshold engagement means more gain reduction applied to peaks. At 1–2dB of gain reduction, limiting is transparent. At 6dB or more, the dynamics of the mix begin to flatten audibly. Set this by watching the gain reduction meter, not by ear alone — the first thing limiting removes is your ability to judge loudness objectively.

Output Ceiling / Output Level -0.1 to -1.0 dBTP for streaming; -0.3 dBTP recommended for lossy encoding

The absolute maximum level the limiter will allow to exit. For streaming delivery, set the output ceiling to -1.0 dBTP as a minimum; -0.5 dBTP is safer for Spotify and Apple Music specifically, which apply additional encoding that can raise inter-sample peaks. For vinyl or CD masters, -0.1 dBFS is common. Never set the output ceiling to 0 dBFS on material going to lossy encoding — the codec processing will push peaks above zero and create clipping in every transcoded file.

Attack 0.01ms – 50ms; typical starting points 0.1–1ms for mastering, 1–10ms for track limiting

How quickly the limiter engages gain reduction once a peak crosses the threshold. On a brickwall mastering limiter with look-ahead, the attack is largely managed by the look-ahead buffer — the control labeled "attack" usually affects the shape of the gain reduction curve rather than its absolute speed. On track limiters without look-ahead, attack is critical: below 0.5ms and you're killing transients; above 5ms and transients pass through before gain reduction kicks in, which is often useful for preserving snare crack and pick attack.

Release 1ms – 500ms; auto/program-dependent mode preferred for mastering

How quickly gain reduction recovers after a peak. Fast release (under 10ms) causes audible clicking or distortion on bass-heavy material because the gain modulation creates its own artifacts at low frequencies. Slow release (over 100ms) can cause sustained pumping where the entire level drops after a loud peak and takes an audible moment to recover. Program-dependent release — available on most modern mastering limiters — analyzes the signal content and modulates release time automatically. Use it as the default and only switch to manual release when program-dependent mode is creating specific artifacts you can identify.

Look-Ahead 0ms – 10ms; 1–3ms typical for mastering, 0ms for real-time performance use

The delay buffer that allows the gain reduction algorithm to see incoming peaks before they arrive at the audio path. More look-ahead means more transparent limiting on transient-heavy material because the gain reduction is fully engaged before the peak arrives. Less look-ahead reduces latency — critical for live performance and monitoring contexts where delay is disruptive. In the mastering chain, always use maximum look-ahead unless you are specifically chasing a harder, more aggressive limiting character where the gain reduction is slightly late and creates a mild saturation-like smearing effect.

Oversampling 1x, 2x, 4x, 8x, 16x; 4x minimum for mastering

Oversampling temporarily increases the sample rate inside the limiter algorithm, revealing inter-sample peaks that would be invisible at the native rate and allowing the limiter to catch them. At 1x oversampling, the limiter processes only at your session's sample rate, missing peaks that form between samples during reconstruction. At 4x, most inter-sample peaks are visible and caught. At 16x, the processing is exhaustive and CPU-intensive but leaves essentially no inter-sample distortion. For final masters going to streaming, 4x oversampling is the practical floor — anything below it means you are not controlling true peaks regardless of what your output ceiling reads.

The relationship between threshold, look-ahead, and release creates the character of every limiter in use. A short look-ahead with fast release sounds aggressive — you hear the gain reduction working, and the transients have a compressed edge rather than a transparent ceiling. A long look-ahead with program-dependent release sounds transparent to the point where even experienced listeners cannot identify that limiting is occurring at 3–4dB of gain reduction. Neither is categorically correct. The aggressive character serves hard rock, metal, and hip-hop where density and perceived loudness impact the emotional delivery of the genre. The transparent character serves acoustic, jazz, and dynamic pop where the listener's perception of space and movement is part of what makes the record work.

Gain reduction and output ceiling interact in a way that confuses producers consistently: raising the input into the limiter and lowering the output ceiling simultaneously is not a neutral move. Raising input increases gain reduction on peaks while raising the average level of everything below the threshold. Lowering the ceiling does not change how hard the limiter is working — it only changes where the ceiling sits. The practical implication is that you should always set your output ceiling first, then raise input gain (or lower threshold) until you reach the target loudness. Adjusting in the opposite order means every ceiling change requires re-evaluating your loudness decision from scratch.

The output ceiling is a delivery specification — set it first and never move it; the threshold or input gain is your only loudness lever, and everything above 4dB of gain reduction is a dynamic range trade-off, not a free loudness gain.

Quick Reference Card

-1.0 dBTP True Peak Output Ceiling

Setting the output ceiling to -1.0 dBTP (true peak) is the universal professional standard for streaming and digital distribution, providing 1dB of headroom to absorb inter-sample peaks created during MP3/AAC encoding and D/A conversion. This number appears in the technical specifications of Spotify, Apple Music, YouTube, and virtually every broadcast standard globally.

These are production starting points for limiting across common contexts — dial these in first, then adjust by ear with the gain reduction meter as your reference, not the waveform display.

