Your Suno song is quiet because it was never mastered. Suno exports a finished mix: across 337 untouched Suno downloads we measured from one creator's library, the median was -15.3 LUFS integrated for WAV and -14.4 for MP3, and none was louder than -10.5. Most sat below Spotify's -14 LUFS reference level, and no volume slider closes that gap — turning up playback raises everything, so your track is still quieter than the next song in the playlist. The fix is one mastering pass that raises the track to -11 LUFS integrated (TrackGleam's default, or -14 on the Low option) while holding true peak at -1.0 dBTP. It takes about two minutes, runs free in your browser, and nothing uploads.
If you want the fix before the explanation, drop your export on the Suno mastering page and listen — the free master is a real download, not a preview clip. The rest of this page is why it happens and how to check the numbers yourself.
Quiet enough to measure, not just to feel. We measured 337 untouched Suno downloads from one creator's library: 218 WAVs and 119 MP3s, downloaded between February 2024 and September 2026. They come from several Suno versions, and no file says which version made it. The WAVs had a median of -15.3 LUFS integrated (the middle 80% ran from -16.3 to -14.0), and the MP3s -14.4 (-16.0 to -12.2). The loudest file read -10.5 LUFS. Most sat below -14 LUFS, the level Spotify normalizes to: 89% of the WAVs and 65% of the MP3s. The median loudness range was about 5 LU, with about 12 dB between loudness and true peak, so these files aren't squashed. The problem is level.
Put that next to a track mastered to -11 LUFS, TrackGleam's default, and the median WAV is about 4 LU quieter. Wherever playback isn't normalized, that gap is easy to hear: the next song jumps out. Nothing is broken in your file; a stage is simply missing from it.
We decoded each file and measured it with an ITU-R BS.1770-4 script that passes the EBU Tech 3341 and 3342 test signals and reads the same loudness as the mastering studio's own meter, to within 0.0001 LU (September 2026). Older downloads carry no Suno tag, so 292 of the 337 were identified as Suno by their file format, name and date. One creator's library is not a sample of all Suno output.
Because volume is relative and loudness is not. Your system volume raises your Suno track and everything else by the same amount, so the gap between your song and the one after it in the playlist never changes. Loudness is a property of the file — how much energy it holds over time — and only the file can change it.
The obvious follow-up is "then just apply gain". That fails for a different reason: a mix's peaks already sit within a few dB of the top of the scale, so plain gain runs them into the ceiling before the track gets loud. The Suno downloads we measured had a median true peak of -3.5 dBTP (WAV) and -2.1 dBTP (MP3), so plain gain takes a typical file only to about -13 LUFS before its peaks reach -1 dBTP. Peak level and perceived loudness are not the same measurement — you can have a file whose peaks touch 0 dBFS and still measures -16 LUFS. Getting louder without clipping requires dynamics processing, not a fader; the mechanics are in making a song louder without clipping.
LUFS (Loudness Units relative to Full Scale) measures perceived loudness across the whole track, weighted the way human hearing actually responds, and gated so silence between phrases does not drag the number down. It is the number streaming platforms use to decide how much to turn your song up or down. Integrated LUFS is the single figure for the whole song; loudness range (LRA) describes how much it moves between quiet and loud sections; true peak (dBTP) is a separate measurement of the highest sample value including the peaks that appear between samples after conversion. Loudness tells you how big it sounds; true peak tells you whether it will distort. You need both, and the full explanation of each lives in the LUFS streaming targets guide and what dBTP actually measures.
A mastering pass raises the track by shaping it rather than just amplifying it: corrective EQ clears the low-mid congestion around 250-500 Hz that eats headroom, gentle compression evens out the level so there is less peak to work around, and a true-peak limiter lands the file on target without letting anything through the ceiling. On TrackGleam the default is -11 LUFS integrated with a -1.0 dBTP true-peak ceiling, and the finished file is re-measured so the numbers you see are the numbers in your download.
Why -11 and not louder? Because past a point you stop gaining anything. Platforms normalize playback down toward roughly -14 LUFS, so a -6 LUFS master does not play louder than a -11 one — it just arrives with less dynamic range intact after being turned down by more. And why -1.0 dBTP rather than 0? Because the lossy encoders every platform uses push inter-sample peaks higher than the file's own sample values; leaving a decibel of room is what keeps that from turning into audible distortion. If you'd rather follow Spotify's own -14 LUFS recommendation, the Low option is one click.
One master covers all of them — platforms that normalize simply play your file back around their own level. As of July 2026:
| Platform | Normalizes toward | What happens to a -11 LUFS master |
|---|---|---|
| Spotify | -14 LUFS (documented) | Turned down about 3 dB, cleanly — density and punch survive |
| Apple Music | About -16 LUFS (Sound Check) | Plays about 5 dB down; Sound Check is user-toggleable |
| YouTube | Around -14 LUFS | Turned down, not up — quiet uploads stay quiet |
| TikTok / Reels / Shorts | Around -14 LUFS, but playback context varies wildly | Density matters more here than the exact number |
| An untouched Suno download (median -15.3 LUFS WAV, -14.4 MP3, in the library we measured) | — | Below the -14 targets; turned up only within peak limits, if at all |
Verified July 2026 — we re-check these quarterly.
