How we measured
12 AI-generated exports, ITU-R BS.1770-4 gated loudness + dBTP true peak, measured client-side in the TrackGleam engine, July 2026.
To make a song louder without clipping, raise its average loudness with a true-peak limiter — up to −14 LUFS integrated with a −1.0 dBTP ceiling — instead of multiplying the gain. Flat "volume booster" gain squares off any peak that's already near 0 dBFS, and Spotify turns everything above −14 LUFS back down at playback anyway, distortion included. The limiter route is free in your browser, with nothing uploaded.
Because it did exactly what it advertised. "Make MP3 louder" sites apply flat gain — a percentage or a fixed number of decibels across the whole file. One popular converter offers boosts from 10% to 200%, or 1 to 30 dB, and its page never mentions a limiter or any clipping protection (checked July 2026). Here's the math that bites you: most finished tracks already peak within a decibel or two of 0 dBFS, the hard ceiling of digital audio. Add 6 dB of flat gain and every peak that was above −6 dBFS has nowhere to go — the waveform squares off at the ceiling. That squared-off top is clipping, by definition, and it's the crunchy distortion you heard.
Most of these sites also decode and re-encode your MP3 to do it, so you keep the clipping and pay a second generation of lossy compression. Your instinct that the tool made things worse is correct. (If it was a mastering service rather than a booster that came back distorted or slammed, our honest rundown of the alternatives covers that comparison.)
Gain moves every sample up or down together — it changes peak height. Loudness is what your ears track: the average energy across the whole song, measured in LUFS (a gated average defined by ITU-R BS.1770). A track can have tall peaks and a quiet body, or modest peaks and a dense, loud body. "Make it louder" correctly means raising the body while keeping the peaks under control — which is a limiter's job, not a gain slider's. The full explainer lives in our LUFS streaming targets guide.
And genuinely quiet files are common, especially unmastered and AI-generated exports. We measured 12 AI-generated exports from our test library: median −15.2 LUFS integrated (range −16.4 to −12.3), with 8 of 12 below the −14 LUFS streaming level — and 3 of 12 already over the −1.0 dBTP true-peak ceiling, meaning flat gain would push them straight into clipping.
12 AI-generated exports, ITU-R BS.1770-4 gated loudness + dBTP true peak, measured client-side in the TrackGleam engine, July 2026.
Loudness normalization. As of July 2026, Spotify plays tracks at about −14 LUFS: masters hotter than that are turned down at playback, and quiet masters are boosted only as far as their own peak headroom allows, with 1 dB held in reserve for encoding. So a naive boost to −9 LUFS gets turned down roughly 5 dB — and the clipping you baked in survives the turn-down. You traded fidelity for a loudness advantage that evaporates the moment anyone presses play. YouTube behaves the same direction — it turns loud content down but never boosts quiet content up, which you can verify on any video via Stats for Nerds — and Apple's Sound Check normalizes near −16 LUFS. If your track sounds quieter than everyone else's on Spotify specifically, the full diagnosis is in why your song sounds quieter on Spotify.
There's a second trap even careful boosters fall into: true peaks. A file whose samples never touch 0 dBFS can still reconstruct above full scale between samples when it's converted back to analog or squeezed through a lossy encoder — that's an inter-sample peak, measured in dBTP. Spotify's own guidance warns that hot true peaks can distort during transcoding, which is why a track that sounded fine on your machine comes back fizzy from the platform. The fix is a limiter with a true-peak ceiling, not a sample-peak one. The concept is unpacked in our true peak explainer, and you can check any file free with the true-peak checker.
Three ingredients. First, a limiter: it raises the body of the track into the space the peaks used to monopolize, transparently when it's not pushed too hard. That's more density, not more amplitude — density is what reads as loud.
Second, a loudness target: −14 LUFS integrated is the streaming sweet spot, loud enough that normalization leaves it alone, restrained enough that nothing gets crushed. Not sure what your material should target? The free what-LUFS-to-master-to tool answers that per platform and genre.
Third, a true-peak ceiling of −1.0 dBTP, so the master survives MP3 and AAC conversion without inter-sample overs. That ceiling is the difference between louder and louder-but-fizzy.
| Method | How it works | Clipping risk | What Spotify does with it |
|---|---|---|---|
| Online volume booster | Flat gain — up to 200% or +30 dB, usually with an MP3 re-encode | High — no limiter; peaks near 0 dBFS square off | Anything pushed past −14 LUFS is turned back down; the distortion stays |
| Peak normalize (DAW/editor) | Raises gain until the tallest peak touches a ceiling | None — but it stops at your tallest peak, so a quiet, peaky song barely moves | Usually still under −14 LUFS, so still quiet |
| TrackGleam free master | EQ + dynamics + true-peak limiting to −14 LUFS integrated / −1.0 dBTP, re-measured after render | Controlled — the −1.0 dBTP ceiling is the whole point | Plays at full level; nothing to turn down |
Verified July 2026 — booster specs from the linked page; we re-check quarterly.
