Free tool

Podcast Loudness Checker

Integrated LUFS and true peak for your episode, measured properly, without opening a DAW.

Looked after by Boom, a loudspeaker.

Choose an episode

or drop it here

MP3, WAV, M4A, FLAC.

Nothing is uploaded. The file is decoded by your browser and the samples are read in this tab, so there is no limit on episode length.

Podcast loudness targets

PlatformLoudnessTrue peakWhere it comes from
Spotify-14 LUFS-1 dBTPPublished by Spotify.
Apple Podcasts-16 LKFS, plus or minus 1-1 dBFSPublished by Apple, one number for every channel layout.
YouTubeabout -14 LUFS-1 dBTPNot published. This is what people measure, so treat it as an observation.
Podcast convention-16 stereo, -19 mono-1 dBTPConvention, not a rule. Apple no longer splits mono from stereo.

The two published numbers are 2 dB apart and you only get to pick one, so the practical answer is -16 LUFS: exactly right for Apple and 2 dB under Spotify, and 2 dB of turn-up is inaudible. Master at -14 and Apple turns you down while you have already given up the headroom.

How it works

  1. 1Drop in the episodeThe file you are about to publish. It stays on your device.
  2. 2Wait for the measureK-weighted, gated and oversampled, in this tab.
  3. 3Read the verdictLoudness, true peak and range, in plain words.
  4. 4Check each platformWhat Spotify, Apple and YouTube will do to it.

Questions

What LUFS should a podcast be?
Aim for -16 LUFS integrated with true peak at or below -1 dBTP, and you will be inside Apple's stated tolerance and close enough to Spotify's -14 that nobody will notice the 2 dB. If your show is mono, -19 LUFS measured is the equivalent, because the standard sums channels and a single channel measures 3 dB lower than the same programme in two.
Does it matter if my episode is louder than the target?
Not in the way people hope. Spotify turns a loud episode down rather than leaving it loud, so mastering to -9 LUFS gains you nothing at playback and costs you the dynamic range you crushed to get there. The one thing that does survive is the damage: a limiter squashing a voice to -9 makes it flat and tiring, and turning it back down afterwards does not undo that.
What is true peak, and why is it different from the peak in my DAW?
The peak meter in most editors reads the highest sample. The actual waveform passes between the samples, and it can go higher there than any sample does. That reconstructed level is true peak, and it is what a listener's decoder produces. It matters because MP3 and AAC encoding shifts peaks around: a file that reads exactly 0.0 dBFS on a sample meter can come out of Spotify's encoder clipping. Leaving 1 dB of headroom is the usual insurance, which is why both Apple and Spotify ask for -1.
My episode measures right but still sounds quiet. Why?
Look at the loudness range. Integrated loudness is one number for the whole episode, so a show with a loud host and a quiet guest can average out perfectly and still be unlistenable in a car. A range above about 10 LU on a talk show usually means somebody needs levelling, not more limiting on the master. Fix the gap between the voices first, then set the overall level.
Is my audio uploaded anywhere?
No. The file is decoded by your own browser and the samples are read in this tab. There is no upload, no account and no server, which is also why there is no limit on how long an episode can be.
How accurate is this compared to a plugin or ffmpeg?
It implements ITU-R BS.1770 properly, K-weighting and both gates, and it is checked against the EBU reference signals and cross-checked against ffmpeg's ebur128 filter on real speech, where the two agreed within 0.04 LU on loudness and 0.01 dB on true peak. The one compromise is that a long episode has to be decoded at a reduced sample rate to fit in browser memory. On speech that cost 0.04 LU at 24 kHz and 0.08 LU at 16 kHz in testing, which is nothing next to the 1 dB tolerance Apple allows. On a show with a lot of music it will be more, so the page tells you which rate it used.
Which file formats work?
Anything your browser can decode, which in practice means MP3, WAV, M4A, AAC, FLAC and OGG. Measure the file you are about to publish rather than the WAV you exported before encoding, because encoding is what moves true peak.
How this measures it

K-weighting: BS.1770 two-stage filter. A shelf and a high pass, so low rumble does not read as loudness the way a plain RMS meter would.

Gating: -70 LUFS, then -10 LU. Silence is dropped first, then anything 10 LU below the average, so pauses do not drag the number down.

True peak: 4x oversampled. The waveform between two samples can be higher than either. Sample peak is shown separately and labelled as such.

Checked against the EBU reference signals a meter is supposed to reproduce, and cross-checked against ffmpeg’s ebur128 filter on real speech, where the two agreed within 0.04 LU on loudness and 0.01 dB on true peak. A browser has to decode an episode in one piece, so a long file is decoded at a lower sample rate to fit in memory. On speech that cost 0.04 LU at 24 kHz in testing; on a show with a lot of music it costs more, so the page tells you which rate it used.

This tool is looked after by Boom.Loud, bassy, big-hearted. Tells you when it is too much.

More Free Tools

Image ConverterCrop Video to VerticalYouTube Thumbnail DownloaderReels Cover CreatorPodcast Media Kit GeneratorPodcast Guest Release Form
View All Free Tools

Shorts by Slice

From the team behind these free tools

One video. A week of posts.

Upload an episode and get back the moments worth posting, cut and captioned.

  • Reads the whole episode, then picks the excerpts worth posting
  • Captions burned in and framed vertical, ready to post
  • Seven days free, no card
Try Shorts by Slice