The asymmetry in that table is the whole story: normalization turns loud files down reliably, but it will not rescue a quiet one, because boosting is limited by the file's own peaks. That is also why the same song can sound quieter on Spotify than it did on your desktop — covered in why your song is quieter on Spotify.
No DAW, no plugins, no account:
1. Export the best file your Suno plan gives you — WAV where available, MP3 otherwise. Master the MP3 if that is all you have; just export the finished master as WAV so you never stack lossy on lossy.
2. Open the mastering tool and drop the file in. Everything runs on your own device through the Web Audio API — you can watch the network tab in your browser's developer tools and see that no audio leaves.
3. Read the "before" numbers: integrated LUFS, true peak, loudness range, measured off your actual file.
4. Master to -11 LUFS / -1.0 dBTP and compare with the volume-matched A/B. Matched volume is the point — it strips out the "louder sounds better" illusion so you are judging tone and punch, not level.
5. Download the WAV. Free, no signup, no watermark, full length.
Releasing more than one track? Batch the whole set to the same target so song 3 does not jump out of song 4 — see mastering a whole Suno album at once and making all your songs the same volume.
Then it was never really a loudness problem — it was a tone or arrangement problem wearing loudness as a disguise, and now you can hear it clearly. Three usual suspects. Low-mid congestion around 250-500 Hz makes a track feel thick but small; cutting there does more for perceived size than any amount of extra limiting. Sub-30 Hz rumble you cannot hear still eats limiter headroom, so a high-pass buys you real loudness for free. And a wide, phasey stereo image can collapse on mono playback, which is where a lot of listening happens — check it before blaming the master.
If the track still reads as harsh, washy or muffled rather than quiet, that is the other half of the AI-export problem set, triaged symptom by symptom in why your Suno song sounds off, with the full walkthrough in how to master a Suno song and the generator-agnostic version in how to master AI-generated music. And if the render itself is broken — garbled syllables, smeared transients — regenerate before polishing; no master fixes what the model baked in.
Because Suno exports a finished mix, not a master. Across 337 untouched Suno downloads we measured from one creator's library, the median was -15.3 LUFS integrated for WAV and -14.4 for MP3, and none was louder than -10.5. Mastering raises the track to a loudness target, -11 LUFS by default in TrackGleam, with true peak held at -1.0 dBTP.
Master it rather than amplifying it. Corrective EQ clears low-mid congestion that eats headroom, compression evens out the level, and a true-peak limiter lands the file at -11 LUFS integrated without letting peaks past -1.0 dBTP. TrackGleam does exactly that free in your browser, with nothing uploaded.
Spotify recommends mastering to -14 LUFS integrated with true peak below -1 dB, or below -2 dB for masters louder than -14, and plays tracks at -14 on its Normal setting (checked September 29, 2026). TrackGleam’s default is -11 LUFS with a -1.0 dBTP ceiling, which is louder: Spotify turns it down at playback, and the -1.0 ceiling does not meet the -2 dB advice for louder masters. Pick the Low option for a -14 file.
No. System volume raises your track and every other track equally, so the gap stays. Plain gain does not work either — a mix already peaks near full scale, so gain clips before it sounds loud. Peak level and perceived loudness are different measurements.
Only within limits. Normalization turns loud masters down reliably, but boosting a quiet one is capped by the file's own peaks, and normalization is off in some listening contexts entirely. Where it is off, a track that measures -15 LUFS plays about 4 dB quieter than one mastered to -11.
Not at a sensible target. The Suno downloads we measured had a median loudness range of about 5 LU and about 12 dB between loudness and true peak — room to work with — and -11 LUFS is well short of the levels where limiting audibly flattens a track. The volume-matched A/B is there to check: if the master does not win at equal level, keep the original.
It is a real download — a full-length WAV or MP3, no watermark, no account, no track limit. Processing runs in your browser on your own device, so nothing uploads. The optional GleamAI master is $1.99 (plus tax where applicable) per track and never required.
Everything here runs in your browser — nothing uploads.
Master a track free — no signup, nothing uploads
Includes the AI Fix presets for AI-generated tracks.
Ready to try it? Master a track free with TrackGleam's online audio mastering or clean up a voice recording — no signup, nothing uploads.
Suno track sounds muffled, quiet, or harsh? A symptom-by-symptom triage: causes, fixes, and free browser mastering that never uploads your audio.
Master your Suno songs free in the browser — no signup, no upload. Hit -11 LUFS at -1.0 dBTP (or -14 on Low), fix mud and sheen, and keep a real WAV.
Every platform's LUFS and true-peak target in one dated table — and why -11 LUFS is the new release default while platforms normalize playback to about -14.
Online volume boosters clip your peaks. Make a song genuinely louder — true-peak limiting to a release-competitive −11 LUFS, free, in your browser.
Every guide and comparison, in one place.