Where mastering sits in the process
Mixing and mastering also tend to ask for different mindsets, not only different tools — mixing is detail-focused, chasing exactly how one part sits against another, while mastering is deliberately zoomed out, judging the track as a single, finished object rather than a collection of parts.
Mixing works with the individual tracks of a song — vocals, guitars, drums and the rest — balancing them against each other. Mastering starts only once that mix is finished and bounced down to a single stereo file; a mastering pass works with that one file, not the individual tracks inside it, which makes it a structurally different job even though both stages lean on EQ, compression and level.
That structural difference shows up practically, too: a mixing engineer can solo an individual guitar track and adjust it in isolation, but a mastering engineer generally cannot — everything reaching them has already been permanently blended into one stereo signal, which is exactly why decisions made at the mixing stage aren’t recoverable later. Mastering works with what mixing already decided, not around it.
Loudness consistency: one track next to another
A big part of mastering is making sure a whole release — an EP, an album, a run of singles meant to sit on one playlist — lands at a broadly consistent perceived loudness, so a listener isn’t reaching for the volume knob between tracks. Individually great mixes can still land at noticeably different loudness levels on their own; mastering is the pass that lines them back up against each other.
This matters even for a single-song release, not only a multi-track album — a track has to sit at a competitive, appropriate loudness next to whatever a listener plays immediately before or after it on a shared playlist, and mastering is the stage where that comparison is actually made and adjusted for, track against track, rather than judged in isolation.
Genres differ here too — a pop single and an ambient instrumental piece are reasonably mastered toward different target loudness ranges entirely, since “consistent” means consistent within the context a track will actually be heard in, not consistent against some single universal number.
Streaming platforms complicate this further by applying their own loudness normalisation on top of whatever a release ships at, which is one more reason mastering toward a sensible, consistent target across a release matters more than chasing the single loudest possible number — the platform will often flatten an overly loud master back down regardless.
Translation: how a mix holds up everywhere
A mix built and judged in one room, on one set of speakers, can reveal problems the moment it’s played somewhere else entirely — a phone speaker, cheap earbuds, a car stereo, a club system. Mastering deliberately checks for and adjusts that spread, aiming for a mix that holds together reasonably well across very different playback situations, rather than sounding perfect in exactly one room and falling apart everywhere else.
A mastering engineer typically checks a mix across several deliberately different playback references during this stage — a pair of studio monitors, a consumer-grade speaker, headphones, sometimes a car — precisely because a mix that only gets checked once, on one system, tends to reveal its blind spots only after release, when it’s too late to quietly fix.
Fresh ears and the final polish
After hours or days inside the same mix, a mixer’s ears adapt and gradually stop hearing it accurately — a well-known hazard of any long session. Mastering is often handed to a different set of ears entirely, or at minimum returned to after real distance, precisely because a listener encountering the track fresh catches things its own creator has stopped noticing. What actually happens at this stage is usually subtle: small equalisation moves, gentle overall compression or limiting, and careful sequencing and spacing between tracks — a polish pass, not a rebuild.
That fresh-ears effect is strong enough that even a mixer working entirely alone benefits from deliberately building distance into the process — finishing a mix, leaving it for a day or more, and returning to master it only once the ear has genuinely reset, rather than pushing straight from mixing into mastering in one unbroken sitting.
What a streaming platform’s loudness normalisation actually does, mechanically
The “Loudness consistency” section above notes that streaming platforms apply their “own loudness normalisation on top of whatever a release ships at” — worth naming the actual mechanism, since it’s more specific than a vague adjustment. Spotify’s own documentation states it plainly: “We adjust tracks to -14 dB LUFS, according to the ITU 1770 standard.” A master that comes in louder than that target gets turned down — “Negative gain is applied to louder masters so the loudness level is -14 dB LUFS” — and a quieter one gets turned up to match, with Spotify’s own explanation adding that it “consider[s] the headroom of the track, and leave[s] 1 dB headroom for lossy encodings to preserve audio quality.”
That’s a real, specific number sitting behind the general claim already made above — not a universal one across every platform, since each service sets and can change its own target, but a concrete illustration of exactly what “the platform will flatten an overly loud master back down” means in practice on one major service.
Why mastering unusually loud can backfire even inside that system
Spotify’s own documentation gives a specific case worth passing on because it sharpens the “What mastering can’t fix” section’s spirit into a numbers problem rather than just a philosophical one: a track measured at -20 dB LUFS with a True Peak of -5 dB FS doesn’t get boosted the full distance up to -14 — it’s capped at -16 dB LUFS instead, specifically to avoid distortion once the file is encoded down to a lossy streaming format.
The practical upshot lines up exactly with what this page already argues about fresh ears and restraint: pushing a master’s loudness right up against the ceiling doesn’t guarantee it reaches a target loudness on a platform that normalises anyway — it can just as easily trade away headroom the platform’s own encoding safety margin was counting on, for a loudness gain that quietly gets capped or reversed the moment the track is streamed.
What mastering can’t fix
Mastering cannot rescue a mix with a genuinely unbalanced vocal, a muddy low end from clashing instruments, or an arrangement fighting itself — those are mix-stage and arrangement-stage problems, and pushing them downstream on the assumption they’ll get “fixed in mastering” almost never works. Mastering polishes and evens out a mix that is already fundamentally right; it isn’t a repair stage for one that isn’t.
This is also why “send it to mastering” is not a rescue plan for a mix a producer isn’t confident in — a mastering engineer receiving a genuinely unbalanced mix can only apply the same broad, whole-mix tools everyone else has (overall EQ, overall compression, overall level), none of which can selectively fix one clashing instrument without affecting everything else in the mix at the same time. The fix for that has to happen earlier, back at the mix, where individual tracks are still separately adjustable.
Where this comes from
Spotify for Artists, “Loudness normalization on Spotify”. support.spotify.com · accessed 11 August 2026.
Common questions
Can I master my own music?
Yes, plenty of home producers do. The harder part is getting the fresh-ears perspective a separate mastering pass depends on, so stepping away from a mix for a day or two before mastering it helps.
Does mastering make a track louder?
Often, yes, as part of getting it to a consistent loudness next to other tracks — but loudness is one piece of mastering, not the whole job, and chasing loudness alone can do more harm than good.
What’s the one-line difference between mixing and mastering?
Mixing balances the separate tracks inside one song; mastering polishes and levels the finished stereo mix, often alongside the other tracks it will sit next to.