Mixing and mastering

Mastering is the last step before release: the final balance of tone and loudness that makes a song sound finished next to other songs on a playlist. AI mastering services take a stereo mix, analyze it and apply that last stage automatically, usually in a few minutes and with a choice of styles or loudness levels.

For an independent band they solve a practical problem. You have a stack of demos, a single you want out next month and no budget to book an engineer for every track. An online master from a service such as LANDR, eMastered or Mixea gives you something release-ready quickly, and you can compare a few versions before you choose.

The catch is that mastering can only work with what the mix gives it. If the vocal is buried, the kick and bass fight each other or the cymbals are harsh, an automatic master will make those problems louder. Spend your effort on the mix first, leave a little headroom, and then master. A simple test is to compare your master with a commercial reference track in the same style at matched loudness.

Assistant plugins work earlier in the chain. Tools such as sonible's smart plugins and Soundtheory's Gullfoss listen to a track and suggest or apply EQ and dynamics settings, which you then adjust. They help a band member who records the band at home and wants a sensible starting point. The open-source Matchering takes a different route: give it your track and a reference song you like, and it matches loudness and tonal balance to that reference.

6 AI music tools in mixing and mastering

Matchering

by Matchering contributors

Open-source tool that masters a track to match the sound of a reference song you choose.

  • Free plan or free version
  • Open source

LANDR

by LANDR Audio

Online AI mastering with style and loudness choices, plus distribution, samples and plugins.

  • Works in the browser

eMastered

Online AI mastering where you upload a mix and adjust the result before downloading.

  • Works in the browser

Mixea

by DistroKid

DistroKid's online mastering tool for preparing tracks before release.

  • Works in the browser

sonible smart plugins

by sonible

AI-assisted mixing plugins such as smart:EQ and smart:comp that suggest settings from your audio.

  • Desktop app or plugin

Gullfoss

by Soundtheory

Automatic EQ plugin that adjusts tone in real time to make a mix clearer.

  • Desktop app or plugin

Keep your unmastered mixes and the session files. If a song finds an audience, you may want a person to master it again for vinyl or for a label, and they will ask for the original mix.

Loudness is the setting people most often get wrong. Streaming services adjust playback volume, so a master pushed as loud as possible gains little and can lose punch. Most services let you pick a gentler loudness, and it is worth trying both. Listen to the quiet intro and the loudest chorus of each version, because heavy limiting shows up there first.

If two band members disagree about a master, export both versions, take a break and listen again the next day on the speakers you know best.

Questions about mixing and mastering

Is AI mastering good enough to release?

For many independent singles and demos, yes, if the mix is already balanced. Listen on headphones, in a car and on a phone, and compare against a commercial song in the same style at the same loudness before you decide.

Can an AI mastering service fix a bad mix?

No. Mastering works on the whole stereo file, so a buried vocal or muddy low end will still be there, often louder. Fix balance and tone problems in the mix first.

What is reference matching?

You give the tool your track and a finished song you like, and it adjusts your track's loudness and tonal balance toward the reference. Matchering is an open-source tool that works this way.

Should we keep the unmastered files?

Yes. Keep the final mix before mastering and the full session. A label, a vinyl cutting engineer or a human mastering engineer will want them later.

Related categories