SourceRatioAttackReleaseThresholdNotes
Master Bus (Streaming)∞:1Look-ahead 1msAuto/Program-1.0 dBTP ceilingMax 3–4dB GR; 4x+ oversampling; true peak mode on
Master Bus (CD/Vinyl)∞:1Look-ahead 2msAuto/Program-0.3 dBFS ceilingAllow up to 6dB GR for loud genres; check for pumping at 4–6dB
Snare / Drum Bus20:10.5ms50ms-6dB below peakCatch rogue hits; retain crack; use GR meter as guide not ears alone
Vocal Track10:15ms80ms auto-3 to -6dB below peakAfter compressor; catch jumps; slow attack preserves consonants
Bass / Sub∞:10.1ms100ms-3dB below peakTames fundamental spikes; use with multiband compression for dense low end
Mix Bus (Pre-Master)∞:1Look-ahead 0.5msAuto-0.5 to -1.0 dBFSSafety net only; if GR exceeds 1dB, remix before mastering
Room / Overhead Bus10:110ms200ms-6dB below peakSlow attack preserves room transient attack; aggressive limiting creates density
EDM / Electronic Master∞:1Look-ahead 2msAuto-0.5 dBTP ceilingAllow 4–6dB GR for genre expectations; sub control upstream is critical
Share

Tools for This Entry

MusicProductionWiki.com
◆ The Producer's Bible
Interactive Tool
LUFS Target and Loudness Penalty Reference
See exactly how much each streaming platform reduces your track. Enter your integrated LUFS, select genre — get per-platform penalties, mastering targets, and delivery guidance.
Platform Loudness Penalties
Genre Mastering Target
Key LUFS Benchmarks
Spotify / YouTube / Tidal-14 LUFS
Apple Music-16 LUFS
Podcast (Apple / Spotify)-16 LUFS
Netflix / VOD delivery-27 LUFS
Broadcast EBU R128-23 LUFS
True Peak ceiling (streaming)-1 dBTP
LUFS measures perceived loudness over time. Platforms normalize loud masters down — mastering louder than target means the platform turns your track down, potentially removing punch. Always check True Peak (-1 dBTP minimum) separately from sample peak before delivery.
◆ The Producer's Bible — MusicProductionWiki.com𝕏 ShareReddit

Signal Chain Position

Signal chain position of Limiter in music production Instrument / Source Raw audio or MIDI pre-processing Clip Gain / Trim Pre-fader level adjustment EQ Tonal shaping before dynamics Compression Dynamic control range shaping Saturation / FX Color and texture creative processing Fader / Automation Volume rides and sends Mix Bus / Master Bus Bus compression and EQ Limiter Ceiling enforcement loudness maximizing ◀ YOU ARE HERE
Instrument / Source
Raw audio or MIDI · pre-processing
Clip Gain / Trim
Pre-fader level · adjustment
EQ
Tonal shaping · before dynamics
Compression
Dynamic control · range shaping
Saturation / FX
Color and texture · creative processing
Fader / Automation
Volume rides · and sends
Mix Bus / Master Bus
Bus compression · and EQ
Limiter
Ceiling enforcement · loudness maximizing
▶ You are here

The limiter lives at the absolute end of every signal chain it occupies — last plugin on the master bus, last processor in the mastering chain, last insert on any track bus where it is used for protection. This is not convention; it is physics. Any processing applied after the limiter can raise the signal above the ceiling you just enforced. If a stereo width processor, harmonic exciter, or subtle EQ shelf follows the limiter in the chain, those processors can push inter-sample peaks above the output ceiling you set, negating the entire function of the limiter. On the master bus, the limiter is always last — no exceptions for convenience or workflow preference.

Interaction Warnings

  • Upstream Saturation / Harmonic Distortion: Saturation and harmonic distortion plugins add odd and even harmonics that increase the RMS energy of a signal even when peak levels remain similar. Running saturation before a mastering limiter means the limiter is receiving a denser, hotter signal than your pre-saturation metering suggested — you will hit more gain reduction than expected. Meter post-saturation before the limiter, not pre-saturation, and recalibrate your threshold setting accordingly.
  • Bus Compression Interaction: Bus compression upstream of the limiter controls the dynamic range of the summed mix before the limiter sees it. Heavy bus compression followed by aggressive limiting is the fastest path to an audibly over-limited, lifeless master. Use the bus compressor to shape the dynamic contour of the mix — typically 2–4dB of gain reduction — and leave the limiter to catch only the remaining peaks. If the limiter is doing 6dB+ of work after bus compression, the bus compressor is not doing enough.
  • EQ After Limiting (True Peak Violation): Any EQ applied after the limiter — even a tiny 0.5dB high-frequency shelf — can create inter-sample peaks that exceed your true peak ceiling. Many mastering engineers use a linear-phase EQ before the limiter precisely to avoid this problem. If you need to EQ after limiting (e.g., a dull mix needing air), add the EQ, then run the signal through a second limiter stage set to catch the resulting peaks. Never assume a post-limiter EQ boost is safe at any amount.

History of Limiting

Limiting emerged not from music production but from broadcast engineering. In the 1930s and 1940s, AM radio transmission required that audio signals never exceed the transmitter's modulation capacity — an overmodulated transmitter caused signal splatter across adjacent frequencies, which was both illegal and audibly catastrophic for listeners. Broadcast engineers at RCA, CBS, and the BBC developed early limiting amplifiers that would clamp the audio signal below the danger point automatically. These early units were crude — heavy, transformer-coupled, relying on tube circuits with slow, program-dependent characteristics — but they solved the fundamental problem of preventing uncontrolled peaks. The Fairchild 670, developed in the 1950s, became the most sophisticated of these early limiters, using a combination of variable-mu tubes and program-dependent time constants that gave it a musicality no subsequent solid-state design fully replicated.

The hardware golden age of limiting in music production arrived with the 1176 LN in 1967, designed by Bill Putnam at Universal Audio. The 1176 used a Field-Effect Transistor as its gain reduction element — faster and more reliable than the tube circuits that preceded it — and introduced the now-legendary "all buttons in" mode: an undocumented configuration that engaged all four ratio buttons simultaneously, producing a distorted, hyper-aggressive character that became synonymous with drum room sound in 1970s rock. Engineers at Record Plant, Sound City, and Olympic Studios discovered this accident and weaponized it across hundreds of records. The Neve 33609, the dbx 160, and the SSL G-Series bus compressor followed, each carving out specific roles — the 33609 for transparent peak limiting on individual sources, the dbx 160 for aggressive VCA-driven gain reduction on drums and bass.

