Why Order Matters in Mastering
Every processor in a mastering chain takes its input from the processor before it. This seems obvious, but the implications are profound. An EQ boost at 3kHz placed before a compressor will cause the compressor to react more aggressively to that frequency range, because the compressor now sees higher levels at 3kHz. The same EQ boost placed after the compressor will simply raise 3kHz without triggering any dynamic response. Same boost, completely different result, depending on where it sits in the chain.
In mixing, you have dozens of tracks and buses, and the order of plugins on each one matters but the consequences are localized. In mastering, you have one stereo bus. Every processing decision affects the entire frequency spectrum, the entire dynamic range, and the entire stereo field simultaneously. There is no "fixing it on the next track." What happens at each stage propagates through every stage that follows.
Professional mastering engineers have converged on a general chain order over decades of practice. There is room for variation -- every engineer has preferences -- but the broad structure is consistent because it is dictated by signal processing logic, not opinion.
The Typical Mastering Chain: An Overview
At the highest level, a mastering chain moves from correction to enhancement to control:
- Correction: Fix problems in the source material (DC offset, subsonic rumble, spectral imbalances)
- Dynamics: Control the dynamic range (compression, multiband, parallel processing)
- Character: Add tonal color and density (saturation, exciter, console emulation)
- Spatial: Shape the stereo image (stereo width, M/S processing)
- Limiting: Set the final loudness ceiling (limiter, clipper, true peak safety)
This order is not arbitrary. You correct problems first so downstream processors are not reacting to artifacts. You compress before you saturate so the saturation receives a dynamically controlled signal. You handle spatial processing after tonal shaping so the stereo image reflects the final spectral balance. And you limit last because the limiter's job is to catch the cumulative output of everything before it.
Stage 1: Correction First
DC Filter
A DC offset is a constant voltage bias in the audio signal that shifts the waveform up or down from the zero line. You cannot hear it, but it wastes headroom and causes every processor after it to work asymmetrically. A high-pass filter at 5-10 Hz removes DC offset without affecting any audible content. This goes first because every processor benefits from a centered waveform.
Subsonic Filter
Content below 20-30 Hz is inaudible but carries significant energy. Rumble from HVAC, microphone handling, or poorly managed bass synths eats headroom and causes compressors and limiters to react to energy you cannot hear. A subsonic high-pass filter, typically a gentle 12 dB/octave slope at 25-30 Hz, removes this energy. Placed early in the chain, it gives every downstream processor a cleaner signal to work with.
Corrective EQ
If the mix has spectral problems -- too much mud at 200-400 Hz, harshness at 2-5 kHz, excessive sibilance -- this is where you address them with surgical cuts. The EQ goes before dynamics processing for a specific reason: if you compress first, the compressor will react to the problematic frequencies, potentially making them worse or causing unmusical gain reduction patterns. Fix the spectrum first, then compress a balanced signal.
The rule: Corrective EQ goes before compression. Creative/tonal EQ can go after. If you are cutting problems, do it early. If you are adding flavor, do it later.
Stage 2: Dynamics Processing
Bus Compression
A stereo bus compressor is the backbone of most mastering chains. Placed after corrective EQ, it receives a spectrally balanced signal and applies gentle gain reduction (typically 1-3 dB) to add cohesion and control. Classic settings: slow attack (20-30ms) to preserve transients, auto or medium release, low ratio (1.5:1 to 2:1). The bus compressor makes the mix feel like one unified piece of music rather than a collection of elements.
Multiband Compression
Multiband compression splits the signal into frequency bands (typically 3-4) and compresses each independently. In mastering, it is used surgically: taming an inconsistent low end, controlling harsh vocal sibilance in the upper mids, or evening out a boomy bass that triggers the bus compressor. It goes after the bus compressor because you want the bus comp to glue the overall mix first, then use multiband to address specific frequency-range dynamics that the broadband compressor missed.
Parallel (NY) Compression
Parallel compression blends a heavily compressed signal with the dry original. It adds density and sustain to quiet elements without squashing transients. In the chain, it typically runs alongside or just after the main compressors. The dry signal preserves the natural dynamics, while the compressed signal fills in the gaps. The blend ratio determines how much "body" you add.
Upward Compression
Where standard compression pushes loud signals down, upward compression pulls quiet signals up. It increases the level of low-level detail -- room tone, reverb tails, quiet instrumental nuances -- without touching the peaks. Placed after the main dynamics stage, it adds perceived fullness without increasing the peak-to-loudness ratio. This is a subtle but powerful tool that many professional mastering chains use.
Stage 3: Character Processing
Saturation
Tape, tube, or transformer saturation adds harmonic overtones that fill out the spectrum and add perceived warmth and density. Placed after dynamics processing, it receives a dynamically controlled signal, which means the saturation behaves consistently rather than reacting wildly to transient peaks. Even-order harmonics (2nd, 4th) add warmth. Odd-order harmonics (3rd, 5th) add grit and presence.
Harmonic Exciter
An exciter generates upper harmonics (typically above 4-8 kHz) to add air and sparkle. Unlike EQ, which boosts what is already there, an exciter creates new harmonic content. Used subtly in mastering, it can add life to a dull mix without the phasing issues that aggressive high-shelf EQ can cause. It sits after saturation because the exciter's output benefits from the harmonically enriched signal the saturator provides.
