The short answer: If your song sounds unbalanced, cluttered, or the vocal fights the beat, you need mixing. If your song already sounds right on your headphones, your car, and your friend's phone — and it's just quiet, dull, or slightly thin next to commercial releases — you need mastering. Nine times out of ten, an independent artist asking "do I need mixing or mastering" actually needs mixing, and mastering is the last 5% that comes after. Buying mastering to fix a mix is the single most common way people waste money on their music.

Diagnose it from what you're hearing

Forget definitions for a second. Play your rough on three sources — good headphones, a phone speaker, and a car — and check which of these describes your problem.

You need mixing if:

  • The vocal disappears under the beat in the chorus, or you have to strain to catch words.
  • The low end is a blur — you can't tell the kick from the 808, and the bass is huge in one room and gone in another.
  • It sounds harsh or fatiguing after 30 seconds, especially in the 2–5 kHz range where sibilance and snare crack live.
  • Ad-libs, doubles, and background stacks feel like separate events instead of one voice.
  • The song sounds "small" — everything is in the middle, no width, no depth.
  • The arrangement feels flat — no lift into the hook, no drop into the verse.
  • Your phone speaker gives you a completely different song than your headphones.

You need mastering if:

  • The balance is genuinely right everywhere, and your only complaint is volume — it's noticeably quieter than reference tracks in a playlist.
  • It's a touch dull or a touch dark compared to the records you're competing with, but nothing is wrong.
  • You have a project — an EP, an album, a mixtape — where the songs were mixed at different times and don't sit together in level or tone.
  • You need deliverables: streaming-ready files, a track order with gaps, ISRCs embedded, a separate louder version for a club or radio drop.

That's the whole diagnostic. Mixing decides what the record is. Mastering decides how it travels.

What happens if you buy the wrong one

This part matters because the mistake is expensive in both directions.

Buying mastering when you needed mixing. A mastering engineer works on your stereo file. They cannot pull the vocal up 2 dB in the second verse. They cannot separate the kick from the 808 because both live in the same 40–80 Hz region of the same file. What they can do is make the problem louder and more obvious. Buried vocal, mastered: still buried, now louder and more compressed, and often harsher, because the limiter grabs whatever is loudest — usually the exact thing that was already too loud. You will get the file back, feel like something is still off, and not know why. That's the trap.

Buying mixing when you needed mastering. Less harmful, but you're paying for a full rebuild of something already working, and you risk losing a vibe you'd already dialled in. If your rough genuinely translates everywhere, say so up front and ask for the smaller service.

The myth worth killing: mastering is not a "make it sound professional" button. It is the final quality-control and translation stage — typically a few dB of gain, a couple of gentle EQ moves under 1.5 dB, some glue, and correct loudness and true-peak targets for delivery. If an engineer's master transforms your song, that's a sign the mix left too much on the table, not a sign the mastering was great.

The 2026 loudness reality — why "make it loud" is the wrong ask

Spotify, Apple Music, YouTube and Amazon all normalize playback. Spotify targets around -14 LUFS integrated by default; Apple Music sits near -16 with Sound Check. Crushing your master to -6 LUFS doesn't make it louder to a listener — the platform just turns it down, and you're left with the squashed transients and dead drums you traded away for nothing.

What actually survives normalization is density and clarity: a controlled low end, a vocal that stays present when the level drops, transients that still punch. Those are mix decisions. This is exactly why "I need mastering, my song isn't loud enough" is usually a mixing conversation in disguise. In R&B and afrobeats especially, where the groove depends on space between elements, over-limiting flattens the thing people came for.

Practical target: aim your master somewhere in the -9 to -12 LUFS integrated range for hip-hop and R&B, true peak at -1 dBTP, and let the platforms do their thing. Genre and arrangement move that number — a sparse ballad and a wall-of-808s trap record don't land in the same place.

Order of operations, and where production sits

Mixing comes first. Always. Mastering is the last thing that happens to a finished, approved mix. If you're wondering mixing or mastering first — there is no version where mastering comes first. Not one.

Where it gets genuinely ambiguous is production versus mixing. Some songs don't have a mix problem; they have an arrangement problem. If your hook doesn't lift, if the second verse is identical to the first, if the 808 pattern fights the vocal cadence — no amount of EQ fixes that. That's production: rewriting a part, re-cutting a vocal, adding or removing elements, re-tuning. It's worth being honest about this before you pay for anything, and it's a five-minute conversation on a call rather than a two-week discovery after you've already booked the wrong service.

Quick sanity check on scope:

  1. Does the song work when someone plays it acoustically or on a phone with no processing? If no → production.
  2. Does the balance hold up across three playback systems? If no → mixing.
  3. Is the balance right but it feels quiet, dull, or inconsistent across a project? If yes → mastering.

What actually drives the cost of each

You'll see a wide spread in the market, from bedroom offers to major-label rates, and the spread is real because the work is genuinely different. What moves the number:

  • Track count and session hygiene. Sixteen well-labelled, consolidated stems is a different job than 94 unlabelled tracks with printed effects and clipped exports.
  • Revisions. Two rounds versus open-ended.
  • What you're delivering. One single versus a ten-track project needing cohesion across all of it.
  • Extra deliverables. Instrumentals, TV mixes, acappellas, clean edits, Dolby Atmos, DJ/club versions.
  • Vocal work included. Tuning, comping, timing, de-essing, stacking — often lumped into "mixing" but real hours.
  • Whether the mix needs rescuing. Distorted source recordings, room ring, mic bleed.

Mastering is almost always the smaller of the two, because it's the shorter job. That's not a discount; it reflects the scope.

Working with Felix on this

Send the rough. That's the whole first step. Felix works with independent artists and small labels out of Toronto — in person if you're local, fully remote if you're not — mostly across R&B, hip-hop, and afrobeats, so the reference points for how these records should hit are already in the room.

What happens: he listens to what you have, tells you straight whether you're looking at a mastering job, a full mix, or a production conversation — including when the honest answer is "this doesn't need what you were about to buy." From there you get scope, timeline, and a real number for your specific song, because pricing depends entirely on the things listed above and any quoted range would be a guess about your project.

If you already know which one you need, great — say so and skip ahead. If you don't, that's exactly what the call is for. Book it through the contact page, send the rough with it, and you'll have a clear answer on which service your record actually needs before you spend anything.