What Is LUFS? Podcast Loudness Explained in Plain Words

You finished an episode. On your headphones it sounds fine. A listener writes in: "This one is much quieter than the last." Nothing about your setup changed, but the loudness did. LUFS is the number that would have told you before you published.

This guide explains what LUFS means, why every platform uses it, and how to hit the right number for your show. When you are ready, the loudness normalizer on the home page measures and fixes a file in one pass, and the companion guide on normalizing a podcast to -16 LUFS walks the steps.

What LUFS Measures

LUFS stands for Loudness Units relative to Full Scale. It is a measure of how loud a whole recording feels to a human ear, averaged over time. It comes from the ITU-R BS.1770 standard, the method broadcasters and streaming platforms all share. The letters look technical, but the idea is simple: LUFS answers the question "how loud does this sound?" instead of "how big is the signal?"

Two details inside the standard do most of the work. First, the meter weighs frequencies. Human ears are most sensitive to the middle range — the range where speech lives — and less sensitive to deep bass and very high tones. The BS.1770 method boosts the middle and trims the extremes before it adds everything up, a step called K-weighting. That is why a bass-heavy track and a speech track with the same raw signal level can measure several LU apart: the meter hears them the way you do.

Second, the meter gates. It ignores long stretches of silence and near-silence, such as the pause before the intro music or a gap between segments. Without gating, a show with long quiet gaps would measure quieter than it feels, and a producer could cheat the number by padding a file with silence. With gating, the number tracks the parts listeners actually hear.

Why LUFS Is Not a Peak Meter

A peak meter looks at the single loudest instant. Two files can both peak at the same point and still feel very different in loudness — one mostly quiet with a single shout, one loud all the way through. LUFS hears the difference because it listens to everything, weighs the frequencies the ear is most sensitive to, and ignores long silences at the start and end.

Here is a worked example. Imagine two 30-minute episodes, both exported so their highest peak touches -1 dB:

A peak meter calls all three files identical. A listener calls them three different volumes. LUFS is the measurement that agrees with the listener, which is why it — and not peak — became the number every platform publishes.

Where LUFS Came From

LUFS is younger than most people assume. Television created it. Through the 1990s and 2000s, viewers complained that adverts blasted louder than the programmes around them, even though every broadcaster swore the peaks were legal. They were: advertisers had learned to compress audio so it felt louder without ever peaking higher. Peak rules could not stop the trick, because the trick never broke them.

Regulators answered with a measurement of felt loudness instead. The ITU published the BS.1770 method in 2006, Europe built EBU R128 on it in 2010, and the US followed with ATSC A/85. Streaming services adopted the same method a few years later, which is why one standard now covers a podcast recorded in a bedroom and a drama broadcast to a continent. When you measure in LUFS, you are using the tool that ended the loudness war.

Why Every Platform Uses It

Platforms want every show at a similar loudness so listeners are not reaching for the volume knob between episodes. Apple Podcasts asks for about -16 LUFS. Spotify and YouTube play everything back at about -14 LUFS. European broadcasters use -23 LUFS under the EBU R128 rule. If your file is far off the target, the platform turns it up or down on its own terms — and a quiet file turned up by a simple gain also turns up its noise.

The targets differ because the listening situations differ. Podcasts are heard on earbuds in noisy streets, so the target sits fairly loud. Broadcast television is heard in quiet living rooms on bigger systems, so the target sits lower and leaves more room for dramatic swings. The measurement is the same everywhere; only the goalposts move.

The Numbers That Matter

A full loudness report has more than one number in it. These are the ones worth learning:

• Integrated loudness (LUFS). The headline number: the average loudness of the whole file, gated so silence does not drag it down. This is the number platforms check.
• True peak (dBTP). The highest point the signal reaches between samples — the number that decides whether a file clips on playback; we explain it in the true peak guide.
• Loudness range, or LRA (LU). How much the loudness moves between quiet and loud parts inside the file; the LRA guide shows why a healthy range makes a show easier to listen to.
• Momentary and short-term loudness. The meter's live readings over 0.4 seconds and 3 seconds. Useful while mixing — you can watch a quiet guest drift down in real time — but no platform grades a file by them.

A real report for a healthy episode might read: integrated -16.1 LUFS, true peak -1.2 dBTP, LRA 6 LU. That file will sit correctly in a podcast queue, will not clip on any phone, and will not make the listener ride the volume knob. Learn to read those three lines and you can judge any episode in five seconds.

