AI Mastering Checklist Before Release — Free MegaMix Trial
Confirm the mix export is mastering-ready
Avoid slamming a loud maximizer across your master bus before mastering. Leave about 3 to 6 dB of peak headroom so mastering can shape tone without fighting limiting artifacts. If the mix already looks like a finished streaming master, you have nothing left to work with except more distortion.
Check that fades and final reverbs are printed the way you intend. Mastering rarely fixes a chorus that cuts off mid-tail or a vocal delay that gets chopped at 3:02. Listen to the last four bars and the first downbeat in solo-ish focus: no clicks, no DC thump, no accidental sine sweep leftover from a sound-design track.
If the vocal is still 2 dB too quiet in the mix, fix that before mastering. A global 2 dB vocal EQ boost in mastering will also lift hats and sibilance.
Set realistic loudness expectations
Spotify, Apple Music, and YouTube normalize playback. Extreme brick-wall limiting often reduces punch after that adjustment. A master that hits -6 LUFS integrated with no transients left will not “win” against a punchier master that lands nearer typical album loudness; it will just sound smaller once levels are matched.
Prioritize punch and dynamics that survive normalization rather than peak loudness alone. Hip-hop can still feel loud with a readable kick. Acoustic singer-songwriter should not be crushed to match a hyperpop single. Use LUFS as a guide, not a religion.
If your only goal is “louder than the last SoundCloud rap beat,” you will over-limit. Goal is translation plus competitive density for your lane.
Stereo image and mono compatibility
Listen in mono and verify lead vocal, snare, and bass remain stable. If the hook stack disappears in mono, you have a width source that is out of phase, often a Haas-delayed double or a stereo widener on the whole mix. Fix that in the mix, not with more mastering width.
Watch low-end stereo widening. Wide energy below 120 Hz causes translation problems on phones, laptops, and club systems that sum to mono in the subs. Keep the 808 and kick as a center anchor.
Choruses can still feel wide from doubles, hats, and FX. Width is arrangement, not a stereoizer on the master.
Final translation passes
Run quick checks on earbuds, phone speaker, car if you have it, and whatever monitors you own. Pay attention to vocal intelligibility and harsh syllables at 3 to 8 kHz. A master that is exciting on studio headphones and painful on AirPods is not done.
Compare with a trusted reference at matched loudness. Take notes: snare too dull, vocal too far, bass disappearing on the phone. Do not turn your master up to “win” the A/B.
Walk away for 20 minutes if you have been limiting for an hour. Ear fatigue reads as “needs more treble.”
Release-day mastering workflow
Treat mastering as a short, separate session from mixing. Same-day masters are possible for simple songs if the mix already translates. Dense metal or very dynamic jazz needs a night of distance more often than another 2 dB of limiter.
Export WAV at the mix sample rate. Encode AAC or MP3 only after you approve the master.
- Import the mix WAV with true-peak headroom, not an MP3.
- Level-match a reference chorus and write three tonal notes only.
- Apply AI mastering in MegaMix AI, then judge punch after matching loudness.
- Mono check vocal, kick, and bass.
- Phone-speaker check lyrics; earbud check sibilance.
- Export WAV, then create distributor encodes from that file.
Pre-release mastering checklist
Print this next to your upload form. Skipping two items is how a “final” master still has a click at 0:00.
If a line fails, go back to mix or re-export. Mastering is a bad place to discover the verse vocal was the rough take.
- Mix peaks have headroom; no mix-bus clipper at -0.1 dB on the bounce.
- Start and end are clean; tails are intentional.
- Mono: vocal and kick still exist.
- Phone speaker: lyrics readable.
- Reference A/B at matched loudness, not raw volume.
- Filename includes version, e.g. Title_master_v03.wav.
Do not master a file that still needs a vocal-up mix revision.
How to master in MegaMix AI after the mix
Use /app for mixing and mastering in the browser, then after the mix feels locked, run the mastering pass on that bounce rather than uploading a crushed MP3. Refine with Josh if the top end is spiky or the low end is heavy on a phone, then export WAV for distribution. Use a free trial to learn your loudness taste against real references, or see plans.
A solid Josh note at mastering: keep punch on the kick, reduce 4 kHz harshness 1 dB, do not widen the sub. That is a mastering request. “Make it as loud as a club track” is how transients die.
You still do not need a DAW for this stage if the mix print is clean. Keep MegaSFX transitions in the mix stage, not as a surprise layer during mastering. For how MegaMix AI mixes then masters in one session, see the Mix AI hub.
FAQ
Should I master MP3 or WAV?
Master from the highest-quality mix export, typically WAV at the native sample rate. Encode MP3 or AAC after mastering for sharing and for distributors that want a lossy extra, not as the master source.
How loud should my master be?
Prioritize sound quality over a single LUFS number. Use integrated loudness as a guide, but trust translation tests more: if the snare disappears after you match loudness to a reference, you went too far.
Can AI mastering replace a human mastering engineer?
It can produce solid commercial loudness and tonal balance for many singles. A sequenced album with intentional quiet movements, or a vinyl cut, may still benefit from a human who hears the whole project in order.
What is the biggest mistake before mastering?
Handing off an over-limited mix with no headroom. Mastering then amplifies distortion and cannot restore punch you already squashed on the mix bus.
Do I need different masters for Spotify and YouTube?
Usually one strong master is enough because platforms normalize. Special cases are vinyl, DJ-focused extra-loud club versions, or a TV-safe mix with tighter peaks.
Should I dither the 24-bit master?
Dither when you reduce bit depth for a specific delivery, such as 16-bit CD. Keep a 24-bit WAV as your archive master and dither only the copy that needs it. More on the Mix AI hub. See Start free trial or plans.