Tutorials Podcasting

What Is LUFS and How Loud Should My Podcast Be

By Infinite Creation · Beginner · ~20 min

A 1960s mid-century comic book style editorial illustration showing two audio engineers at a mixing desk, with a loudness meter visible in the frame.

Overview

LUFS is the loudness unit used by every major streaming platform to normalise audio so that all content plays back at a consistent perceived volume. Get this right and your podcast won't sound quiet next to other shows, or harshly loud when listeners use auto-volume on their phones.

What You Need

A finished podcast episode (WAV or MP3)
Audacity (free), or any DAW with an integrated loudness meter
Optional: a free LUFS meter plugin like Youlean Loudness Meter

Steps

Understand what LUFS measures

LUFS (Loudness Units Full Scale) measures the perceived average loudness of audio over time, not the peaks. Unlike dBFS, which only measures instantaneous signal level, LUFS accounts for how humans hear loudness across a whole programme. −16 LUFS sounds about the same subjective volume as a television news broadcast.

Learn platform loudness targets

Each platform has a normalisation target. Spotify: −14 LUFS for music tracks, with no published target for podcast episodes. Apple Podcasts: −16 LUFS. YouTube publishes no single official target, though −16 to −13 LUFS is the commonly cited range. If you submit at −14 LUFS, Apple Podcasts will turn you down. If you submit at −19 LUFS, Spotify will turn you up. Mastering to −16 LUFS is a sensible middle ground. It's the most commonly cited podcast standard.

Measure your podcast's loudness

In Audacity: select your entire track → Analyze → Contrast to get an approximate measure, or install the Youlean Loudness Meter plugin for a proper integrated LUFS reading. Load your exported episode and play it through the meter from beginning to end. Note the Integrated LUFS value.

Normalise to target in Audacity

Select your whole track → Effect → Loudness Normalization. Set the target to −16 LUFS. This applies a uniform gain adjustment to the whole episode so that the integrated loudness hits the target. This is not the same as clipping or compression. It's a single, clean gain pass.

Check your true peak ceiling

After normalising, check that your true peak doesn't exceed −1 dBTP. True peak is different from sample peak. It accounts for inter-sample peaks that can clip during codec encoding. In Audacity: Analyze → Find Clipping. If you have peaks above −1 dBTP, apply a −1 dB limiter before exporting.

Tested: bringing a voice to -16 LUFS under -1 dBTP

We ran steps 4 and 5 in FFmpeg 8.1.2 on 2026-09-24, reading loudness and true peak with its ebur128 meter. The voice was 12 seconds of speech from the "Daniel" voice built into macOS 26.6.2, set to -24 LUFS to stand in for a quiet recording. We brought it to -16 LUFS the ways a podcaster using FFmpeg might, and checked each result against the -1 dBTP ceiling in step 5.

What was done	Integrated loudness	True peak
The recording as made	-24.0 LUFS	-6.3 dBTP
Gain of +8.0 dB to reach -16 LUFS	-16.0 LUFS	+1.7 dBTP
Then alimiter limit=-1dB level=0	-16.1 LUFS	-0.9 dBTP
Then alimiter limit=-1dB, auto level left on	-15.1 LUFS	+0.1 dBTP
Then alimiter limit=-2dB level=0, and gain back to -16 LUFS	-16.0 LUFS	-1.5 dBTP
The row above encoded to MP3 at 128 kbps	-16.4 LUFS	-2.0 dBTP
loudnorm to -16 LUFS, one pass	-16.2 LUFS	-1.0 dBTP
loudnorm to -16 LUFS, two passes, linear=true	-16.0 LUFS	-1.0 dBTP
loudnorm to -14 LUFS, two passes, linear=true	-14.8 LUFS	-1.0 dBTP

This voice peaked 17.7 dB above its loudness, so the gain in step 4 alone took it to +1.7 dBTP, over full scale. At -16 LUFS a recording stays under -1 dBTP only if its peaks sit less than 15 dB above its loudness, and at -14 LUFS less than 13 dB. This voice needed the limiter in step 5 at either target.

FFmpeg's alimiter watches sample values, not true peak. At a limit of -1 dB it held every sample to -1.0 dBFS. The true peak was -0.9 dBTP, just over the ceiling. It also turns its own output back up by default. Left that way, the same limiter gave +0.1 dBTP and -15.1 LUFS. What worked was limiting 1 dB lower, with level=0, which cost 0.4 LU, then adding that back as gain: -16.0 LUFS at -1.5 dBTP. Encoding that file to MP3 lowered its peaks here, and it read 0.4 LU quieter.