The digital era introduced precision that hardware limiters could never match and sterility that engineers immediately resented. Software limiters in the early DAW era — Waves L1 in 1994, then L2 in 1996 — achieved genuinely brickwall behavior using look-ahead buffering and IDR dithering algorithms that hardware could not approach. The L2 became the industry standard master bus limiter of the late 1990s and 2000s, appearing on virtually every commercially mastered record during the loudness normalization era preceding streaming normalization. The problem was that the precision invited abuse — engineers could now push 10, 12, even 15dB of gain reduction through the L2 and deliver a file that technically never exceeded 0 dBFS while sounding like a wall of compressed noise. The response was plugin emulation: iZotope, FabFilter, and Softube began modeling the non-linear behavior of analog hardware, reintroducing saturation and program-dependent characteristics into digital limiting.

The arrival of streaming platform normalization — Spotify in 2013, Apple Music and YouTube in 2015, Tidal following shortly after — fundamentally changed the strategic value of aggressive limiting. LUFS-based normalization means that a master pushed to -7 LUFS and a master mastered to -14 LUFS are played back at the same perceived loudness. Over-limited masters are turned down, losing the competitive loudness they were over-limited to achieve, while retaining all the dynamic damage the aggressive limiting caused. Mastering engineers now target -14 LUFS integrated for Spotify, -16 LUFS for podcast platforms, and -1.0 dBTP as the true peak ceiling across all major destinations. The limiter's role shifted from loudness maximizer back to its original function: ceiling enforcer. The dynamic range that the loudness war destroyed is slowly being reclaimed.

"The loudness war destroyed dynamic range in music for twenty years. LUFS normalization gave it back. Now the question is whether producers will use that space."

— Bob Katz, Mastering Engineer — author of Mastering Audio — Mastering Audio: The Art and the Science — Third Edition

Limiting evolved from a broadcast safety device into the central competitive tool of the loudness war, and has now been realigned by streaming normalization into its original purpose: an intelligent ceiling, not a loudness machine.

How Producers Use Limiting

The workflow for mastering-grade limiting begins before you open the limiter plugin. Your mix needs to hit the limiter's input with no more than 3–4dB of peak material above your target ceiling — if the kick and snare are driving the loudest peaks and they are 8dB above where you want the ceiling, those 8dB need to be addressed through gain staging, bus compression, or targeted peak control before the limiter. Set your output ceiling first: -1.0 dBTP for streaming delivery, with true peak mode and 4x oversampling enabled. Then raise the input (or lower the threshold) until you see 2–4dB of gain reduction on the loudest sustained passages — not peaks, sustained passages. If the gain reduction meter is jumping to 6dB only on occasional transients and sitting at 1–2dB everywhere else, you have found the right threshold. If you are seeing sustained 6dB+ gain reduction across the majority of the track, you are over-limiting and you need to bring level down upstream before the limiter.

On individual tracks and buses, a limiter functions differently than it does at the master stage. A limiter on a vocal track after the compressor is a peak catcher, not a loudness tool — set the ceiling 2–3dB above where the compressed vocal typically sits, so the limiter only engages on the loudest consonants and the occasional shout or jump. The release time here matters more than at the master stage: too slow and the vocal sounds ducked after every hard consonant; too fast and you hear the gain snapping back as a click or a distortion artifact on sibilant material. A limiter on a drum bus works similarly — set the ceiling so only the occasional rogue hit engages gain reduction, and use a slower attack (5–10ms) deliberately to let the transient crack pass through before the limiter clamps the body of the hit. That transient survival is the difference between a drum bus that hits and one that thwacks.

1. Open your Master track and navigate to the end of your effects chain. 2. Click 'Audio Effects' and insert 'Limiter' (native Ableton device) as the last plugin in the chain. 3. Set the 'Ceiling' parameter to -1.00 dB (this is a true peak ceiling in Ableton's Limiter). 4. Enable 'Look-Ahead' by clicking the LAH button. 5. Lower the 'Gain' knob until the gain reduction meter shows 1–3dB of movement on the loudest peaks. 6. Monitor the output level meter in Ableton's master track to confirm no output exceeds -1dBTP. 7. Use the integrated LUFS meter in Ableton 11/12 (View > LUFS Meter) to verify your integrated LUFS target. For mastering, consider third-party limiters like FabFilter Pro-L 2 for more detailed metering and control.

1. On your Stereo Out channel strip in Logic Pro, click the last insert slot and select 'Dynamics > Adaptive Limiter' or 'Dynamics > Limiter' from the plugin menu. 2. Use Logic's Adaptive Limiter for mastering — set 'Out Ceiling' to -1.0 dB and 'Gain' to achieve 2–4dB of gain reduction on peaks. 3. Enable 'Optimize for Loudness' if targeting streaming LUFS targets automatically, or set 'Lookahead Time' to 2ms for manual control. 4. Alternatively, insert a third-party limiter (FabFilter Pro-L 2, iZotope Ozone Maximizer) in the last slot. 5. Open the Loudness Meter plugin (Utilities > Loudness Meter) on the Stereo Out to monitor LUFS integrated and true-peak levels in real-time. 6. Bounce the master using File > Bounce > Project or Section, ensuring 'Normalize' is set to 'Off' to preserve your limiter's output ceiling.

