Mixing vs mastering: which one does your track actually need?

What mixing does

Mixing works with your session open. It takes the separate stems — kick, bass, pads, leads, vocals, percussion — and decides how loud each one is, what part of the frequency spectrum it owns, how much of the stereo field it occupies and how it behaves dynamically against everything else. The output is one stereo file that sounds like a record instead of a stack of correct parts.

• A bass that disappears on laptop speakers and swamps a club system: mixing.
• A vocal that fights the lead synth for the same 3 kHz: mixing.
• A drop that feels smaller than the build-up before it: mixing.
• A kick that never locks with the bass: mixing.

What mastering does

Mastering starts where mixing ends: with a finished stereo file. It sets the final tonal balance, controls the dynamics as a whole, decides the loudness the record will live at, and delivers the formats each destination needs — a 16-bit WAV for stores, a higher-headroom file for streaming normalisation, sometimes a separate club version. It is measured in decisions, not in plugins.

What to deliver, and the pre-master level

1. For mastering: one stereo WAV, 24-bit, at the session sample rate (44.1 or 48 kHz), with no limiter on the master bus.
2. Leave around 3 to 6 dB of headroom — peaks landing near -4 dBFS. A file already squashed to 0 dB leaves the mastering engineer nothing to work with.
3. For mixing: every stem exported from bar one to the end of the track, same length, same sample rate, dry unless the effect is part of the sound design.
4. Name the files. "Audio 04.wav" costs someone an hour of guessing.

You can master a track you mixed yourself, and plenty of good records are made that way. What you cannot do is master your way out of a mix that does not work — the loudness will go up and every problem will go up with it.

See mixing & mastering at Epical Tune→