The Mathematics of Mastering Peak Limiters and Streaming Targets
The final peak limiter in an audio mastering signal chain serves two crucial purposes: catching rapid inter-sample transients to prevent digital clipping, and increasing perceived acoustic loudness (LUFS) to match commercial industry standards. However, dialing in incorrect limiter attack and release times creates catastrophic side effects—either destroying transient punch and causing unnatural "breathing" pumping, or generating harsh harmonic buzzing across 808s and sub-bass synthesizers.
Why Limiter Release Must Match Song BPM
When a kick drum or snare hits the limiter threshold, the gain reduction circuit immediately attenuates the audio level. How fast the limiter releases (recovers back to unity gain) determines the groove and transparency:
- Too Fast: If the release time is shorter than one-half of a bass waveform's period (for a 50Hz sub, a single wave takes 20 ms), the limiter opens and closes within a single cycle of the sine wave. This wave-shaping distortion sounds like gritty, buzzy square-wave distortion.
- Too Slow: If the release time is overly sluggish (e.g. 300 ms on a 130 BPM dance track), the limiter remains pulled down during the subsequent downbeat, sucking the life and dynamic energy out of following instruments (limiter pumping).
- Tempo-Synced Sweet Spot: Setting the release to an exact musical note division (such as 1/32 or 1/16 note based on
Release ms = (60,000 / BPM) × Note Division) ensures the limiter snaps back to 0 dB gain reduction just in time for the next groove element.
True Peak (ISP) and Streaming Normalization
Standard peak sample meters only measure the value of discrete digital audio samples. When audio is converted back to analog through a DAC (Digital-to-Analog Converter) or compressed into lossy codecs (Ogg Vorbis, AAC, MP3) for Spotify, YouTube, or Apple Music, the reconstructed continuous waveform can rise 0.5 to 1.5 dB higher than the highest digital sample. These are known as Inter-Sample Peaks (ISPs).