The short answer
Streaming services measure the integrated loudness of your song in LUFS and play it at their own reference level. A master louder than that reference is turned down; it does not sound louder to the listener, only denser. Keep true peaks at or below −1 dBTP so the lossy versions the services create do not distort.
| Platform | Plays at about | Louder masters | Quieter masters | Peak ceiling |
|---|---|---|---|---|
| Spotify (Normal) | −14 LUFS | turned down | turned up, only as far as their peaks allow | −1 dBTP (−2 dBTP if louder than −14 LUFS) |
| Spotify (Loud setting) | −11 LUFS | turned down | turned up with a limiter | −1 dBTP |
| YouTube | −14 LUFS | turned down | left as they are | −1 dBTP |
| Tidal | −14 LUFS | turned down | left as they are | −1 dBTP |
| Amazon Music | −14 LUFS | turned down | left as they are | −1 dBTP |
| Deezer | −15 LUFS | turned down | turned up | −1 dBTP |
| Apple Music (Sound Check) | −16 LUFS | turned down | left as they are | −1 dBTP |
| Club and DJ sets | no normalization | play louder | play quieter | −0.3 dBTP |
These are the levels the services publish or that measurements agree on; each one also changes its own behaviour from time to time, so treat them as targets, not laws.
What LUFS actually measures
LUFS (Loudness Units relative to Full Scale) is the loudness measure defined in ITU-R BS.1770. Instead of reading peaks or plain RMS, it filters the signal the way the ear weights frequencies (K-weighting: less low end, a little more presence) and averages it over time.
- Integrated loudness: the whole song, with quiet passages gated out (an absolute gate at −70 LUFS and a relative gate 10 LU under the average). This is the number streaming services use.
- Short-term loudness: a 3-second window. Useful to see how loud the chorus is compared with the verse.
- Momentary loudness: a 400 ms window, the fastest reading.
- Loudness range (LRA): how much the loudness moves across the song, in LU. Ballads and film music have a wide range; dense dance tracks a narrow one.
Two songs at −14 LUFS integrated sound about equally loud even if one peaks much higher than the other. That is why services normalize with LUFS and not with peaks.
True peak, and why −1 dBTP
Digital audio is stored as samples, but the waveform your speakers reproduce passes between them. Between two samples just under 0 dBFS the real wave can go over 0 dB: an inter-sample peak. A true-peak meter oversamples the signal to find those peaks, and reports them in dBTP.
Streaming services convert your WAV to lossy formats (Ogg Vorbis, AAC, Opus). Those encoders change the waveform slightly and can push peaks higher, so a master that touches 0 dBTP can clip after encoding. Spotify asks for −1 dBTP, and −2 dBTP if your master is louder than −14 LUFS. A true-peak limiter at −1 dBTP keeps every version clean.
Should you master to −14 LUFS?
Not necessarily. Normalization means a master at −8 LUFS is turned down about 6 dB on Spotify: it ends up as loud as a −14 LUFS master, but with less dynamics, because the limiting that made it loud stays in. A master near −14 LUFS keeps its punch and its quiet moments, and nothing is turned down.
Many genres are still mastered louder on purpose. Dense, upfront sound is part of how trap, EDM, metal or pop sound, and DJs, radio and some platforms play files without normalization. The useful question is not “how loud” but “how dense”:
- Acoustic, jazz, classical, singer-songwriter, film: around −16 to −14 LUFS, where the music breathes.
- Most pop, hip-hop, R&B, rock: around −12 to −9 LUFS for a fuller, more forward sound.
- EDM, trap, metal, club edits: around −9 to −7 LUFS when density is part of the style; keep the true peak in check.
Whatever you choose, compare masters at the same loudness. Louder always sounds better for a few seconds, so a fair comparison turns that advantage off; it is how you hear whether more limiting really helped.
Why your song sounds quieter than others on Spotify
- Its peaks are high compared with its loudness (a high peak-to-loudness ratio). Spotify only turns quieter songs up as far as their peaks allow, so a quiet master with sharp peaks stays quiet.
- Its tone has less energy between about 2 and 5 kHz, where the ear is most sensitive. Two songs at the same LUFS can feel different in loudness because of their balance.
- The listener has Apple Music or Spotify set to a different level, or normalization turned off. Then the louder master wins, which is why some genres keep mastering loud.
How to check your song for free
Measure the integrated loudness and the true peak of the final file, not of your session: dither, sample-rate conversion and the export itself change both. OBRA Audio Mastering measures your mix and its three masters in the browser, nothing is uploaded, and its Platforms view shows how Spotify, Apple Music, YouTube and the rest will play each one.
Frequently asked questions
What LUFS is best for Spotify?
Spotify plays at about −14 LUFS in its default setting. Mastering near −14 LUFS keeps your dynamics and nothing is turned down; louder masters are turned down to the same level. Keep the true peak at −1 dBTP, or −2 dBTP if your master is louder than −14 LUFS.
Is −14 LUFS too quiet?
Not on streaming services, where louder masters are turned down to about the same level. It can sound quiet in a DJ set or next to unnormalized files, which is why dense genres are often mastered around −9 LUFS.
Does YouTube turn quiet songs up?
No. YouTube turns louder uploads down to about −14 LUFS and leaves quieter ones as they are, so a quiet master plays quieter than others.
What is the difference between LUFS and dB?
dBFS readings are peaks or averages of the raw signal. LUFS is a loudness measure weighted for human hearing and gated over time (ITU-R BS.1770), so it tracks how loud music sounds rather than how high its peaks go.
What true peak should I use?
−1 dBTP for streaming, −2 dBTP if the master is louder than −14 LUFS, and around −0.3 dBTP for files that go straight to DJs or clubs.