Audio tools
AI Stem Mastering Engine
The final step. Upload your finished vocal covers or beats to apply real per-preset dynamics compression, true-peak limiting, and tonal EQ, then normalize to standard streaming loudness targets (LUFS).
Upload final mix
1 creditUpload File
or drop your file here
MP3, WAV, FLAC · max 100 MB
Target parameters
-14.0 LUFS
-24 (Quiet)-6 (Loud)
Set by the streaming preset — switch to “Custom Parameters” to change it.
Upload a final mix to enable this.
Analysis & output
Nothing mastered yet
Your mastered audio and LUFS statistics will appear here.
Questions answered
Frequently asked questions
The preset picks it for you: Streaming aims at −14 LUFS integrated, which is roughly where Spotify, YouTube and Apple Music normalise; Radio at −11; Club at −9. Every preset limits to a −1 dBTP true-peak ceiling so the file doesn't clip on conversion. Choose Custom to unlock the slider and set your own target. The integrated loudness, true peak and loudness range actually achieved are reported on the result.
They are real signal processing, not just different numbers. Streaming adds gentle bus compression and a touch of air above 8 kHz. Club lifts the low end, compresses harder and adds a limiter. Radio high-passes the sub, compresses firmly and brightens the top. Custom is deliberately transparent — loudness normalisation only, no tonal colouring — so a target you dial in yourself arrives uncoloured.
One credit per master by default, and you need to be signed in; the live price is always shown before you commit. You get back a WAV. If the job fails or times out, the credit is refunded automatically as part of the same transaction that marks the job failed — you are never charged for a master you didn't receive.
Master the final full mix, last. If you are building stems, separate first, reassemble, and master the combined result — mastering a stem in isolation and then remixing undoes most of what the mastering chain did. Accepted inputs are MP3, WAV, FLAC, M4A and OGG.