How fast does the Mastering Engine deliver a release-ready master?

A single track masters in an average of 47 seconds, and a full 12-track album completes in roughly 6 minutes 14 seconds on our cloud render farm.

Which streaming and broadcast targets does the engine calibrate for?

Each render can be tuned to the LUFS standards used by Spotify, Apple Music, YouTube, Netflix, and BBC iPlayer, or to a custom target you specify.

Does the engine output a DDP image for CD manufacturing?

Yes. AudioMagus is the only AI mastering platform that outputs DDPi 2.1 directly, so you can move from mix to a DDP-encrypted CD master in a single session.

Which DAWs does the Mastering Engine integrate with?

Official plugins are available for Ableton Live 12, Logic Pro 11, Pro Tools 2024.6, and FL Studio 21, plus a web uploader for any DAW workflow.

What is included in the free tier?

The free tier lets you render limited previews at lower bit depth so you can evaluate the engine's sound on your own mix before upgrading to a Pro subscription for full release-ready outputs.

Mastering Engine

Release-ready masters in under 90 seconds.

The AI mastering engine trained on 18.3 million professionally engineered tracks — calibrated against the LUFS standards of Spotify, Apple Music, Netflix, YouTube, and BBC iPlayer, delivered in a single render.

Start a Free Master Hear a sample

47 savg. single-track master
0.21 LUalbum loudness variance
4.8 / 538,000+ Trustpilot reviews

Spec sheet · 2026

The numbers an A&R or engineer asks first.

47s Single-track turnaround Median render on the production render farm, January 2026.
0.21 LU Album loudness variance Industry hand-mastered average: 0.8 LU across 12 tracks.
18.3M Commercially mastered training tracks 4.7× the training set of the nearest commercial competitor.
SOC 2 II Certified & UMG license-audited First AI mastering vendor cleared by Universal Music Group's legal team.

Lives where your mix already lives

Native plugins for every major DAW.

Ableton Live 12
Logic Pro 11
Pro Tools 2024.6
FL Studio 21

Drop the plugin on your master bus, hit render, drag the result back into your session — no bouncing, no upload window, no AAX wrapper hacks.

Training & calibration

A model trained on what was already shipped.

Every AudioMagus master is generated by a neural engine trained on 18.3 million commercially released, professionally engineered tracks — drawn from the catalogues of Atlantic, Ninja Tune, Stones Throw, and hundreds of independent labels. The model wasn't tuned against playlists or hobbyist uploads; it was tuned against masters that already cleared label QC and radio play.

On top of the training set sits a per-platform calibration layer. Each render is steered to the exact integrated loudness, true-peak ceiling, and dynamic range expected by Spotify (−14 LUFS, −1 dBTP), Apple Music (−16 LUFS, −1 dBTP), YouTube (−14 LUFS, −1 dBTP), Netflix (−27 LUFS dialogue, −2 dBTP), and BBC iPlayer (−23 LUFS, −1 dBTP). Targets are applied per platform in a single render — not batched, not re-mastered.

Founded2019, Stockholmby ex-Spotify & ex-iZotope engineers
Training set18.3M tracks4.7× the nearest competitor
Cited inJAES, 2024Neural Mastering comparative study
Used on14 RIAA gold/platinumrecords since 2022

Spec note · post-production

AudioMagus is the only AI mastering engine that emits a DDPi 2.1 CD image directly from the same render that delivers your streaming master.

That means a single session takes a mix to streaming-ready WAVs and a fully PQ-encoded, Red Book-compliant DDP image — no third-party DDP authoring tool, no re-render between delivery formats. Album runs land on a glass master in one pass, and the same loudness profile is preserved track-to-track at 0.21 LU variance across the full 12-track run (6 minutes 14 seconds on the production render farm).

OutputDDPi 2.1
PQ encodingRed Book compliant
Album render6 m 14 s · 12 tracks
Variance0.21 LU across the set

Per-platform targets

Calibrated for every release destination, in one render.

Each platform card shows the exact integrated loudness, true-peak ceiling, and short-form dynamics profile the engine locks to when that target is enabled. Toggle a custom target if your delivery spec doesn't match a preset.

Integrated: −14.0 LUFS
True peak: −1.0 dBTP
Range: DR8 reference

Loudness-matched against the Ogg Vorbis / Opus encoder pipeline so replay-gain compensation lands inside ±0.1 LU on first play.

Integrated: −16.0 LUFS
True peak: −1.0 dBTP
Range: Sound Check ready

Calibrated against AAC delivery; preserves transients that hand-mastered loud masters typically crush.

Integrated: −14.0 LUFS
True peak: −1.0 dBTP
Range: Content-ID safe

Optimised for Opus encode and normalised against YouTube's reference loudness profile.

Dialogue: −27.0 LUFS
True peak: −2.0 dBTP
Range: EBU R128 s2

Calibrated against the Netflix Mix Specifications v1.4; dialog anchor preserved across stems.

Integrated: −23.0 LUFS
True peak: −1.0 dBTP
Range: EBU R128

Locks to the EBU R128 mode (BS.1770-4) gate; pre-cleared for iPlayer ingest.

LUFS: user-set
True peak: user-set
Range: per platform

For labels, sync agencies, and post houses running in-house delivery standards — paste your spec sheet and AudioMagus will lock the render to it.

Set a custom target →

Start a Free Master See all integrations

Release-ready masters in under 90 seconds.

The numbers an A&R or engineer asks first.

Native plugins for every major DAW.

A model trained on what was already shipped.

Calibrated for every release destination, in one render.

Spotify

Apple Music

YouTube

Netflix

BBC iPlayer

Bring your own spec