FFmpeg's loudnorm filter chose its dynamic mode every time, which varies the gain through the file rather than applying the single gain step 4 describes. Run once, it also wrote the file at 192,000 Hz, so set the output rate. Run twice with linear=true, it still switched to dynamic mode. FFmpeg's manual says it does that when one gain would push the true peak over the target, as it would have here. At -16 LUFS it hit the target. At -14 LUFS it stopped at -14.8 LUFS.

To read a file's integrated loudness and true peak: ffmpeg -i episode.wav -af ebur128=peak=true -f null -

Validated. Measured 2026-09-24 with FFmpeg 8.1.2 on macOS 26.6.2, Apple silicon. One synthetic voice with no music, laughs or room noise. How far a real episode's peaks sit above its loudness varies, so measure your own. We did not test Audacity or any platform's own normalisation.

Pro Tips

Apply LUFS normalisation as the very last step, after all EQ, compression, and noise reduction is done.
Never boost audio that was too quiet at recording time with LUFS, all you'll do is raise the noise floor. Fix the gain before mixing.
Use our Podcast Level Meter to check your mic level in real time before recording.

Knowledge Base

What You'll Learn

Measure and normalize podcast loudness to LUFS standard. Understand why -16 LUFS matters for platform distribution. Use loudness meters to achieve consistent volume across all platforms.

Understanding LUFS

LUFS (Loudness Units relative to Full Scale) measures perceived loudness over time. Different from peak level (-6 dB peak can be -16 LUFS or -23 LUFS average depending on dynamic range).

Platform Standards

Apple Podcasts: -16 LUFS (integrated).
Spotify: -14 LUFS (integrated) for music tracks. Spotify publishes no target for podcast episodes, and its podcast ad specification is -16 LUFS (plus or minus 1.5).
YouTube: -16 to -13 LUFS. No single official target is published. This is the commonly cited range for video content.
Broadcast/EBU: -23 LUFS (European broadcast standard).

Metering Tools

Youlean Loudness Meter: Free, accurate, simple.
FabFilter Pro-Metering: Professional suite with loudness measurement.
Waves WLM Plus: Industry standard, cloud-based loudness database.

Workflow

1) Mix your episode to natural volume. 2) Load loudness meter on master output. 3) Check integrated loudness (full episode average). 4) If lower than target (-16 LUFS), gain up all tracks 2-4 dB. 5) Re-measure. 6) Repeat until -16 LUFS achieved. 7) Export.

Where This Fits

A loudness target is where a mix ends, and Mixing a Full Song covers the work before it.

FAQ

Q: If I mix to -16 LUFS, will it sound quiet on Spotify?
A: Spotify publishes -14 LUFS as its target for music tracks and does not publish a target for podcast episodes, so we cannot say how it will treat your file. Its podcast ad specification is -16 LUFS (plus or minus 1.5), and -16 LUFS is a common podcast standard. Check how the episode plays back on the platform.

Vocabulary

LUFS. Loudness Units Full Scale. The unit used to measure a programme's perceived average loudness over its full duration, rather than its instantaneous peak level. The measure every major streaming platform normalises against.

Integrated loudness. The single loudness figure LUFS produces for an entire recording, averaged across its whole duration rather than measured at any one moment. What a platform's normalisation target actually compares against.

True peak. The highest level an audio signal's waveform reaches in continuous time, including points that fall between the digital samples. Can exceed the sample peak a basic meter shows, which is why a separate true-peak check matters after normalising.

dBTP. Decibels True Peak. The unit true peak is measured in. A true peak above -1 dBTP risks clipping or distortion when the file is later converted or encoded, even if it looked clean in the original file.

Gain. An adjustment to a signal's overall level, applied uniformly across a recording. Loudness normalisation works by applying a single gain change calculated to bring the integrated loudness to a target, distinct from compression or limiting, which alter dynamic range rather than only level.

Sources

Read 22 September 2026. Apple's figure and EBU R128 are also cited, with the same numbers, on this site's Standards page.

Spotify, Loudness normalization on Spotify. Spotify's own wording: we adjust tracks to -14 dB LUFS, according to the ITU 1770 standard, and recommends mastering to -14 dB integrated LUFS with true peak kept below -1 dBTP. The article covers music tracks and does not mention podcasts.
Spotify Ads, Podcast ads minimum requirements. Read 6 October 2026. Spotify's wording: overall loudness -16 LUFS, an integrated average of -16 LUFS (+/- 1.5 LUFS) and a true peak limit of -2.0 dBTP. It applies to podcast ads, not to episodes.