1. Open the Master Mixer track (master channel in FL Studio's Mixer). 2. Click an empty slot in the Effects chain on the far right of the Master track and select 'Fruity Peak Controller' — no, instead select your limiter plugin: navigate to Effects > Dynamics > 'Fruity Peak Controller' is incorrect; use 'Maximus' (FL Studio's native multiband maximizer/limiter) or insert a third-party VST limiter. 3. In Maximus, click the 'Master' band tab. Set 'Ceiling' to -1.0 dB. Lower the 'Volume' knob to push the signal into the limiter until you see 1–3dB of gain reduction on the GR meter. 4. Enable 'Soft Knee' for more transparent behavior on sustained material. 5. For third-party limiters: right-click any Mixer slot, select 'Insert,' and load your plugin (e.g., FabFilter Pro-L 2). Place it as the absolute last effect in the chain. 6. Use Edison or the Playlist's Export function (File > Export > Wave File) to render your master, monitoring the peak meter to confirm the ceiling is held at -1.0 dBTP.

1. On your Master Fader track, navigate to the Inserts section and click the last available insert slot. 2. Select 'Plug-In > Dynamics > Limiter' — Pro Tools ships with BF-2A and other Bomb Factory dynamics, or use a third-party limiter via AAX format (Waves L2, FabFilter Pro-L 2 AAX, Sonnox Oxford Limiter). 3. Set the output ceiling to -1.0 dBTP — if using a true-peak-capable plugin, engage the ISP / True Peak mode. 4. Reduce the threshold or increase the input gain until the GR meter shows 1–3dB of movement on the loudest transients. 5. Insert the Avid LUFS Meter (if available in your Pro Tools version) after the limiter to monitor integrated and short-term LUFS. Alternatively, use iZotope Insight 2 for comprehensive loudness metering. 6. Bounce to disk using Bounce to Disk (Cmd+Alt+B) with the Master Fader as the source, ensuring 'Dither' is placed after the limiter if dithering to 16-bit delivery.

The diagnostic for a well-set limiter is A/B comparison at matched loudness. Bypass the limiter and reduce the playback volume to match the perceived loudness of the limited signal. If the bypass version sounds more alive, more spacious, and more dynamic while perceived loudness is equal, you are over-limiting — the limiter is removing dynamic information that was delivering emotional impact. If the bypass version sounds similar in character but slightly more erratic in level, with occasional peaks that feel uncontrolled, the limiter is doing exactly what it should: catching outliers without affecting the body of the signal. The bypass test at matched loudness is the only objective diagnostic for whether your limiter setting is transparent or destructive.

What to listen for as the limiter engages: the kick drum is your primary reference point. A well-limited kick retains its attack transient — the initial click or punch that defines the hit — and has its body and sustain controlled by the limiter. An over-limited kick sounds like the transient has been rounded off, the attack smeared into a softer, less defined impact, and the low-frequency sustain compressed into a flat blob. The snare should retain its crack — that sharp initial spike that tells you a drum was physically struck — and only have its room tail and body controlled. If the snare sounds dull or pillowed rather than crisp, your attack time is too short or your gain reduction is too deep. Bass notes should retain their fundamental pitch and harmonic definition; if the bass starts sounding like a constant hum rather than distinct notes, the release time is too slow or the gain reduction is interacting with the low-frequency content in a way that needs to be addressed upstream.

Set the output ceiling first, raise input until 2–4dB of sustained gain reduction appears, then use the kick and snare transients as diagnostic tools to confirm the limiter is catching peaks rather than flattening the dynamics that make the track feel physical.

Limiting by Genre

Genre conventions around limiting are not arbitrary — they reflect the listening contexts, emotional expectations, and competitive loudness standards of each market. A metal master that hits -8 LUFS is doing exactly what the genre demands; a jazz master at the same level has been destroyed. Understanding where each genre sits on the loudness-dynamics spectrum is the difference between a master that fits its context and one that fights it.

GenreRatioAttackReleaseThresholdNotes
Trap∞:1<0.1ms (look-ahead)Auto / 30–80ms-6 to -10 dBFSBrickwall ceiling at -1.0 dBTP; pre-limit 808 and hi-hat peaks with clip gain to prevent the limiter from monopolizing GR on sub transients
Hip-Hop∞:1<0.1ms (look-ahead)Auto / 50–100ms-4 to -8 dBFSPreserve kick and snare punch; no more than 3–4dB GR; upstream bus compression should do the heavy dynamics work
House∞:1<0.1ms (look-ahead)Auto / 40–80ms-4 to -6 dBFSWatch for kick-driven pumping; use mid-side limiting or pre-limit the low bus separately to keep consistent gain reduction across the mix
Rock∞:1<0.1ms (look-ahead)Auto / 60–120ms-3 to -6 dBFSDense guitar transients can cause inter-sample overs — true-peak mode is essential; target -10 to -12 LUFS for rock/metal delivery
Mastering∞:1<0.1ms (look-ahead 1–3ms)Auto-1 to -4 dBFSMaximum 3–4dB GR; true-peak ceiling at -1.0 dBTP; target -14 LUFS integrated for Spotify/YouTube; -16 LUFS for Apple Music
Share

Deviate from genre conventions when the production's emotional intention conflicts with the expectation — a deliberately lo-fi hip-hop track mastered conservatively at -16 LUFS is making a statement; an EDM track mastered at -18 LUFS is probably a mistake. Always check your target platform's normalization level against your genre's expectation before committing to a ceiling and loudness target, because the platform's algorithm will determine what the listener actually hears, regardless of what your meters read in the session.

