Mastering

Online Music Mastering That Actually Masters

Not a volume knob and not a one-size-fits-all filter — KROMA applies an ordered mastering chain (EQ → compression → harmonics → loudness) tuned on AI-generated music and rough mixes, then lands the file on your platform's loudness standard.

KROMA masters a track to streaming loudness in one click — pick a platform preset (Spotify, Apple Music, YouTube, Club) and the EQ, compression, harmonics and limiting chain runs automatically. No mastering engineer or plugin suite needed.

Interactive A/B Demo

Electronic Track → Club Master

Midnight Club tightens the low end and brings the master to club loudness — the kick hits harder without clipping the limiter.

Open in Studio →

Demo: “Volatile Reaction” by Kevin MacLeod — CC BY 3.0

How it works

  1. 1

    Drop in your track

    WAV or MP3 — a Suno/Udio generation, a demo bounce, a voice memo idea, a bedroom mix.

  2. 2

    Choose a mastering style

    Pristine Clear for release polish, Vocal Focus to pull the singer forward, Kinetic Sub for low-end weight, plus gentler texture options.

  3. 3

    Set the platform target

    Spotify (−14 LUFS), Apple Music (−16), YouTube (−14), Club (−9) — or match the source loudness to keep dynamics untouched.

  4. 4

    Download the master

    A release-ready file: balanced, glued, on-loudness — upload it and it survives the platform's normalization untouched.

No mastering engineer needed — the chain is already built

A mastering chain is a sequence of specialist decisions: tonal balance, dynamics, harmonic saturation, ceiling, loudness target. KROMA packages that chain as platform presets — the same processing a studio applies, run automatically on upload. The limiter ceiling and LUFS target for each platform are set for you; you just listen.

What 'mastered' actually means here

Mastering is the last step that makes a song sound like a release instead of a bounce: tonal balance (no mud, no harsh edge), dynamic glue (the song breathes but holds together), harmonic density (the subtle saturation that makes things feel expensive), and finally loudness that matches the platform's playback standard.

KROMA applies that order — EQ, then compression, then harmonics, then a loudness stage under a limiter — the same chain a mastering engineer would run. What you get back is not just louder; it's consistent, so Spotify's normalization leaves it alone.

Tuned for AI-generated music

AI songs fail mastering in predictable ways: a thin, rolled high end above ~12 kHz, dynamics that were never glued, and loudness that lands several dB off target. The mastering styles here were built on exactly that material — restoring air, adding low-end weight, and settling the mix before the loudness stage.

It works just as well on human-made demos, voice memos and rough mixes — anywhere the song is good but the bounce isn't.

Master a section before the whole song

The segment selector lets you master just the chorus, or a 30-second window, and hear it spliced back into the original — so you can compare styles on the part that matters before committing to a full render. Processing a section costs less and iterates faster.

Questions

How is this different from a normalizer or loudness tool?

A normalizer only changes gain — it can't fix tonal balance, dynamics or thin high end. KROMA runs EQ, compression and harmonic stages first; the loudness target is the last step, not the only one.

Which platform target should I pick?

Spotify −14 LUFS for Spotify/YouTube uploads, Apple −16 for Apple Music, Club −9 if you want a hot DJ-ready file. 'Match source' keeps your original dynamics — use it when the mix already sounds right.

Will it over-compress my song?

No — the compressor glues dynamics gently and the limiter only catches peaks. If you want zero dynamic change, choose 'Match source' loudness and the lighter enhancement styles.

Can I master only part of the track?

Yes — the segment selector processes a time range (e.g. the chorus) and splices it back seamlessly, so you can audition styles cheaply before rendering the whole song.

Does it work on Suno or Udio exports?

Yes — that's a primary use case. AI outputs are usually off-target on loudness and thin above 12 kHz; the mastering styles were tuned specifically to fix both.

Master a track →

Free to try — no install, works in the browser.