The Mastering Chain

A mastering chain is a short line of processors on your stereo mix. Each one makes a small change. Added together, those small changes turn a good mix into a finished, release-ready master.

The Typical Chain Order

Here is a common order. Each step is explained below.

Corrective EQ→Compression→Tonal EQ→Saturation→Stereo imaging→Limiter→Dither

A typical mastering chain. The limiter and dither always come last.

Why this order?

Fix before you enhance. Remove problems first, so the compressor doesn't react to them.
The limiter goes near the end, because it sets the final peak level. Anything after it could push peaks over the ceiling.
Dither is the very last step, and only when you reduce bit depth.

You won't always need every step. If the mix sounds great, a master might be only a limiter and a tiny EQ move. Never add a processor just because it's on the list.

Note

The golden rule of mastering: small moves. Most EQ changes are 0.5–2 dB, compression is 1–2 dB of gain reduction, and you bypass every plugin often to check it actually helps.

Step 1: Corrective EQ

Corrective EQ removes problems in the mix as a whole.

Sub rumble: a gentle high-pass filter around 20–30 Hz removes energy you can't hear but that eats headroom. Use a gentle slope (12 dB/octave) and check it doesn't thin the kick.
Resonances: a narrow cut of 1–3 dB on a boomy note in the bass (often 60–200 Hz) or a harsh spot in 2–5 kHz.
Mud: a wide, small cut around 200–400 Hz if the whole mix sounds clouded.

Many engineers use a linear-phase EQ for mastering. It avoids the phase shift of normal EQs, at the cost of extra latency and a small "pre-ringing" artifact on steep settings. For gentle moves, either type works.

Step 2: Compression

Mastering compression is gentle. It holds the mix together and controls dynamics slightly. It isn't there to make things loud.

Mastering compressor settings

Ratio 1.5:1 to 2:1 Threshold set for 1–2 dB of gain reduction on the loudest parts Attack 10–30 ms (slow enough to keep transients) Release 100–300 ms, or auto; time it to breathe with the groove Knee soft Make-up gain match the bypassed loudness

Listen for the mix feeling more "together" and steady. If the kick loses punch, lengthen the attack. If you hear pumping, shorten or lengthen the release, or reduce the gain reduction.

Multiband compression

A multiband compressor splits the audio into frequency bands (often 3 or 4) and compresses each one separately. It's useful when only one area has a problem:

The low end jumps up on some bass notes: compress only the band below 120 Hz.
The cymbals or vocal get harsh when loud: compress only around 3–8 kHz.

Multiband is powerful but easy to overuse. It can change the mix balance in ways you don't notice at first. Use it only when you can name the problem it fixes.

Step 3: Tonal EQ & Mid/Side

Tonal EQ shapes the character of the whole song with broad, gentle curves.

Low shelf +/-0.5–1.5 dB around 80–100 Hz for more or less weight.
High shelf +0.5–2 dB from about 8–12 kHz for "air" and openness.
A wide, small dip around 2–4 kHz if the mix tires your ears.

Mid/side EQ

Mid/side (M/S) processing splits a stereo mix into two parts:

Mid: what is the same in both speakers (usually kick, bass, snare and lead vocal).
Side: what is different between the speakers (wide pads, stereo reverb, panned guitars).

Common M/S moves:

High-pass the side channel below about 100–150 Hz. This keeps the bass mono and solid, which helps clubs and vinyl.
A small high-shelf boost on the side to make the top end feel wider.
A small presence boost on the mid to bring the vocal forward.

Step 4: Saturation & Stereo Imaging

Saturation

Gentle saturation from a tape, tube or console emulation adds harmonics. It can make a mix sound warmer and fuller, and slightly rounds off peaks so the limiter works less. Keep the mix (dry/wet) control low, often 10–30%, and level-match when you compare.

Stereo imaging

A stereo imager makes the mix wider or narrower. In mastering, use it carefully:

Widen only the mids and highs, if anything. Keep everything below about 120 Hz mono.
Always check in mono after widening. If the vocal or snare drops in level, you've gone too far.
Watch a correlation meter: it should stay mostly between 0 and +1. Values near -1 mean phase problems.

See Panning & Stereo for more about width and mono compatibility.

Step 5: The Limiter

A limiter is a compressor with a very high ratio and a very fast attack. It stops peaks from going above a set level, called the ceiling. Turning up the input (or lowering the threshold) pushes the track into the limiter and raises the overall loudness.

