Liner Notes

How mastering actually works

The decisions behind records you already know, and the engineering underneath them. Written for people who want to understand what is happening to their audio — not to be sold a preset.

technicalloudness6 min

Why boosting the EQ doesn't make your master louder

You add 3 dB at 80 Hz, the meter doesn't move, and it looks like the tool ignored you. It didn't — the limiter gave the gain back on purpose.

Read →
technicalchain7 min

The eleven stages, and why they are in that order

A mastering chain is not a pile of effects. It is a sequence, and most of what makes one master sound better than another is the order rather than the settings.

Read →
technicalchain8 min

Two compressors in a row, and why that isn't a mistake

The chain compresses twice, with settings that look almost contradictory. The contradiction is the point — one is holding the record together, the other is giving it punch.

Read →
technicalchain8 min

Three ways to make a peak smaller

Saturation, limiting and clipping all do the same arithmetic and sound completely different doing it. Which one does the work is the difference between a master that breathes and one that doesn't.

Read →
technicalchain8 min

The stages that decide what it sounds like

Two EQs, a de-esser, a multiband compressor and a width control. These are the stages with no correct setting — and the two of them that can be copied from a reference record.

Read →
technicalanalysis8 min

What we mean by bright

Warm, bright, airy, lean, scooped. Everyone uses these words and almost nobody agrees what they mean — so here is exactly what each one is worth in decibels, and which record sits at each extreme.

Read →
technicalanalysis8 min

Punchy, dynamic, compressed — three words, two measurements

The vocabulary for dynamics splits along a fault line most people never notice, and a quarter of the library sits on the wrong side of it. Public Enemy hits harder than Dire Straits and has a third of the range.

Read →
technicalanalysis7 min

Reading a spectrum, and what it hides

Ten numbers describe a record's tonal balance well enough to act on. They also leave out most of what makes it recognisable.

Read →
technicalreferences8 min

What a finished master can tell you, and what it can't

You can measure any record ever released. That still won't tell you what was done to it — and mistaking one for the other is the most common error in reference matching.

Read →
historyloudness9 min

The loudness war, and the record that ended the argument

For thirty years records got louder and lost their dynamics. Then streaming turned the volume back down and the whole strategy stopped paying.

Read →
historyreferences8 min

What makes a mastering iconic

Four different things get called "legendary mastering", and one of them is a warning rather than an aspiration. Telling them apart is most of the skill in choosing a reference.

Read →
historyproduction8 min

Kind of Blue, 1959 — a room, three tracks, and almost no second takes

The best-selling jazz record ever made was cut in two afternoons, mostly first takes, in a converted church. It is also slightly out of tune, and nobody noticed for thirty years.

Read →
historyproduction8 min

Abbey Road, 1969 — the desk that changed the sound

The Beatles' last recorded album sounds smoother than everything before it, and the main reason is not the songs or the band. It is a new mixing console with a compressor on every channel.

Read →
historyproduction7 min

Thriller, 1982 — when the format decided the tracklist

The best-selling album in history was shortened because vinyl physically could not hold it and stay loud. The constraint made it better.

Read →
historyproduction7 min

Nevermind, 1991 — three people, three jobs, one argument

Butch Vig produced it, Andy Wallace mixed it, Howie Weinberg mastered it, and the band ended up unhappy with how good it sounded. It is the clearest lesson available in who does what.

Read →