1. Drop your file into the TrackGleam mastering tool. Nothing uploads — every stage runs on your own device, and you can verify that in your browser's developer tools: no audio in the network tab. No account, either.
2. Read the analysis card first: integrated LUFS, true peak, loudness range, crest factor, stereo width, plus BPM and key. Now you know exactly how far under streaming level you're starting.
3. The engine renders a master targeting −14 LUFS integrated with a −1.0 dBTP ceiling — then re-measures the finished file, so the numbers in your report are what's actually in the download, not a promise.
4. A/B the original against the master, volume-matched at streaming loudness, switching with the A key. Matched volume is the honest test — louder always sounds "better," so the comparison removes that trick. Then download the WAV or MP3, free, no watermark.
Don't take the engine's word for it. Run the boosted file from earlier and the mastered file through the free loudness penalty checker and compare what each platform would do to them — the boosted file shows a turn-down penalty, the −14 LUFS master shows essentially none. Then run both through the true-peak checker: expect overs on the booster output and a clean −1.0 dBTP reading on the master. Two files, four numbers, no listening bias involved.
When the music is supposed to breathe. Ambient, lofi, acoustic, and classical material often sounds best below streaming loudness with its dynamics intact — that's why TrackGleam ships 16 genre presets including quieter, genre-appropriate targets for ambient and lofi, and why the report shows LRA and crest factor instead of pretending one number fits all (there's a standalone dynamic range meter too). Two honest limits while we're here: no limiter can un-clip distortion that's already printed into your source — if the booster output is the only copy you have, go back to the original file — and if you're cutting vinyl, book a human mastering engineer; an automated chain doesn't replace those ears for that job.
Same physics, batch workflow. Drop all of them into the queue at once and preview one free; any one-time credit purchase — from $1.99 — unlocks the bulk queue forever, with no subscription and no account, and "Master all" then runs the free engine on every track with no per-track cost. That's the one paid thing on this page, so it's worth saying plainly: single tracks are free forever, bulk is not. The full credit breakdown is on the pricing page.
Raise the average loudness with a true-peak limiter instead of adding flat gain: target −14 LUFS integrated with a −1.0 dBTP ceiling. That is what a mastering chain does — TrackGleam renders the master free in your browser, then re-measures both numbers so you can verify them in the report.
Those tools multiply the signal by a percentage or add fixed decibels with no limiter. If your file already peaked near 0 dBFS — most finished tracks do — every boosted peak squared off at the digital ceiling, which is clipping by definition. Many boosters also re-encode your MP3, adding a second generation of loss.
Only up to the normalization level. Spotify plays tracks at about −14 LUFS: quiet masters are boosted only within their peak headroom, and anything louder is turned back down with its limiting damage intact. Pushing past −14 LUFS buys nothing.
Not meaningfully — genuinely changing loudness means re-rendering the audio. Tag-based gain tricks only adjust playback metadata that many players and streaming platforms ignore. The clean move is to master from the best copy you have and download the result as WAV, so you never stack a second lossy encode.
Around −14 LUFS. YouTube turns loud content down but does not boost quiet content up, so a −14 LUFS / −1.0 dBTP master plays at full level. You can see the adjustment YouTube applied to any video under Stats for Nerds.
No. Ambient, lofi, acoustic, and classical material often sounds best below streaming loudness with dynamics intact — that is why genre presets with quieter targets exist. Judge with a volume-matched A/B so raw loudness cannot masquerade as quality.
Not with TrackGleam — analysis and mastering run entirely in your browser on your own device, and you can confirm in your browser's developer tools that no audio appears in the network tab. No account, no watermark, free WAV and MP3 downloads.
Master a track free — no signup, nothing uploads
Every master is measured: integrated LUFS, true peak, loudness range.
Every platform's LUFS and true-peak target in one dated table — Spotify, Apple Music, YouTube, Amazon, Tidal, TikTok — and how to hit -14 LUFS free.
Four reasons your track sounds quiet on Spotify — only one is loudness. Diagnose density, normalization, and the quiet-track penalty, then fix it free.
True peak (dBTP) measures peaks between samples. Why 0 dBFS masters still clip after lossy encoding, and why -1.0 dBTP is the streaming-safe ceiling.
Every guide and comparison, in one place.