How to read a loudness report: LUFS, true peak, and LRA
The three numbers that describe how loud a record actually is, what they measure, why louder is not automatically better — and what they said about a real album before and after mastering.
Curriculum
The curriculum, from the work
This is the body of knowledge the work has actually produced — each subject named with its status and the evidence behind it, written up as the projects finish teaching it.
- PublishedReading a loudness report: LUFS, true peak, and LRAITU-R BS.1770 / EBU R 128 methodology with sources, read against the measured Wings & Prayers album.Read →
- Evidence readyBefore and after: what mastering measurably changedTwelve complete pre-master/reference-master measurement pairs are published in the case study — one ceiling, narrowed spread, dynamics kept.
- Evidence readyAlbum-level analysis: consistency, outliers, and the hardest trackPer-track data published for the whole record: a 0.8 LU core, remixes spanning 4.6 LU, and track 09 visible as the limiting problem in two numbers.
- PublishedLevel-matched A/B: building an honest comparisonWhy louder wins unmatched comparisons, and how the shipped instrument controls it — attenuation by each track’s measured delta, worked example included.Read →
- Evidence readyThe silence sentinel: when 0.0 LUFS means the meter is lyingThe bug and its fix — negative infinity, JSON null for non-finite values — are documented from the album’s first production run. Code-level detail stays private.
- In progressMeasurement versus judgment: what the meter cannot decideThe Decision Engine is the current milestone; the engineer-decides boundary is being drawn in software right now, and the writing should wait for it.
- Not yetThe wider report: PLR, crest factor, DC offset, stereo correlation and widthThe engine measures these; none are published yet. They appear when they can be reproduced from published artefacts (ADR 0007).
- Not yetInside true-peak oversamplingThe concept is covered in the published article; an implementation piece needs publishable engineering evidence it does not yet have.
Statuses are a statement about evidence, not a schedule. An entry moves when the work moves — and measurements appear only when they can be reproduced from published artefacts, the same rule the case study holds itself to.