LUFS vs the Numbers You Already Know

If you came up on older tools, you may know dBFS, RMS, or the "normalize to 0 dB" button. They are not interchangeable:

RMS comes closest, and on steady material RMS and LUFS often land within a couple of units of each other. But RMS does not weight frequencies or gate silence, so on real speech with pauses and music beds the two drift apart — sometimes by 3–4 LU. When a platform asks for LUFS, an RMS reading is a guess, not a measurement.

How to Use LUFS in Your Workflow

Measure every finished episode before you upload it. If the number is off, normalize it to the target instead of guessing with a volume slider. A two-pass normalizer measures first, applies one exact gain, holds the true peak under the ceiling, and then measures again so the report proves the result. That is exactly what LoudNorm does, and batch shows can process a whole season at once with the workflow in the batch normalization guide.

To see the routine in action, take three real reports from one show's archive. Episode 41 measured -21.3 LUFS with a true peak of -3.1 dBTP: a clean but quiet recording, fixed with a straight +5.3 dB of gain to land at -16.0. Episode 42 measured -16.2 LUFS and needed nothing. Episode 43 measured -13.8 LUFS with peaks already at -0.4 dBTP: a hot master, turned down 2.2 dB, which also pulled its true peak safely under the ceiling. Three different problems, one routine, and a queue where no episode surprises the listener. Without the measurement step, episodes 41 and 43 would both have shipped — one whispering, one shouting.

A simple routine that works for any show:

1. Finish everything else first. Editing, EQ, compression, music — loudness is measured on the final file, so it comes last.
2. Measure the export, not the project. Meters inside an editor can be set to different standards. Measure the actual MP3 or WAV you will upload.
3. Compare against your platform target. Stereo podcast: -16. Mono: -19. Streaming-first: -14. Broadcast in Europe: -23.
4. Normalize if you are more than 1 LU off. Inside ±1 LU, no listener can tell. Outside it, fix it — fixing a quiet recording needs more than gain when the file is far under.
5. Keep the report. A saved report per episode turns listener complaints into a two-minute check instead of an argument.

Common Mistakes

Normalizing to a peak instead of a loudness. "Normalize to -1 dB" in many editors means peak normalization. As the worked example above shows, three files can share a peak and differ by 13 LU. Peak normalization is not loudness normalization, and the names are confusingly close.

Measuring before the final edit. Every cut, fade and music bed changes the integrated number. A measurement taken before the last edit describes a file that no longer exists.

Forgetting that mono plays louder. A mono file is doubled into both ears at playback and sounds about 3 LU hotter than it measures. That is why mono targets -19 instead of -16 — the full story is in the mono vs stereo loudness guide.

Trusting one upload to fix itself. Some platforms turn quiet files up, some do not — YouTube famously leaves quiet uploads quiet. Publishing far off target and hoping the platform sorts it out gives every platform a different problem to solve differently.

Chasing decimals. -16.0 vs -15.8 is not a difference a listener can hear, and most platforms accept ±1 LU. Spend your effort on being consistently close, not on being exact to a tenth.

Frequently Asked Questions

Is LUFS the same as LU? Almost. LUFS is loudness measured against full scale — an absolute number. LU is the unit for differences and relative readings, such as loudness range. A file is -16 LUFS; it is 2 LU louder than another file.

Why are LUFS numbers negative? Because they are measured down from the loudest possible digital signal, 0 dBFS. A quieter file sits further below zero. Nothing in digital audio can measure above 0 LUFS integrated — that would mean the average is louder than the maximum.

Can I measure LUFS without special software? Any BS.1770 meter works: most modern editors have one built in, and an online tool like the one on this site measures a file in the browser with nothing to install. What matters is the standard (BS.1770), not the brand of meter.

Do I need a different file for every platform? Usually not. The common podcast targets are within 2 LU of each other, and one good master at -16 LUFS serves Apple and Spotify acceptably. Broadcast delivery is the exception — it needs its own master at -23 LUFS.

What if my episode measures exactly on target but still sounds quiet next to other shows? Check the loudness range. A wide-range episode at -16 LUFS has quiet passages well under -16, and those passages are what listeners compare. The range problem is fixed in the mix, not with more gain.

References

• https://www.apple.com/apple-podcasts/ — Apple Podcasts, whose publisher guidance sets the -16 LUFS reference for stereo episodes.
• https://artists.spotify.com/ — Spotify for Artists, which documents the -14 LUFS playback normalization for music and podcasts.
• https://tech.ebu.ch/publications/r128 — EBU R128, the European broadcast loudness recommendation built on ITU-R BS.1770.