Limiter settings

Ceiling -1.0 dBTP (true peak; -2 dBTP for very loud masters) True peak ON Gain / input raise until the target loudness, typically 1–4 dB of reduction Release auto, or 50–200 ms Lookahead default (often 1–5 ms)

True peak (dBTP) measures the peaks that appear between samples when audio is converted back to analog or encoded to MP3/AAC. A ceiling of -1 dBTP leaves room for those peaks, so lossy streaming versions don't distort.

Listen for limiter damage

Signs of over-limiting: the kick and snare lose their snap, the mix pumps or "breathes", distortion appears on loud bass notes, and the chorus no longer feels bigger than the verse. When you hear any of these, back off the gain. Learn about loudness targets in Loudness & LUFS.

Step 6: Dither

When you reduce the bit depth (for example from 24-bit to 16-bit for a CD), the extra detail is cut off. That causes small, harsh-sounding errors called quantization distortion, mostly heard in quiet fades and reverb tails.

Dither adds a tiny amount of special noise before the reduction. This turns the harsh distortion into a smooth, very quiet hiss that is far less noticeable. Noise shaping moves that hiss to frequencies where our ears are less sensitive.

Dither only when reducing bit depth, such as 24-bit → 16-bit.
Dither once, as the very last step. Dithering twice adds extra noise.
If you export a 24-bit file for streaming, dither is usually optional; many engineers still apply it. A 32-bit float export needs none.

A common limiter ceiling for streaming masters is dBTP.

Dither is added when you reduce the , and it is always the step.

Mastering Plugins

You can master well with stock plugins. Well-known tools include:

Tool	Type	Notes
Stock DAW tools	Free with your DAW	Ableton: EQ Eight, Glue Compressor, Multiband Dynamics, Limiter. FL Studio: Fruity Parametric EQ 2, Maximus, Fruity Limiter. Logic: Channel EQ, Linear Phase EQ, Multipressor, Adaptive Limiter.
iZotope Ozone	All-in-one mastering suite	EQ, dynamics, imager, maximizer and an assistant that suggests a starting chain.
FabFilter Pro-L 2	Limiter	Widely used, with true-peak limiting and detailed loudness metering.
FabFilter Pro-Q / Pro-MB	EQ / multiband	Precise EQ with linear-phase and mid/side modes; flexible multiband dynamics.
Loudness meter	Metering	Many DAWs include one, and free meters exist. You need LUFS and true-peak readings.

Try it in your DAW

Import your mix (peaks around -6 dBFS) and a level-matched reference into a new session.
On the master [Ableton: Main; FL Studio: Master; Logic: Stereo Out], add an EQ: high-pass at 25 Hz, plus one small cut or boost (under 2 dB) based on the reference.
Add a compressor: 2:1, 30 ms attack, auto release, 1–2 dB of gain reduction.
Add a limiter with the ceiling at -1 dB (true peak on if available). Raise the gain until you see 2–3 dB of reduction on the loudest parts.
Add a loudness meter after the limiter and note the integrated LUFS.
Bypass the whole chain at matched level. Keep only the changes that clearly help.
Export a 24-bit WAV for streaming, and a 16-bit WAV with dither for CD.

1. Which processor comes last in a mastering chain?

Corrective EQ
Compressor
Stereo imager
Dither (after the limiter)

The limiter sets the final peak level, and dither is applied once at the very end when reducing bit depth.

2. How much gain reduction is typical for mastering compression?

About 1–2 dB
About 10–15 dB
At least 20 dB
None, compressors are never used

Mastering compression is gentle glue, not heavy squashing.

3. Why do mastering engineers often high-pass the side channel below about 100–150 Hz?

To make the vocal louder
To add reverb
To keep the low end mono and solid
To remove sibilance

Wide, out-of-phase bass is weak in mono and problematic on clubs and vinyl. Mono bass is focused and solid.

4. Why set a limiter ceiling to -1 dBTP instead of 0 dBFS?

It makes the track louder
It leaves room for inter-sample peaks and lossy encoding, so the file doesn't distort
It is required for vinyl only
It removes the need for dither

Peaks between samples and MP3/AAC encoding can push a 0 dBFS file into clipping. A true-peak ceiling of -1 dBTP prevents this.

5. When is multiband compression the right tool?

On every master, always
To make the mix wider
To change the tempo
When one frequency area has a dynamic problem, such as bass notes jumping out

Multiband compression treats only the band with the problem. Use it when you can name what it fixes.