Hardware vs Plugin vs Stock

The gap between hardware and plugin limiting has narrowed dramatically since 2015, but it has not closed. Hardware limiters — the Neve 33609, the Manley Variable Mu, the Chandler Limited TG1 — introduce transformer saturation, tube harmonics, and circuit-level non-linearity into the gain reduction process. These are not bugs. They are colorations that interact with the limiting behavior in ways that make heavy gain reduction sound musical rather than clinical. A digital brickwall limiter at 6dB of gain reduction sounds like 6dB of gain reduction. A Fairchild 670 at 6dB of gain reduction sounds like someone opened a room. The plugin world has responded with sophisticated modeling — FabFilter Pro-L 2, iZotope Ozone Maximizer, and Sonnox Limiter offer transparent algorithmic precision that hardware cannot match in controllability, while Slate Digital's VBC and Waves CLA-76 attempt to recapture the non-linear warmth through circuit emulation. The honest answer is: for transparent true-peak mastering delivery, the best plugins outperform most hardware. For character and color in limiting, hardware still wins in ways that are immediately audible.

AspectHardwarePlugin
True Peak ControlCannot measure or enforce true peak; relies on analog headroomFull true peak metering and ceiling enforcement with oversampling
Sonic CharacterTransformer saturation, tube harmonics, and circuit warmth at gain reductionAlgorithmically transparent or modeled analog color depending on design
Precision / RepeatabilityComponent drift, temperature sensitivity; no exact recallPerfect recall, total automation, no drift
Look-AheadNot available; reacts to signal after it arrives1–10ms look-ahead buffers standard on all mastering limiters
OversamplingN/A — analog domain has no sample gridUp to 32x oversampling available for inter-sample peak detection
Cost / Access$2,000–$20,000+ for quality hardware limiters$0 (stock DAW) to $200 for industry-standard plugins
Free Tier
Limiter No6 Tokyo Dawn Labs
SAFE Limiter Semantic Audio
Mid Tier
Ceilings Mastering The Mix
Invisible Limiter G2 A.O.M.
Pro Tier
Pro-L 2 FabFilter
Ozone Maximizer iZotope
Oxford Limiter V3 Sonnox

Use hardware limiting when you are deliberately chasing the warmth and color of a specific circuit's gain reduction character — and you have budget for maintenance and calibration. Use plugin limiters for all mastering delivery work where true peak compliance, oversampling precision, and perfect recall are non-negotiable. The stock DAW limiter is adequate for peak protection on individual tracks; for mastering, FabFilter Pro-L 2, iZotope Ozone Maximizer, or Sonnox Limiter at 4x oversampling with true peak mode represent the current professional baseline.

Before and After

Before

Without proper limiting, the master may contain peak transients that clip the output stage or breach 0 dBFS during D/A conversion, causing harsh digital distortion on playback. The mix may also sound dynamically inconsistent — quiet in the context of other commercial releases when played back on streaming platforms.

After

With correct limiting, the output ceiling is enforced at -1.0 dBTP with transparent gain reduction of 2–3dB on peaks — the mix sounds cohesive, competitive in loudness against commercial releases, and free from digital clipping artifacts across all playback systems and after lossy encoding.

As you engage the limiter and begin raising the input level, listen specifically to the transient behavior of the loudest elements first. A kick drum that previously had a 6dB peak above the body of the mix should retain its physical impact — the felt thud rather than just the frequency — as the limiter catches the ceiling. If that physical impact disappears and the kick flattens into the mix, you have passed the threshold where limiting is transparent. Vocally, a limited mix should sound controlled rather than compressed — the vocal sits in the mix without jumping, but its breath and consonant detail remain intact. The moment vowels start sounding like they are behind a slight blanket and consonants lose their edge, back the input off 1–2dB and reassess from that position.

Limiting In The Wild

These seven tracks represent the full spectrum of limiting decisions in commercial production — from the textbook damage of aggressive loudness maximization to the strategic restraint of streaming-optimized mastering. Listen to each one with the gain reduction meter behavior in mind: ask yourself how much work the limiter is doing, and whether the result serves the production or fights it.

Kanye WestPOWER (2010), My Beautiful Dark Twisted Fantasy. Produced by Kanye West, S1.
Notice the immediate, wall-of-sound density from the intro — the master limiter is pushed hard against the sample and drums, creating a relentless, crushed loudness. The transients are controlled but the track still breathes, demonstrating skilled brickwall limiting on a commercial master.
MetallicaDeath Magnetic (2008), Death Magnetic. Produced by Rick Rubin.
An infamous example of over-limiting: listen for the complete absence of dynamic range as the guitars and drums sound perpetually clipped and distorted due to the limiter ceiling being set far too aggressively. This track became a textbook case study in the dangers of competitive loudness maximization.
Daft PunkGet Lucky (2013), Random Access Memories. Produced by Daft Punk, Pharrell Williams.
The master limiter is deployed with surgical restraint — peaks are controlled and the LUFS target is met while the groove and dynamics remain fully intact. Listen for how the kick and bass interplay retains punch and clarity, a hallmark of transparent limiting done right.
Dr. DreStill D.R.E. (1999), 2001. Produced by Dr. Dre, Scott Storch.
The limiter on this master creates cohesion and density across the piano loop, 808, and vocal without audible pumping or distortion. Notice how the overall loudness feels competitive for its era while transients in the snare remain crisp and defined — hallmarks of well-calibrated limiting.
Billie Eilishbad guy (2019), WHEN WE ALL FALL ASLEEP, WHERE DO WE GO?. Produced by Finneas O'Connell.
Finneas used the limiter conservatively, targeting streaming normalization rather than maximum loudness — the result is a mix that actually sounds louder perceptually due to its dynamic contrast. Listen to how the sub bass and vocal breathe freely, demonstrating that smart limiting can outperform aggressive limiting on modern streaming platforms.
Kendrick LamarHUMBLE. (2017), DAMN.. Produced by Mike Will Made-It.
The stark, minimalist production allows the limiter to catch only the hardest transient peaks without squashing the intentional space in the mix. Notice how the snare crack retains its physical impact — the limiter ceiling was set conservatively, letting inter-sample peaks be caught by true-peak limiting rather than hard clipping.
The WeekndBlinding Lights (2019), After Hours. Produced by DaHeala, Metro Boomin, Oscar Holter, Max Martin.
The synth-driven mix hits a tight output ceiling while maintaining the retro-tinged movement of the production — listen for how the synth arpeggios maintain their rhythmic envelope despite the high average loudness. This demonstrates mastering-grade limiting where density and transient integrity coexist at a streaming-competitive level.

