Loudness Metering
Peak meters stop clipping and loudness meters track perception; LUFS and true peak each watch one end.
Related
The problem
How loud a track feels is not the same as how high its peaks sit. A peak meter only says whether the signal passed full scale; a loudness meter says how loud the ear judges it. Mastering, podcasting and streaming delivery all set targets in loudness, so both rulers matter.
How it works
A peak meter reads the largest amplitude of a sample or between neighbouring samples in dBFS, and anything above zero clips. A loudness meter applies frequency weighting that mimics the ear's sensitivity to the midrange, then gates and integrates over time to produce LUFS. Short-term loudness uses a three-second window, integrated loudness covers the whole programme, and true peak interpolates between samples to catch overshoots a sample peak misses. Streaming platforms turn down tracks that are too loud, and the more they turn down, the more dynamics get squeezed.
Settings and practice
While mixing, watch peaks and leave 3 to 6 dB of headroom. While mastering, watch integrated loudness and set it to the delivery target instead of chasing maximum level. Keep true peak below minus one dBTP to leave room for lossy encoding. Use short-term loudness to judge balance between sections and integrated loudness for the whole track. Align loudness before comparing two versions, otherwise the louder one always seems better. Phones, televisions and cars play back at very different levels, so do not set the standard in one room only.
What to listen for
On the soul loop, hear how kick impact relates to overall weight; on the live ballade, whether quiet passages survive the applause; on the organ hymn, whether sustained passages stay steady. The numbers are a reference, and the ear still has to confirm that the rise and fall between sections is intact.
Listen in the library
All tracks from midicn-lib — inline badge is the license tier (C1 commercial / C2 non-commercial / C3 study only).
Sources
- ITU-R BS.1770 响度测量方法的用途，依公开标准说明核对，未抄原文
- 真峰值与采样峰值之差，依数字音频通识核对，未抄文本
Used for fact-checking only; all prose is original.