Console Emulation
Console emulation models the cumulative nonlinearities of passing audio through an analog mixing console -- the subtle saturation of each channel strip, the crosstalk between channels, the transformer coloration. This is the "glue" that makes analog mixes sound cohesive. In the chain, it sits after other character processors to apply its shaping to the fully colored signal. The effect is subtle but audible: a sense of width, warmth, and spatial cohesion that digital processing alone does not achieve.
Stage 4: Spatial Processing
Stereo Width
Stereo width processing adjusts the relationship between the mid (center) and side (stereo difference) signals. Widening the sides makes the mix feel more expansive. Narrowing the sides makes it more focused. This goes after character processing because saturation and excitation can shift the stereo balance, and you want to make spatial adjustments based on the final tonal character, not a version that will change.
Mid/Side EQ
M/S EQ lets you equalize the center and sides independently. Common moves include cutting low frequencies from the sides (to keep bass centered for mono compatibility) and gently boosting highs on the sides (to add air and space). This sits near the end of the chain because it is making final spatial refinements to a signal that has already been dynamically controlled, harmonically shaped, and width-adjusted.
Stage 5: Final Limiting and Safety
Look-Ahead Limiter
The primary mastering limiter sits at the end of the chain. Its job is to catch every peak that exceeds your target ceiling and reduce gain smoothly. Look-ahead technology lets the limiter anticipate transients a few milliseconds in advance, resulting in transparent gain reduction without the distortion artifacts of reactive limiting. Set the ceiling to -1.0 dBTP and push the input gain until you reach your target LUFS.
Brickwall Limiter
A brickwall limiter guarantees absolute compliance with the ceiling. Some chains use a dedicated brickwall after the primary limiter as a second safety net, catching any transients that the look-ahead limiter does not fully contain. It should be doing almost no work -- if the brickwall is triggering constantly, the primary limiter needs to work harder.
Safety Clipper
The absolute last stage. A safety clipper hard-limits any stray sample that exceeds the true peak ceiling. In a well-set chain, it catches fractions of a dB at most. Its existence is insurance: even if something unexpected passes through the limiters, the clipper guarantees broadcast-compliant output. A ceiling of -1.0 dBTP is the broadcast standard, ensuring no clipping occurs during codec conversion on streaming platforms.
Three-stage limiting is not overkill. Each stage serves a different purpose: the look-ahead limiter handles musical gain reduction, the brickwall catches outliers, and the safety clipper guarantees absolute compliance. Together, they let you push loudness confidently without worrying about stray peaks.
LuvLang's 20-Stage Chain
LuvLang's mastering engine implements a full professional signal chain with 20 processing stages in the correct order. Here is the complete signal flow:
- DC Filter -- removes DC offset
- Subsonic Filter -- removes inaudible low-frequency energy
- Unlimiter -- restores dynamics to over-compressed mixes
- Bass Mono -- sums bass below threshold to mono for playback compatibility
- 7-Band EQ -- corrective and tonal equalization
- Resonance Notches -- surgical removal of resonant frequencies
- Dynamic EQ -- frequency-dependent compression
- Multiband Compression -- 4-band Linkwitz-Riley crossover network
- Bus Compressor -- broadband glue compression
- NY Compression -- parallel compression for body and density
- Upward Compressor -- lifts low-level detail
- Transient Shaper -- controls attack and sustain character
- Exciter -- harmonic enhancement
- Warmth -- tube-style saturation
- Console Emulation -- analog console modeling
- M/S EQ -- independent mid and side equalization
- De-Esser -- sibilance control
- HF Limiter -- high-frequency limiting for harshness control
- Stereo Width -- stereo image adjustment
- Look-Ahead Limiter + Brickwall + Safety Clipper -- three-stage final limiting at -1.0 dBTP
Every stage is genre-aware, with presets that adjust parameters based on the type of music being mastered. The chain mirrors what you would find in a top-tier mastering facility, running in real time in your browser.
Common Mistakes in Chain Ordering
These are the chain ordering mistakes that cause the most problems. Each one is fixable by understanding the logic above.
- EQ after the limiter. Any boost after the limiter pushes peaks above your ceiling, undoing the limiting. If you need to EQ after limiting, you need to limit again. Instead, put your EQ before the limiter.
- Compression before corrective EQ. If the mix has a 6 dB buildup at 300 Hz, the compressor will react disproportionately to that frequency. Cut the problem first, then compress a balanced signal.
- Stereo width before compression. Widening the stereo image before compressing can cause the compressor to pump unevenly as the side content fluctuates. Compress first for consistent dynamics, then shape the width.
- Saturation after limiting. Saturation generates harmonics that add energy to the signal. After a limiter, that added energy has nowhere to go -- it either clips or forces the limiter to re-engage. Put saturation before limiting.
- No DC filter at all. Many engineers skip the DC filter because they cannot hear the problem. But DC offset causes asymmetric waveforms that waste headroom and make compressors and limiters behave unpredictably. It takes nothing to include and fixes a real problem.
- Too many stages doing the same job. Three compressors each doing 4 dB of reduction is worse than one compressor doing 3 dB and a parallel compressor adding body. Stacking the same type of processing multiplies artifacts. Use different tools for different goals.
The mastering chain is not a mystery. It is a logical sequence where each stage prepares the signal for the next. Understand the purpose of each processor, respect the signal flow, and your masters will be cleaner, louder, and more professional.