The lesson across these seven tracks is that the limiter's audibility is always a choice, never an accident. Metallica's Death Magnetic made the limiter audible by pushing it past all reasonable limits in pursuit of competitive loudness — and the resulting distortion became a cultural case study. Billie Eilish's "bad guy" made the limiter invisible by refusing to push it, and the track sounds louder than records with three times its dB level because Finneas understood that perceived loudness is a psychoacoustic phenomenon, not a meter reading. The limiter is as transparent or as characterful as the decisions made in the 60 minutes before the session ends.

Types of Limiting

Limiting vs Compression

See the full comparison: Compression

Limiting vs Saturation

See the full comparison: Saturation

Not all limiters perform the same function, and choosing the wrong type for the job produces either inadequate peak control, unnecessary coloration, or audible artifacts that undermine the mix. The distinctions between limiter types are not marketing categories — they reflect genuinely different gain reduction mechanisms and time constant behaviors that produce different sonic results on the same material.

Brickwall Limiter FabFilter Pro-L 2 / iZotope Ozone Maximizer

The definitive mastering limiter type — ∞:1 ratio with look-ahead buffering and true peak detection that prevents any signal from exceeding the output ceiling under any condition. Use this as the final processor in every mastering chain and as the last insert on any mix bus where delivery compliance matters. Not appropriate for track-level limiting where some transient slippage is musically desirable.

VCA Peak Limiter dbx 160 / SSL G-Series (in limiting mode)

Voltage-Controlled Amplifier-based limiters are the fastest and most aggressive of the hardware types — capable of near-instantaneous gain reduction that catches transients before they escape. The cost is hard-knee gain reduction that can sound punchy and exciting on drums but harsh on full mixes. Use VCA limiters on drum buses and individual aggressive sources where speed matters more than transparency.

Optical Limiter Teletronix LA-2A / Neve 33609

Optical limiters use a light-dependent resistor as the gain reduction element — inherently program-dependent because the optical cell's response speed varies with signal content. The result is a musically intelligent gain reduction character that sounds like someone with excellent taste is riding a fader. Ideal for vocals, acoustic instruments, and full mix limiting where you want the dynamics shaped rather than clamped. Not appropriate for strict peak control.

Variable-Mu / Tube Limiter Fairchild 670 / Manley Variable Mu

The oldest limiter topology and still the most musical at heavy gain reduction — variable-mu tube designs use the tube's non-linear gain characteristics to produce saturation and compression simultaneously. Heavy limiting through a Fairchild 670 creates harmonic density rather than dynamic flattening. Use for mastering stages where character and warmth are priorities, always followed by a brickwall plugin limiter for true peak compliance.

Multiband Limiter Waves L3 Multimaximizer / iZotope Ozone Maximizer (multiband mode)

Splits the signal into frequency bands and applies independent limiting to each — the low end can be clamped aggressively without affecting the high-frequency limiting character. Critical for bass-heavy genres where a full-band limiter would reduce the entire mix every time the kick or sub hits. The risk is over-processing: too much multiband limiting creates an unnatural spectral consistency where nothing in the mix moves relative to anything else.

Clipper / Soft Clipper Kazrog True Iron / FabFilter Saturn (clip mode) / Decapitator

Not a traditional limiter but functions as a transient ceiling control through waveform clipping rather than gain reduction. A soft clipper rounds the tops of peaks rather than attenuating them, introducing harmonic distortion that adds density and perceived loudness without reducing dynamic range as aggressively as a standard limiter. Use before the brickwall limiter to control transients and reduce the amount of gain reduction the brickwall limiter must apply — this is a primary technique in modern loudness maximization for rock and electronic music.

Match the limiter type to the function — brickwall for delivery compliance and true peak, VCA for speed on drums and aggressive sources, optical and variable-mu for musical character, multiband for genre-specific frequency control, and clipper-plus-brickwall for maximum loudness with minimum dynamic damage.

The Producer's Verdict

The most expensive mistake in modern mastering is treating the limiter as the starting point of loudness rather than the endpoint of it. If your limiter is working hard, your mix is not ready — period. Every decibel of gain reduction above 4dB is a dynamic decision being made by an algorithm instead of by you, and algorithms do not know that the snare crack matters more than the average level. Make your loudness decisions in the mix through compression, saturation, and intelligent arrangement before the limiter ever touches the signal. The limiter's job is to catch the last 2–3dB, not the first 10.

