Podcast Loudness Explained: -16 LUFS vs -14 LUFS
Every major platform turns your podcast up or down to match everything else in the listener’s queue. You cannot opt out of it. What you can do is master to the right number so the adjustment is small and nothing is damaged on the way.
Here is what the numbers mean, where each platform sits, and why −16 LUFS remains the safe answer for most shows.
What LUFS actually measures
LUFS stands for Loudness Units relative to Full Scale. Unlike a peak meter, which reports the single loudest instant, LUFS measures perceived loudness over time, weighted to match how human hearing responds to different frequencies.
That distinction matters. Two files can peak at exactly the same level and still sound twice as loud as each other, because one of them is loud consistently and the other has a few loud spikes in a quiet programme.
LUFS is a negative number because it is measured down from digital maximum. −16 LUFS is louder than −23 LUFS. One LU is roughly one decibel.
You will also see LKFS, which Apple uses. LUFS and LKFS are the same measurement under two names.
Where each platform sits
| Platform | Target | Notes |
|---|---|---|
| Apple Podcasts | −16 LKFS ±1 dB | True peak no higher than −1 dBFS |
| Spotify | −14 LUFS | Normalises on playback |
| YouTube | −14 LUFS | Normalises on playback |
| Amazon Music | −14 LUFS | Normalises on playback |
| Most podcast convention | −16 LUFS | Mono or stereo |
| Broadcast (US, ATSC A/85) | −24 LKFS | For comparison only |
Apple publishes −16 LKFS as a specification. The streaming music platforms sit at −14 because they were built for music, which tolerates a louder average.
Why −16 is the safe choice
The platforms differ, so you have to pick. The reason to pick the quieter of the two targets comes down to what happens in each direction.
If you master too quiet and the platform turns you up, it raises everything — including your room noise, your air conditioning, and the hiss in the quiet passages. Listeners hear a noticeably noisier show.
If you master too loud and the platform turns you down, nothing is damaged. The file is simply played quieter. You lose nothing except the loudness you were chasing.
The asymmetry is the whole argument. Being turned down is harmless; being turned up is not. Master at −16 LUFS and Spotify plays you 2 dB quieter than its target, which no listener will ever notice.
The exception is a show that lives almost entirely on one platform. If 90 per cent of your listening is on YouTube, master to −14.
True peak, and why −1 dBTP
True peak measures the peaks that appear when a digital file is converted back to analogue — which can be slightly higher than anything visible in the file itself. Lossy encoding to MP3 or AAC adds a little more.
Apple specifies a true peak ceiling of −1 dBFS. That headroom absorbs both effects. A file that peaks at 0.0 dB in your editor can clip audibly after encoding.
Set a true peak limiter at −1 dBTP as the last step in your chain and the problem disappears.
Mono or stereo?
For a talk show, mono. A single voice has no stereo information worth preserving, and mono halves the file size at the same bitrate — or doubles the effective quality at the same size.
One thing to watch: the loudness measurement of a mono file summed to stereo can read differently depending on your meter. Measure the file as you will deliver it.
How to actually hit the target
- Edit and mix first. Balance the speakers against each other before you think about overall loudness. Mastering cannot fix a mix where one guest is 8 dB quieter than the host.
- Gentle compression on each voice. Enough to even out delivery, not enough to hear. A 3:1 ratio with a few decibels of gain reduction is plenty for speech.
- Measure the whole episode with a loudness meter in integrated mode. Not a section — the whole thing, because that is what the platform measures.
- Apply gain to reach the target. If you measure −19.5 LUFS and want −16, add 3.5 dB.
- True peak limit at −1 dBTP.
- Re-measure. Limiting changes the integrated figure slightly.
Free tools that measure integrated LUFS include Youlean Loudness Meter and the loudness normalisation built into Audacity and Reaper. You do not need to buy anything.
The mistake worth avoiding
Chasing loudness with heavy compression and a hard limiter is a habit carried over from music mastering, and it does not transfer. A heavily compressed podcast measures at target and sounds exhausting — the dynamics that make speech feel natural are gone, and listeners fatigue faster without knowing why.
Because every platform normalises anyway, the loudness you win that way is taken straight back. All you keep is the damage.
Hit the target, protect the dynamics, and let the platform do its job.
If you would rather hand this over, our podcast editing service masters every episode to the loudness target for the platform your show actually lives on, with true peak held at −1 dBTP.
