Restore the dynamics of an over-compressed track
Does the track sound flat and like a brick, squashed by heavy compression or limiting, or was it delivered that way by an AI generator? The lab’s Already loud style gives back the punch and the crest factor. It emphasises the transients and opens up the sound, without adding hard clipping. It runs in the browser, no DAW needed.
What Already loud does
Limiters and AI generators often push the whole track against the ceiling. The transients are flattened, and the crest factor, the gap between the peaks and the average level, drops. The track sounds dense but tired. Already loud works in two steps:
- It opens the squashed section. A slow-acting downward expander gently lowers long, loud sustain, so the contrast between sections comes back.
- It restores the attacks. The transient content, the fast part of the envelope, is emphasised, which brings back the hit of drums and the pluck of strings.
The stereo image is preserved, because the envelope is shared between the channels. The stage runs before compression and loudness normalisation, and the lab sets the final level.
What it does not do
Information that a limiter has cut cannot be brought back. What you get is the feel of the impact and more air, not the original mix. The best results come from tracks where the transients still survive under the compression. A track squashed to the last decibel may change very little.
The style also keeps its own loudness target, −12 LUFS in the lab, instead of the usual platform level. It is not a way to make a track louder: that is the job of mastering, and a too-hot input leaves little headroom.
Who it is for
- AI-generated music (Suno, Udio). Generators often return a track as a brick. Already loud opens it before release.
- Old masters squeezed for loudness. Records made in the loudness-war era can breathe again.
- Mixes that were compressed before mastering. If the source is already over-compressed, the stage returns part of the punch.
- Anyone who streams. Platforms turn loud tracks down anyway, so an open, dynamic track keeps its advantage.
How to use it
- Open the lab and upload a WAV, MP3 or FLAC file.
- Choose the Already loud style. Mastering in the lab runs on that choice.
- Check the analysis tiles: they show the crest factor and the headroom, so you can compare the result with the input.
- Listen, then download the result.
If the track is clipped as well as squashed, check it on the De-clipper page first. Clipped peaks and a squashed sound are two different problems.
Questions
Can the dynamics of an already squashed track be restored?
The limiting cannot be undone, but the perceived punch can be brought back. Already loud emphasises the transients and raises the crest factor at the same loudness, without hard clipping.
Does it work on AI-generated tracks from Suno or Udio?
Yes. Generators often deliver a track close to the ceiling with a small crest factor. Already loud gives back the hits and the air, and mastering then brings the level to the target.
Does it add distortion?
The stage works on the signal envelope and does not hard-clip peaks. If the audio would go over the ceiling, the whole buffer is scaled down, so the restored dynamics stay in proportion.
What is crest factor?
The difference between the peak level and the average loudness. A brickwalled track has a small crest factor: its transients are flattened and the sound feels tiring.
Read also: De-clipper · Audio restoration · True peak · AI artifact cleanup · Mastering for Suno
Open a squashed track?
Already loud is in the lab. Upload the track and pick the style.
Open the lab