Mixing vs Mastering: What Each One Does to Your Song
Mixing is what happens to the tracks. Mastering is what happens to the song. Everything else follows from that.
What mixing does
A mix starts with a pile of separate recordings: a lead vocal, doubles, ad-libs, a beat that might be one stereo file or thirty stems. Mixing decides how those pieces sit against each other. Which one is in front. How loud the kick is against the 808. Whether the vocal is dry and close or sitting back in a room. Where the ad-libs live in the stereo field.
At 24K the mix starts with the vocal, because that is what people hear first. Gain staging comes first, then EQ by hand on the take itself, then a chain built for that voice: subtractive EQ to take out what is not needed, compression that keeps the performance instead of flattening it, saturation for weight, and space last. Reverb and delay go on at the end, not the start, because you cannot judge how much room a vocal needs until it sits right dry.
Every one of those moves is made on an individual track, with access to all the others. If the chorus is not lifting, the mix engineer can ride the vocal up, pull the hats back and let the 808 breathe for those eight bars. Nobody later in the chain has that access.
What you get back from a mix is a stereo file with the balance settled. It is not yet loud, and it is not meant to be.
What mastering does
Mastering works on that one stereo file. There is no vocal fader any more. There is the whole song, and a short list of moves that apply to all of it at once: overall tone, overall dynamics, level for streaming, and the file spec the release needs.
Because it works on the whole, mastering is a narrower job than people expect. It can make a song brighter, tighter, louder, more consistent from verse to chorus. It cannot turn the vocal down two dB without turning the hi-hat down two dB.
The way it is done here: the mix is listened to on real systems before anything touches the master bus. The monitors in the control room, then a car, headphones and a phone speaker. The translation check comes before the processing, not after, because the whole question in mastering is whether the song holds up somewhere other than the room it was mixed in. Then it is levelled for streaming. The analogue gear in the room includes an SSL Fusion, a Wes Audio ngBusComp and a Bettermaker Mastering Limiter 2.0. Gear like that belongs to the last stage. It does not rescue the one before it.
What you get back from mastering is the finished file, at the right level, that sounds the same on a phone as it did on the monitors.
Why mastering problems are usually mix problems
This is the part that saves people weeks.
Most of what surfaces in mastering is a mix problem. A vocal two dB too forward. A low end that only exists on one system, huge on the Yorkville subwoofer and gone on a phone. A chorus that stops lifting the second time round. Each of those shows up the moment the song plays on the second system, and each of them is a mix decision, not a mastering one.
A mastering engineer can push against them. Cut where the low end is piling up, clamp the vocal peaks with multiband compression, squeeze the chorus. But each of those moves lands on the whole song. Pull the build-up at 60 Hz and you thin out the kick that was fine. Compress the vocal peaks and the snare loses its snap. A mix problem fixed at the mastering stage always takes something from somewhere else.
So at 24K, when something surfaces in mastering, it gets fixed in the mix. That is why mixing and mastering here comes back as two files: the master and the corrected mix. You are not getting a master that hides a problem. You are getting a mix with the problem removed, and a master of that.
The same logic is why it is better to send stems than a bounce. Whatever is baked into a stereo file stays baked in. Export as WAV, 16-bit or 24-bit, one file per track, at the session sample rate, and the engineer can reach the thing that needs changing.
Why the order matters
Mix, then master. Not out of tradition, but because of what each stage can see.
Mixing runs with every track available. Mastering runs blind to them. So decisions that need track-level access have to be made while that access still exists. Once you master, you have committed to the balance. If you then hear that the vocal is buried, the route back is to reopen the mix, change it, and master again.
That is also why a mix should not be delivered loud. If the mix has already been limited to a competitive level, the mastering stage has nothing left to work with: the peaks are gone, the dynamics are flat, and any tonal move has to fight a limiter that has already closed. Leave headroom. Loudness is the last job, and it belongs to the last stage.
And there is a stage before both of them: recording. The mix can only shape what the microphone caught. A vocal tracked in an untreated room carries that room into the mix, and the mix carries it into the master. This is why the booth at 24K is isolated at the wall, with the insulation built into the structure and acoustic foam over the reflective surfaces on top, rather than foam glued to the walls. A take that does not carry the room does not need the room EQ’d out later. The rule runs the whole way down the chain: fix it at the earliest stage that can see it.
What you get back, and when
Mixing and mastering at 24K is priced per song, whether you recorded here or somewhere else, and paid in full at checkout. Turnaround is three days from the day the files land.
Before anything starts, an engineer listens to what you sent and says honestly whether it is workable. If the vocals were recorded too poorly to save, the studio says so and refunds in full rather than sending back something that will not hold up. Once work has begun, it is not refundable.
If you would rather sit in the room, there is an in-person option: a four-hour session with the engineer, one song, and you hear the decisions being made as they are made. Details are on the services page, or if you are still working out what your song needs, get in touch first.
Files go to info@24kstudios.ca. If you recorded at 24K, the engineer already has the session.
Questions people ask
Is mastering just making a song louder?
No. Loudness is one part of mastering, but the stage is mainly about how the whole song translates across systems: overall tone, overall dynamics and a level suited to streaming. A master cannot change the balance between individual tracks.
Can mastering fix a vocal that is too loud in the mix?
Not without side effects. Mastering works on the whole stereo file, so any move that pulls the vocal down also pulls down everything sharing that frequency range. At 24K a problem like that is fixed in the mix, and you get back both the master and the corrected mix.
What do I get back from mixing and mastering at 24K Studios?
Two files: the finished master, levelled for streaming, and the corrected mix it was made from. Turnaround is three days from the day the files land, and the work is priced per song whether you recorded at 24K or elsewhere.
What format should I send stems in for mixing?
WAV, 16-bit or 24-bit, one file per track, at the session sample rate. Send them to info@24kstudios.ca. An engineer listens first and says whether the files are workable before any work starts.
Can I sit in on my mix and master?
Yes. 24K offers a four-hour in-person session with the engineer for one song, in the room, so you hear every decision as it is made.
24K Studios is at 550 Queen St E, Toronto, open 9 AM to 2 AM every day. Music · Voice & corporate · Gear · Book studio time