Mastering & delivery
LUFS vs peak level: measure loudness without mixing to a number
Use integrated and short-term loudness alongside peak meters to compare masters and avoid loudness-driven mix mistakes.
LUFS describes programme loudness using a defined measurement method; peak level describes the largest signal excursions. Two songs can share a peak ceiling while one is much denser and louder. Matching peaks alone is therefore a poor way to compare their tonal balance.
By Spectraforge Audio2 min read
Published
Guide review: Captum Records and AS Mastering. AI-assisted writing and source checks. How these guides are made.
Use loudness for comparison and peaks for headroom
Use loudness readings to establish a fair starting level or meet a supplied specification. Use sample and true-peak readings for peak headroom. A loudness number is not a score for impact, clarity or dynamics: those decisions still require listening.
Choose a window that matches the question
EBU-mode momentary loudness covers 400 ms, short-term covers three seconds, and integrated loudness covers the measured programme using gating. Reset the integrated measurement before analysing a whole song. A chorus loop and a full-song measurement cannot stand in for each other.
For a release report, include the intended opening and ending and identify the meter mode. For a local balance question, inspect short-term loudness through the relevant phrase. If you have measured only a section, name that section. Otherwise a saved meter history can look like a precise claim about a file you never actually measured.
Worked comparison: −9 LUFS and −12 LUFS masters
Suppose two complete masters measure −9 and −12 LUFS integrated. Reduce the louder one’s playback by about 3 dB as an initial comparison, without re-limiting either file. Route both around your mix’s processing. Compare similar passages, then refine the listening level if their verse/chorus structures differ.
Now judge the kick’s shape, vocal position and cymbal comfort. If the apparent improvement disappears when levels match, more limiting has not supplied that improvement. If one isolated hit triggers large reductions while the arrangement remains comparatively sparse, revisit that event before forcing the whole master toward a number.
Separate artistic loudness from destination requirements
A delivery specification is an instruction for a submitted file; playback normalisation is what a service may do later. Check the current destination documentation rather than mastering every song to a remembered platform figure. Playback gain cannot restore attacks removed from your source master.
When a target is required, measure the complete export and check true peak separately. Save the filename, meter mode, integrated LUFS and maximum dBTP. Then listen to a quiet verse, the chorus transition and the ending without watching the display. This confirms whether the measured result still carries the musical shape you wanted.
Technical references
- EBU Tech 3341: momentary, short-term and integrated loudness
- ITU-R BS.1770-5: loudness and true-peak measurement
Listening exercises are proposed practice routines. Worked examples illustrate the stated setup; they are not reports of our listening tests. Editorial process and corrections.