A mastering chain can contain the same processors in two studios and produce radically different results. The difference is often not the plugin collection but mastering chain order: what each processor receives, how much it changes the signal, and whether the next stage is solving a problem or exaggerating it.
For serious producers, chain order is not a fixed recipe. It is a way of controlling cause and effect. A limiter asked to manage low-end excess, aggressive vocal peaks and harsh cymbals at once will sound strained. Address those issues earlier, with the right tools and modest moves, and the final limiting stage can concentrate on level rather than rescue work.
Why mastering chain order changes the result
Every processor alters the conditions for the one that follows. EQ before compression changes the frequency balance entering the detector circuit. Compression before saturation changes how consistently the saturator is driven. Clipping before limiting can reduce short peak events, allowing the limiter to work with fewer audible artefacts.
This matters because mastering processors are level-sensitive. A broad 1 dB high-frequency lift before a limiter may make a master feel brighter, but it can also cause cymbals, hats and sibilant material to trigger gain reduction more often. Put that same lift after a gentle first limiting stage and the tonal outcome may be cleaner, although the final ceiling still needs protection.
The practical question is not, “What is the correct chain?” It is, “What must be corrected before the next process can work transparently?” That distinction prevents the familiar habit of adding processors because a template says they belong there.
A dependable mastering chain order
A common starting order is corrective EQ, tonal or dynamic control, gentle enhancement, peak management, final limiting and metering. It is useful precisely because it follows the signal problems from broadest to most time-critical. Yet every stage should earn its position.
Start with level, monitoring and reference context
Before inserting processing, gain-match the source and audition it against relevant references. Do not compare your unprocessed premaster at a much lower loudness than a commercial release and assume the tonal balance is wrong. Perceived bass, brightness and punch all shift with level.
Leave sensible headroom on the premaster. There is no universal number required by every plugin, but a file that is already hard-clipped or limited to within a fraction of 0 dBFS gives the mastering chain very little room to operate. If the mix has obvious faults, return to the mix session where possible. Mastering can refine a balance; it is not a credible substitute for fixing a buried kick, a brittle vocal or a resonant bass note at source.
Use corrective EQ before broad dynamics
A transparent equaliser is usually the first active stage because unwanted subsonic energy, low-mid buildup and narrow resonances can cause downstream processors to react unnecessarily. A high-pass filter may remove information below the useful range of the track. A restrained cut around a persistent resonance can create more usable headroom than a large boost elsewhere.
This does not mean the first EQ should be a surgical exercise. Excessive notching can hollow out the programme and introduce phase shifts or unnatural movement. Work with wide, low-gain adjustments unless a specific problem is clearly audible and consistently present.
Dynamic EQ is especially valuable when a problem appears only intermittently. For example, a low-mid band that becomes congested only when the bass and lower vocal register overlap may be better controlled dynamically than cut throughout the track. Placing this stage before compression stops the compressor from overreacting to those moments.
Compress only when movement needs control
Mastering compression is often less about reducing dynamic range than shaping macro movement. A low ratio, moderate attack and slow or programme-dependent release can add cohesion, particularly on a mix where the drums, bass and vocal feel disconnected.
Order becomes critical here. If you want compression to respond less to the sub-bass, clean up the extreme low end before it. If you want the compressor to emphasise the weight of the kick and bass relationship, you may leave more low end intact or use a sidechain high-pass filter rather than EQing the programme itself.
A single broadband compressor is not automatically the best choice. Dynamic EQ or multiband compression can be more precise when only one area becomes unstable. The trade-off is complexity: split-band processing can alter the groove and tonal continuity if it is driven too hard. If the whole track does not need compression, do not add it simply because it appears in a mastering template.
Where saturation, stereo processing and colour belong
Colour processors usually sit after major corrective work and before the final peak-management stages. Saturation, tape emulation and transformer-style processing can add density, soften transients and generate harmonics that help a mix translate on smaller speakers. But they also raise average energy, particularly in the midrange, so they affect how hard the clipper and limiter will need to work.
Use saturation before clipping when you want the clipper to catch peaks created or reshaped by the colour stage. Conversely, placing a gentle clipper first can prevent very sharp drum transients from driving a saturator in an unpredictable way. Neither choice is universally superior. The first favours cohesive density; the second can favour cleaner, more controlled harmonic processing.
Stereo imaging should be approached with even more restraint. Low frequencies are normally kept central because wide sub information can create compatibility issues and unstable translation on club systems. If you widen upper harmonics or the sides, do it before the final limiter so the limiter responds to the actual stereo programme. Then check mono compatibility and monitor the side channel independently. Apparent width that disappears in mono is not an enhancement.
Clipper before limiter, but not by default
For modern electronic music, clipping before limiting is one of the most useful mastering chain decisions. A clipper can shave the tallest kick, snare and percussion peaks with less pumping than a limiter forced to catch every transient. This lets the limiter deliver competitive loudness with less continuous gain reduction.
The danger is treating clipping as invisible free level. Excessive clipping flattens drum impact, turns high-frequency transients into brittle distortion and can make a master fatiguing even when its integrated loudness looks impressive. Oversampling may reduce aliasing in some clippers, but it does not make an overdriven setting musical.
Try the clipper with small increments, then level-match its output against the bypass signal. Listen for changes to kick body, snare snap and vocal edge rather than looking only at peak reduction. If clipping makes the groove feel smaller, reduce it or move the task back into the mix.
Place the final limiter near the end
The final limiter should generally be one of the last stages because it establishes the delivery ceiling and catches remaining peaks. Keep its job narrow. If it is constantly pulling several decibels to control a harsh chorus or boomy drop, investigate the earlier chain rather than accepting the distortion as normal.
A second limiter can be effective when each stage performs very little work. One may catch brief peaks, while the final limiter establishes the final loudness and true-peak ceiling. This approach is not inherently more transparent, however. Two poorly configured limiters still create two layers of artefacts.
After the limiter, use a dedicated meter for integrated loudness, short-term loudness, true peak and stereo correlation. Metering is not processing, so its placement is straightforward: it must measure the actual delivered output. If you add a final gain trim, dithering stage or codec-preview tool after the limiter, place a meter after those stages as well.
Dither belongs at the final export stage
Dither is the last process when reducing bit depth, such as exporting a 24-bit master to 16-bit for a format that requires it. Do not place dither before EQ, compression or limiting, because subsequent processing alters the noise-shaped signal and defeats its purpose.
For 24-bit or 32-bit floating-point delivery, dither may not be necessary depending on the workflow and final format. The important point is procedural: apply it once, at the end, only when the bit-depth conversion calls for it. Repeated dithering is not extra protection.
Build chains around the material, not genre folklore
A sparse acoustic production may need corrective EQ, a touch of dynamic control and a limiter. A dense techno master may benefit from dynamic EQ, saturation, clipping and staged limiting. A bass-heavy track intended for club playback may require mono low-end management and more careful peak control than an indie mix aimed primarily at streaming.
The most revealing test is to bypass individual stages while matching level. If removing a processor makes the master no worse, it is probably not contributing enough to justify its side effects. SOUNDUNDERCONTROL’s practical standard is simple: each stage should make the following stage easier, not create a new problem for it to hide.
When a master starts feeling overworked, stop rearranging plugins for a moment. Identify the one event that is costing you loudness or clarity, solve that event at the earliest sensible point in the chain, then let the final limiter finish a job it can perform cleanly.