Use WhenPeak control and delivery compliance are requiredMaster bus, track bus safety, or streaming delivery — wherever a hard ceiling is a technical or creative requirement rather than a preference.
Avoid WhenThe mix hasn't been gain-staged properly upstreamA limiter on a poorly gain-staged mix will work twice as hard and sound half as good — fix gain staging and bus compression first, then apply the limiter.
Starting Ratio∞:1 with 1ms look-ahead, true peak mode onFor mastering, always brickwall with true peak and minimum 4x oversampling; for track limiting, 10:1 with 5ms attack is a sensible starting point before dialing by ear.
Watch ForSustained GR above 4dB and kick transient roundingIf the gain reduction meter is pegged above 4dB on the loudest passages or the kick drum sounds pillowed rather than punchy, you are past the transparent zone.
Pairs WithBus compression + soft clipper upstreamBus compression shapes the dynamic contour, soft clipping controls transient peaks, and the brickwall limiter catches the remainder — this three-stage approach minimizes audible limiting artifacts.
Common ErrorSetting output ceiling to 0 dBFS for streaming deliveryCodec encoding creates inter-sample peaks that exceed 0 dBFS — always set the true peak ceiling to -1.0 dBTP minimum for any file going through lossy encoding.
Share This Verdict

The limiter is the last honest tool in the chain — it tells you exactly how much work you left undone in the mix by how hard it has to work to deliver the file.

Common Mistakes with Limiting

Limiting mistakes are almost always upstream problems that get diagnosed at the limiter. The limiter is the place where every bad gain staging decision, every over-compressed bus, and every mix that was never finished properly becomes audible — because the limiter is the only processor in the chain with nowhere left to send the problem. Recognizing that the mistake is rarely in the limiter itself is the first step to fixing it correctly.

Setting the Output Ceiling to 0 dBFS for Streaming

A 0 dBFS ceiling in your DAW reads clean because your DAW only measures sample values. When that file is encoded to MP3, AAC, or OGG by the streaming platform, inter-sample peaks — peaks that form between samples during digital reconstruction — are created above 0 dBFS, producing real distortion in the listener's playback. Set true peak ceiling to -1.0 dBTP minimum, enable true peak mode in your limiter, and use 4x oversampling. This is not optional for any professional delivery; it is the minimum specification for files going to Spotify, Apple Music, or YouTube.

Using the Limiter as the Primary Loudness Tool

Pushing 8, 10, or 12dB of gain reduction through a brickwall limiter in pursuit of competitive loudness is the fastest path to a dynamically dead master that streaming normalization will turn down to the same level as a well-mastered track at -14 LUFS. The loudness competition the limiter was being weaponized for no longer exists on streaming platforms. Use parallel compression, saturation, and careful headroom management in the mix to build perceived loudness before the limiter sees the signal.

Skipping True Peak Mode on Mastering Limiters

Standard peak limiting measures sample values only. True peak mode oversamples the signal, interpolates between samples, and measures the actual reconstructed peak values that a DAC will produce. Without true peak mode, you can deliver a file that reads -1.0 dBFS on every meter in your session and still contains inter-sample peaks of +0.3 dBTP or higher. Enable true peak mode on every mastering limiter, every time, for every streaming delivery. There is no production context where true peak mode on a mastering limiter is the wrong choice.

Applying Processing After the Limiter

Any EQ, stereo width, harmonic exciter, or any other processor placed after the limiter in the chain can raise the signal above the ceiling you just enforced. A 1dB high-frequency shelf after a brickwall limiter will push high-frequency inter-sample peaks above your ceiling, negating the limiter's function entirely. The limiter is always last. If you discover you need EQ after limiting, add the EQ, then add a second limiter stage after it set to the same ceiling. Never remove the limiter from the last position.

Not Checking at Matched Loudness

Bypass comparisons without level matching are meaningless — the louder version always sounds better because human hearing perceives louder as higher quality at up to 1dB difference. To diagnose whether your limiter is helping or hurting, bypass it and manually reduce the playback volume to match the limited version's perceived loudness. At matched levels, if the bypass sounds more alive and dynamic, you are over-limiting. If they sound comparable in character with the limited version having better controlled peaks, the limiter is set correctly. Every A/B comparison involving a limiter must be level-matched.

Ignoring Release Time on Low-Frequency Heavy Material

A fast release time on a limiter receiving bass-heavy material — kick-heavy hip-hop, sub-driven EDM, heavy metal with tuned low end — creates gain modulation at the rate of the bass notes, which is in the audible frequency range. At 120 BPM with 8th-note bass content, a 50ms release creates modulation at around 4Hz — audible as a low-frequency wobble or pumping artifact. Set release to 100ms or use program-dependent release on bass-heavy material, and consider using a multiband limiter or addressing the low-end transient behavior through a clipper before the brickwall limiter stage.

Every audible limiting artifact is a symptom of a decision made upstream — fix the gain staging, the bus compression, and the mix balance first, and the limiter will operate transparently on what remains.

Red Flags and Green Flags

Red Flags

  • 🔴 Consistent gain reduction of 6dB or more on the master limiter — the mix needs more dynamic control upstream before the limiter stage.
  • 🔴 Audible pumping or 'breathing' on sustained elements like pads or reverb tails, indicating the limiter's release time is too fast for the program material.
  • 🔴 Distortion or harshness on kick transients and snare hits — the limiter is being asked to catch unmanaged transient peaks that should have been addressed with compression or clip gain earlier in the chain.

Green Flags

  • 🟢 Gain reduction meter moving only on the loudest transient peaks, staying at 0dB the majority of the time — the limiter is working as a safety net, not a loudness engine.
  • 🟢 The limited mix sounds perceptually identical to the pre-limited signal when A/B'd at matched loudness, with no audible coloration or pumping.
  • 🟢 True-peak output is consistently at or below -1.0 dBTP across all sample rates, ensuring zero inter-sample overs on all playback systems and streaming platforms.

