Short answer: mastering is the final polish on a finished stereo mix, a light and careful pass that makes the track translate everywhere and land at a sensible loudness. The usual chain is gentle and in order: corrective EQ first, then a touch of compression to glue it, optional saturation or width, and a limiter last to set the level. Moves are small, a dB or two, and none of it can rescue a mix that is not already good.
Mastering is the last step before a track goes out, working on the single stereo file rather than the parts inside it. The settings here are starting points, not fixed rules, and every track wants a slightly different touch.
Mastering is the final polish on a finished stereo mix, the step that makes it translate on every system and sit at a sensible loudness. It works on the single stereo file, the two-channel bounce of your whole song, not the separate tracks inside it. By this point the mixing is done, and mastering is the light pass over the top of it.
It is doing two jobs. The first is translation: making sure the track holds together on a phone speaker, cheap earbuds, a car, a laptop and a big system alike, rather than only sounding right where you made it. The second is level: bringing the track to a loudness that fits alongside other releases without clipping or sounding weak. Both are gentle, corrective moves, not a rebuild.
What mastering is not is a rescue. It cannot un-muddy a cloudy mix, un-harsh a brittle one or pull apart parts that fight each other, because it only has the finished stereo file to work with and can no longer touch the pieces. If something is wrong in the balance, the fix is back in the mix, not here. Getting the mix clear first, see why does my mix sound muddy, does more for the final sound than any mastering move.
Start from the best mix you can make. Mastering works on the finished stereo file, so it can nudge the overall tone and loudness but it cannot separate parts, lift a buried vocal or undo mud. Those all live in the mix. If a master is not coming together, the honest answer is almost always to go back and fix the mix, then master again.
Corrective EQ first, then gentle compression to glue the mix, then optional saturation or width, and a limiter last to set the loudness. Each stage does only a little, and you can skip any middle step the track does not need.
The order matters because each stage feeds the next. You balance the tone before you compress it, glue before you colour it, and limit last so the final level sits on top of a mix that is already even. Here is the usual chain.
| Stage | What it does | How much |
|---|---|---|
| Corrective EQ | Balances the overall tone. Trims a boomy or harsh spot, adds a little air if it needs it. | Small cuts and lifts, roughly a dB or two. |
| Compression (glue) | Gently pulls the mix together so it moves as one piece. | Light, around a dB or two of gain reduction at most. |
| Saturation or width (optional) | Adds warmth and a sense of loudness, or opens the stereo image a touch. | Subtle, and only if the mix asks for it. Easy to overdo. |
| Limiter (last) | Raises the whole track to its final level and catches the last peaks. Sets loudness and the true peak ceiling. | Set to your loudness target, true peak at or below -1 dBTP. |
Treat this as the standard shape, not a fixed law. Some tracks want nothing more than a limiter and a small EQ move, others take all four stages. The rule that holds across all of them is restraint: gentle moves, a dB or two at a time, checked by ear against the unprocessed mix so you can hear whether each step is actually helping.
For most streaming, aim for around -14 LUFS integrated with your true peak at or below -1 dBTP, and only push louder when a genre genuinely calls for it. You set both at the limiter, so loudness is the one number you dial in on purpose at the end.
The reason to stop at the target is that streaming services play every track at a set loudness and turn louder masters down to reach it. Crushing a master past the target does not make it play louder than the track before it, it just trades away the dynamics that make it feel alive. Leaving the true peak at or below -1 dBTP keeps a little headroom so the file does not clip when it is encoded for streaming.
The exact numbers move over time and differ by platform, so treat them as a target zone rather than a fixed figure. For the per-service targets and why louder does not win, see how loud should my master be.
When the track translates on several different systems, holds up next to a reference you trust, and every further move stops making it better. Knowing when to stop is the real skill here, and it is easy to keep going long past the point of help.
Check on more than one system. Studio monitors, headphones, a laptop, a phone and a car all tell you something different. A master that sounds right on all of them is finished; one that only works on your monitors is not. The small, honest speakers catch problems the big ones flatter.
Compare against a reference. Pick a released track you like in a similar style, match it to a similar loudness so the louder one does not simply seem better, and switch back and forth. It keeps your tone and your level honest and stops you drifting.
Rest your ears and stop early. Take breaks, and come back the next day if you can. Overprocessing creeps in when you keep nudging a master that is already done, so if you cannot clearly hear an improvement, the move is not needed.
You do not have to master it yourself. A mastering engineer brings fresh ears and a treated room, and an automated online service will get a track to a sensible loudness quickly and cheaply. Either is a fair choice, especially for a release that matters or when your own ears are too close to the song to judge it.
We do not make a mastering tool, and nothing on this page needs one. What we make is instruments. The Collection is our three together, ARGISH, SILT and REHEAT, with three separate licence keys: ARGISH is a self-playing drone synth, SILT is a tape-loop instrument, and REHEAT writes acid lines. They give you clean, well-defined source sounds, so the mix that reaches mastering is already clear and easy to finish, and the final polish has less to correct. Every one ships as an Audio Unit, a VST3, an AAX for Pro Tools and a standalone app on macOS, signed and notarized, and a VST3 on Windows.
You can hear one free in your browser first, no install and no account, then decide.
We write these when there is something worth writing down. One email when a new one lands or a new Tunary instrument ships. No newsletter, no schedule.