Normalize Podcast Loudness and Check True Peak

LUFS vs. peak: two different measurements

Peak level (dBFS) tells you the single highest sample in a file — a brief click or a loud consonant can spike it without the file actually sounding loud overall. LUFS (Loudness Units Full Scale) measures perceived loudness averaged over time, using the same underlying loudness model behind the EBU R128 broadcast standard. It is a far better match for how loud a listener actually experiences your episode, which is why platforms and podcast apps normalize around LUFS rather than peak.

Because of that mismatch, two files can share the same peak level and sound very differently loud, or share the same LUFS reading and have very different peaks. Normalizing loudness and controlling peak are two separate jobs, and the volume booster treats them as two separate controls: Loudness normalize for LUFS, and a true-peak ceiling for peak.

Integrated, short-term, and loudness range

The analysis panel reports three related but distinct loudness numbers. Integrated LUFS is the average loudness across the entire file — this is the number a target like "−16 LUFS" refers to. Short-term max is the loudest moment measured over a rolling few-second window, useful for spotting a section that is louder than the rest even if the overall average is on target. Loudness range (LRA), measured in LU, describes how much loudness varies between quiet and loud parts of the episode — a wide range means big swings between soft and loud moments, which can mean a listener has to keep reaching for the volume control.

Alongside those, the panel also shows sample peak and RMS. Sample peak is the highest individual digital sample in the file, a simpler and less useful measure of loudness than LUFS but still worth a glance. RMS gives a rough average signal level over time; it is a useful sanity check alongside LUFS, though LUFS remains the more accurate measure of how loud a listener actually perceives your episode to be, since it accounts for how human hearing weighs different frequencies.

Loudness targets

Different destinations expect different loudness levels. These are the common reference points built into the tool.

Mono vs. stereo podcasts

The volume booster offers separate podcast targets for stereo (−16 LUFS) and mono (−19 LUFS) files, and it is worth exporting at the format your show actually uses. A single spoken-word feed with no stereo separation is commonly distributed as mono, since it does not need two channels and keeps file size down; the lower LUFS target reflects common podcast delivery practice for mono spoken-word content. If your episode has stereo music beds or multiple mic positions panned across the field, export it as stereo and use the −16 target instead.

Why true peak, not just sample peak

Sample peak only looks at the exact digital samples in your file. True peak, measured in dBTP, estimates the actual waveform peak between samples using 4x oversampled detection, which catches "inter-sample peaks" that sample peak misses entirely. Those inter-sample peaks matter because your episode will likely be re-encoded to a lossy format somewhere in its distribution chain, and a peak that looked safe at the sample level can push over 0 dBFS and clip after that re-encoding. Setting a true-peak ceiling — commonly −1 dBTP — leaves enough headroom to survive that later encoding step cleanly.

If a target requires the limiter to work unusually hard to hit both the LUFS target and the true-peak ceiling at once, the tool shows a warning. That is worth listening to: heavy limiting can introduce audible pumping or flatten dynamics more than intended, so it is often better to check for a source-level problem, such as a section recorded far too hot, than to just push through the warning.