When you see red flags in your limiting — consistent gain reduction above 4dB, kick transients that sound rounded rather than sharp, bass notes that blur into each other rather than remaining distinct — the diagnostic path always runs upstream. Check the mix bus for over-compression that is compressing peaks before the limiter sees them, leaving no headroom for the limiter to work transparently. Check individual track buses for elements that are louder than the mix balance requires. Check whether a saturation plugin upstream is raising RMS energy without raising peaks, meaning the limiter is receiving a denser signal than your pre-saturation metering suggested. The limiter is the last place to look for the problem; it is almost always the first place the problem becomes audible.

Your Progression with Limiting

The three stages of development with limiting mirror the three stages of understanding dynamics broadly: first you learn to use it correctly as a safety device, then you learn to integrate it into a chain where its workload is minimal, then you learn to deploy it as a deliberate creative choice with full awareness of the tradeoffs you are making at each dB of gain reduction. Every producer who has ever over-limited a master had the right instinct — more loudness, more presence, more impact — but applied it at the wrong point in the signal chain with the wrong tool. Understanding where each stage's knowledge gap lies is what makes progression from beginner to advanced something you can accelerate deliberately.

Beginner

Insert a brickwall limiter as the absolute last plugin on your master bus, set the output ceiling to -1.0 dBTP, and gently lower the threshold until you achieve no more than 1–2dB of gain reduction on peaks. This gets you clipping protection without audible artifacts while you learn to hear what the limiter is doing. Enable true peak mode and 4x oversampling immediately — these settings should never be off in a mastering context, even at the beginner stage, because the consequences of ignoring them are invisible in the session and audible everywhere the file is played back.

Intermediate

Learn to use the limiter in conjunction with proper gain staging and bus compression upstream — aim for only 2–4dB of gain reduction from the limiter while hitting target LUFS for your primary streaming platform. Study the LUFS integrated reading alongside the gain reduction meter to understand the relationship between loudness and dynamic damage. Introduce a soft clipper before the brickwall limiter to pre-shape transients and reduce the limiter's workload. A/B your limited masters against commercial references at matched LUFS levels — not matched peak levels — to calibrate your loudness targets against what the genre actually requires.

Advanced

At the advanced stage, the limiter is the last decision in a chain of deliberate choices — you are selecting limiter algorithm character based on the music's emotional content, comparing brickwall transparency against soft clipper density, and adjusting release curves to preserve specific transient behavior in the mix. Advanced use includes using the limiter's gain reduction behavior as a diagnostic tool: watching which elements are driving the most gain reduction tells you which elements need attention in the mix before the mastering stage. The advanced practitioner never reaches the mastering chain limiter without knowing exactly how much gain reduction to expect and why.

Progression with limiting is progression with the entire dynamics chain — the limiter only becomes a transparent tool when everything upstream has been addressed with enough precision that the limiter has almost nothing left to do.

Frequently Asked Questions

A compressor reduces gain at a controlled ratio (e.g., 4:1) above the threshold, allowing some dynamic variation to pass through. A limiter uses a very high ratio — typically 10:1 to ∞:1 — effectively preventing any signal from exceeding the threshold ceiling. In practice, limiters are used for absolute peak control and loudness maximization, while compressors shape dynamics more musically.

The brickwall limiter should always be the absolute last plugin in the master bus chain — after EQ, bus compression, saturation, and stereo imaging. Its job is to catch any peaks that escape all upstream processing and enforce the final output ceiling. Placing any processing after the limiter risks introducing new peaks that exceed your ceiling.

For most streaming platforms including Spotify, Apple Music, and YouTube, a true-peak ceiling of -1.0 dBTP is the standard. Some engineers use -1.5 dBTP or -2.0 dBTP to provide additional headroom for MP3/AAC encoding artifacts, which can add 0.5–1.5dB to inter-sample peaks after lossy compression. Always use a true-peak limiter, not just a sample-peak limiter, for streaming masters.

A general professional guideline is that if the limiter is applying more than 3–4dB of consistent gain reduction, the mix needs additional dynamic control upstream rather than forcing the limiter to do heavy lifting. More than 6dB of gain reduction from a brickwall limiter will almost always introduce audible distortion, transient smearing, or pumping artifacts.

True-peak limiting accounts for inter-sample peaks — values that occur between digital samples and can exceed 0dBFS during digital-to-analog conversion or lossy encoding, even if no individual sample clips. Standard sample-peak limiters miss these overs, which can cause distortion on playback devices and streaming platform rejection. Always use a true-peak-compliant limiter for any final master destined for release.

Limiters on individual tracks are legitimate tools — using them on drum bus channels, synth leads, or bass tracks to catch errant transients before they hit the mix bus is common practice. However, track-level limiting should be used for control and protection, not for loudness; reserve loudness maximization exclusively for the master bus stage after all mixing decisions are finalized.

Look-ahead causes the limiter to delay the audio output by a few milliseconds so it can analyze upcoming peaks and begin applying gain reduction before those peaks arrive, resulting in more transparent limiting with less transient distortion. This introduces a small amount of latency, which is irrelevant in mastering but can cause issues when monitoring in real-time. Use look-ahead whenever you can — most modern mastering limiters have it enabled by default.

Spotify normalizes to approximately -14 LUFS integrated; Apple Music normalizes to -16 LUFS; YouTube normalizes to -14 LUFS. Mastering louder than these targets means the platform will simply turn your track down, wasting any extra limiting. Targeting -14 LUFS integrated with a -1.0 dBTP ceiling is the current professional standard for most streaming-destined masters, giving you maximum perceived loudness without over-limiting.

What level did this entry match?

Also in The Bible

The Producer's Briefing
The Producer's Briefing — practical technique, gear intel, no fluff.