Suno loudness report
We measured 248 Suno exports uploaded for mastering. Most are not too loud for Spotify. The bigger issue is how little they breathe.
The numbers
| Measure | Result |
|---|---|
| Unique Suno exports | 248 |
| Median integrated loudness | −14.3 LUFS |
| Middle half of tracks | −15.0 to −13.5 LUFS |
| Louder than Spotify's −14 LUFS | 36% |
| Louder than −10 LUFS | 6% |
| Quieter than −16 LUFS | 8% |
| Median turn-down on Spotify (tracks above −14) | 0.9 dB |
| Median loudness range (LRA) | 5.8 LU |
| Loudness range under 5 LU | 40% |
| Median crest factor | 14.6 dB |
What it means
Suno exports are mostly at streaming loudness already. The common advice “Suno songs are way too loud, turn them down” fits only a small minority: 6% of tracks are louder than −10 LUFS. Half of them sit between −15 and −13.5 LUFS, where Spotify changes the level by about a decibel at most.
The weak spot is dynamics. 40% of the exports have a loudness range under 5 LU: verses and choruses play at nearly the same level. On such tracks a generic “make it loud” master flattens what is left and brings the metallic AI shimmer forward.
Crest factor is healthy. The median peak-to-average ratio is 14.6 dB, and no export in the sample was crushed below 9 dB. The tracks are not brick-walled; they are evenly loud.
Before you release a Suno song
- Measure loudness, true peak and loudness range: free LUFS checker.
- Keep true peak at or under −1 dBTP so lossy encoding does not clip. See true peak explained.
- Clean the AI artifacts before limiting, not after. See AI artifact cleanup.
- Aim for about −14 LUFS for Spotify and YouTube. See Spotify loudness.
- Master one track free every day: free AI mastering.
How we measured
Each file was measured as uploaded, before any processing: integrated loudness by ITU-R BS.1770 with gating, loudness range by EBU R128, crest factor on the worst channel. We counted one measurement per unique file name containing “suno” among masters completed between 26 July and 2 October 2026, excluding internal test runs. The sample comes from people who chose to master their track, so it is not a random sample of everything made with Suno. Only aggregate numbers are published; no audio, titles or user data.
You may cite these numbers with a link to this page (CC BY 4.0).
Are Suno songs too loud for Spotify?
Usually not by much. In our sample of 248 Suno exports the median is −14.3 LUFS, close to Spotify's −14 LUFS reference. About a third are louder than −14 LUFS and get turned down a little (median 0.9 dB); only 6% are louder than −10 LUFS.
What is the real problem with Suno exports, then?
Dynamics and artifacts. 40% of the exports have a loudness range under 5 LU, so the song barely breathes; a generic limiter on top of that exposes the metallic shimmer typical of AI-generated audio.
How was this measured?
Integrated loudness (ITU-R BS.1770, gated), loudness range (EBU R128 LRA) and crest factor were measured on the original upload, before any processing. One measurement per unique file name containing “suno”, July–October 2026.
Should I master a Suno song before releasing it?
If it will go to Spotify, Apple Music or YouTube, yes: check loudness and true peak, keep true peak at or under −1 dBTP, and clean the artifacts before limiting. You can check any track free with the LUFS checker and master one track free every day.
Can I use these numbers?
Yes, with a link to this page. They are aggregate measurements only: no audio, titles or user data are published.