How loud should my master be? Real numbers from 83 commercial releases
Ask ten producers how loud a master should be and you get ten numbers. We measured 83 commercial releases instead. Here is where they actually sit, and how to decide for your own track.
· 5 min read
Where commercial releases sit
Across 83 recent pop and K-pop releases we measured, integrated loudness clustered around −9 to −10 LUFS, with true peaks between −0.5 and −1 dBTP. Dance and hip-hop releases run a little hotter; acoustic and ballad releases a little quieter, around −11 to −12.
That is the file loudness. On Spotify or YouTube every one of those tracks is turned down to −14 on playback. The producers know that; they master loud anyway because the same file goes to downloads, radio, DJs and sync, where there is no normalization.
The cost of loudness
Loudness is paid for with peak-to-loudness ratio (PLR): the gap between the peaks and the average. A −15 LUFS Suno track with peaks at −4 dBTP has a PLR of about 11 dB. A −10 LUFS master with peaks at −1 dBTP has a PLR of 9 dB. Getting there means removing 2 dB of that gap — through saturation, compression and limiting — and every dB removed is a little less punch.
Up to a point that is fine and even desirable; density is part of what makes a master sound finished. Past that point it becomes pumping and distortion. A good master spends as little PLR as it can to reach the target.
Choosing a target
| Your track is going to… | Target | Why |
|---|---|---|
| Spotify / YouTube only | −14 LUFS | Maximum dynamics; the platform plays it at full level |
| Apple Music only | −16 LUFS | Sound Check target |
| Everywhere (streaming + downloads) | −9 to −10 LUFS | Holds up where there is no normalization |
| Clubs, DJs | −8 to −9 LUFS, −0.8 dBTP | No normalization; loudest file wins |
| Acoustic, ballad, jazz | −11 to −12 LUFS | Genre convention; dynamics are the point |
In MasteringOne the release target menu does exactly this. The engine reports the limiter's gain reduction so you can see how much loudness cost: under 3 dB is comfortable, and the report flags it if the limiter had to work harder than the source warranted.
FAQ
Is louder always better?
Only in a direct comparison with no normalization. On streaming platforms it is neutral at best; past the platform target it is worse, because you lose dynamics and gain nothing.
What is the limiter gain reduction in my report?
How many dB the limiter had to cut on the loudest peaks to reach the target. Low numbers (under 2–3 dB) mean the master got there mostly through tone and density, which is what you want.
Try it on your own track
Free 60-second preview. No account needed.