The usual answer is -16 LUFS, and it is a good one. The reasons behind it are often garbled, though: Spotify and YouTube are said to require numbers they have never published, and the -19 LUFS figure for mono is credited to organisations that never recommended it. This article sorts out what each platform documents, what is only convention, and what to master to.
The short answer
Master stereo episodes to an integrated loudness of -16 LUFS with true peaks no higher than -1 dBTP. For a single-channel mono file, -19 LUFS is the common convention, for reasons explained below. These numbers follow Apple's published recommendation, and none of the other large platforms we checked publishes a different podcast target.
LUFS and LKFS are the same unit. Apple writes LKFS and the EBU writes LUFS; both mean loudness measured with ITU-R BS.1770. A difference of 1 LU is the same as 1 dB.
Apple Podcasts: -16 LKFS, ±1 dB
Apple publishes the clearest loudness figure for podcast episodes. Its audio requirements recommend that audio is "preconditioned so the overall loudness remains around -16 dB LKFS, with a +/- 1 dB tolerance, and that the true-peak value doesn't exceed -1 dB FS", with both values calculated according to ITU-R BS.1770-5. That preconditioning has to happen before encoding.
Apple also gives the reason for the peak limit: audio compression algorithms "typically don't modify the loudness and might clip the signal if the recommended true-peak value is not respected." In other words, an AAC or MP3 encoder can produce peaks higher than the file you give it. Apple adds that content with loudness or Sound Check metadata will consistently play back at -16 dB.
The wording is a recommendation ("we recommend"), not a rejection rule. It is still the most specific published figure for podcasts, which makes it the sensible reference for any show.
Spotify: no published podcast target
Spotify for Creators accepts MP3, M4A and WAV episode files, in mono or stereo, with no file-size limit. It publishes no loudness target for podcast uploads.
The -14 LUFS figure often quoted for Spotify comes from Spotify for Artists, and it describes music playback normalization. Spotify adjusts tracks to -14 LUFS in its default Normal mode, -11 LUFS in Loud and -19 LUFS in Quiet, at playback, without changing the uploaded file. Spotify does not say whether that normalization applies to podcast episodes, so "Spotify requires -14 LUFS for podcasts" is not something Spotify has published.
The one spoken-word loudness figure Spotify does document is for audio ad creatives: -16 LUFS ±1.5 LU. It applies to ads, not to episodes.
YouTube: no official number either
YouTube publishes no loudness target. Its help pages describe Stable volume, which is on by default and works by "continuously adjusting volume levels to reduce variations in sound", and automatic adjustments to volume levels and sound mixing that creators can opt out of for future uploads. Neither page gives a LUFS value.
The -14 LUFS figure linked to YouTube comes from third parties who read the "content loudness" value in Stats for nerds and measured playback. It is a useful observation for video work, not a YouTube specification. Loudness normalization explained covers how these systems behave.
Where -19 LUFS for mono comes from
A BS.1770 meter measures each channel and adds up their energy. A mono file has one channel, so it measures 3.01 LU lower than the same audio copied to both channels of a stereo file. Yet when a player sends a mono file to both earbuds or both speakers, it sounds as loud as that stereo copy. A mono file mastered to -16 LUFS can therefore play about 3 LU louder than stereo shows at -16 on players that do not compensate.
The documented source of "-16 LUFS stereo, -19 LUFS mono" is Google's loudness guidance for Assistant and Actions content, which recommended -19 LUFS for mono "rather than -16 LUFS" because mono and stereo "need to be offset by 3.01 LUFS". That page is now deprecated, but the convention stuck.
The figure is often credited to the AES, but it does not appear in the AES streaming guidelines. AES TD1008 notes only that mono played on two channels "will measure 3 LU higher" and recommends that players use channel-mode metadata. Apple publishes a single figure, -16 LKFS, with no separate mono number. So -19 LUFS for mono is a convention, not a platform rule. Mono vs stereo podcasts explains how to choose.
What the AES recommends
AES TD1008 (2021), the Audio Engineering Society's guidelines for internet audio streaming and on-demand distribution, recommends -18 LUFS for content where speech can be measured, such as podcasts and talk, measured as dialogue loudness with an upper tolerance of +1 LU. It advises keeping integrated loudness above -20 LUFS and true peak no higher than -1 dBTP at the input of a lossy encoder.
TD1008 has since been superseded by AES77-2023, whose content we have not verified, so treat -18 LUFS as a documented alternative rather than the current AES position. For most shows, -16 LUFS remains the practical choice because it matches Apple. A -16 LUFS episode played by an app with a lower reference level is simply turned down a little.
Leave true-peak headroom
Loudness and peaks are separate limits. A -16 LUFS master can still clip after encoding if its peaks are too high. Use a true-peak limiter, which oversamples the audio to catch peaks between samples, set to -1 dBTP or slightly lower, then measure the encoded file. AES TD1008 notes that high-rate encoders, around 256 kbps, may work with as little as -0.5 dBTP, so low-bit-rate episodes are where extra margin matters most. True peak vs sample peak explains why.
Podcast loudness at a glance
| Platform or source | Published loudness | Peak ceiling | Status |
|---|---|---|---|
| Apple Podcasts | -16 LKFS ±1 dB | True peak no higher than -1 dBFS | Official recommendation, measured per BS.1770-5 |
| Spotify podcasts | None published | None published | Accepts MP3, M4A and WAV, mono or stereo |
| Spotify music playback | -14 LUFS (Normal), -11 (Loud), -19 (Quiet) | Keeps 1 dB of headroom for lossy files when adding gain | Official, music only |
| YouTube | None published | None published | -14 LUFS is a third-party observation |
| AES TD1008 (2021) | -18 LUFS for speech (+1 LU), dialogue-measured | -1 dBTP at the encoder input | Superseded by AES77-2023 |
| Google Assistant (deprecated) | -16 LUFS stereo, -19 LUFS mono | Not stated | Origin of the mono convention |
The loudness targets reference lists more standards, and the free LUFS meter measures an episode's integrated loudness and true peak in your browser.
How to hit the target
- Clean up first: remove rumble, reduce steady noise and tame sibilance, because loudness processing makes all of them louder.
- Even out the levels of different speakers with a leveler or gentle compression, so the integrated figure reflects the whole episode.
- Apply gain to reach -16 LUFS integrated, or -19 LUFS for a mono file.
- Limit true peaks to -1 dBTP or a little lower with an oversampling limiter.
- Encode, then measure the encoded file. If it misses, correct the gain or ceiling and encode again.
LufsReady's podcast presets follow these steps: -16 LUFS with a -1 dBTP ceiling for stereo, -19 LUFS for mono, and a re-measurement of the encoded file, which is rendered again if it lands outside the target. How it works describes each stage.
Sources
- Apple Podcasts for Creators: Audio requirements
- Spotify for Artists: Loudness normalization
- Spotify for Creators: Publishing audio episodes
- Spotify Advertising: Audio ad specs
- YouTube Help: Control video volume on your device
- YouTube Help: Video and audio quality enhancements
- Google Assistant: Audio loudness guidelines (deprecated)
- AES TD1008: Loudness guidelines for internet audio streaming and on-demand distribution (archived copy)
- ITU-R BS.1770: Algorithms to measure audio programme loudness and true-peak audio level
Check your file in seconds
Measure loudness, true peak and noise floor for free in your browser, or master a file to the exact target with a report.