The Mixing Guide
You know what the effects do. Mixing is not the same skill. Mixing is deciding how loud everything is, where it sits, and what gets out of the way — and the overwhelming majority of that is done with two controls that aren't plugins at all: the fader and the pan pot. Reach for a compressor because a vocal feels buried, and you'll spend an hour fixing a problem that was three decibels of fader. This guide starts where a mix actually starts, and only picks up a plugin once it has run out of simpler answers.
What Mixing Actually Is
A mix is a hierarchy of attention. At every moment in a song, something is the most important thing — the vocal, the riff, the drop — and everything else exists to support it without competing. Mixing is the craft of making that hierarchy audible. It is not the craft of putting a plugin on every channel.
This is why the order of operations matters so much. Almost every problem a beginner tries to solve with processing is really one of three simpler problems, and they should be checked in this order:
1. Is it too loud or too quiet?
Then it's a fader problem. Fix it with the fader. This is most problems.
2. Is it in the same place as something else?
Then it's a panning problem, or a register problem. Move one of them.
3. Are two things fighting for the same frequencies?
Then it's an arrangement problem first, and an EQ problem only if the arrangement can't change.
4. Only now: processing.
EQ, compression, saturation. Powerful, and completely unable to rescue a mix whose balance is wrong.
Mixing and mastering are different jobs. Mixing is about the relationship between the tracks and happens with all of them in front of you. Mastering treats the finished stereo file as one object. If you're solving a balance problem on the master bus, you're mixing — go back and do it on the channel.
Gain Staging & Headroom
Gain staging is making sure every stage of your signal path receives a sensible level. Get it wrong and everything downstream gets harder for no reason: plugins behave unpredictably, the master bus clips before you've made a single creative decision, and you spend the whole mix pulling faders down.
The target is simple. Aim for each channel to peak somewhere around −12 to −6 dBFS, and your master to peak no higher than −6 dBFS with everything playing. That empty space at the top is headroom, and it is not wasted — it's the room the mastering stage needs to work in, and the room your mix bus needs so that adding one more element doesn't push the whole thing into the ceiling.
Two reasons this matters more than it sounds like it should. First, analog-modelled plugins have a sweet spot. A tape emulation or a modelled compressor is calibrated to expect a particular input level; feed it something 15 dB hotter and you get distortion you didn't ask for. Second, loud sounds better. Your ear reads any louder version of the same thing as an improvement, so if a channel is simply running hotter than the others, you will mistake it for the better-sounding one and build a whole mix around a lie.
One thing you don't have to worry about: modern DAWs mix internally in 32-bit float, so a channel going a little over zero won't actually destroy the audio the way it would have on tape or on a 16-bit console. But your plugins still care, your master bus still clips on output, and your judgement still gets fooled by loudness. Stage the gain anyway.
The Static Mix — Faders First
Here is the single most valuable habit in mixing, and the one most beginners never learn: before you insert a single plugin, get the song sounding as good as you can using nothing but the faders. No EQ, no compression, no reverb. Just relative volume. This is called the static mix or the rough balance, and a good one is startling — most of the way to a finished record.
Do it like this. Pull every fader to the bottom. Bring up whatever the song is built on — usually the drums, or the main instrument if it's a song that lives on one. Then bring up the most important element, which in most popular music is the lead vocal, and set it where it feels right against that foundation. Then add everything else, one at a time, always asking one question: does this need to be louder than that, or quieter? Anything you can't find a level for is a candidate for deletion.
| Move | Why it works | Watch for |
|---|---|---|
| Start at silence | Building up beats pulling down — you decide what earns its place | Faders creeping upward all session; if everything's up, nothing is loud |
| Foundation first | Everything else is judged relative to the rhythm section | Mixing the vocal against nothing, then rebuilding around it |
| Lead element second | It's the thing people came for; it sets the ceiling for everything else | Burying it, then compressing to "bring it out" |
| Mix at low volume | Quiet playback flatters nothing — the balance has to work on its own | Loud playback making a bad balance feel exciting |
| Mix in mono, at first | Removes stereo width as a crutch and exposes masking immediately | Parts that only "work" when spread apart |
That last one is worth its own sentence. Building the static mix in mono is the fastest way to find out which parts are actually fighting. Stereo hides collisions: two sounds occupying identical frequencies can seem fine simply because one is left and one is right. Collapse to mono and the conflict is undeniable — and since a great many listeners hear music on a phone speaker, a single laptop speaker, or in a club with mono bass, that conflict was going to reach them anyway.
Panning & the Stereo Field
If the fader decides how loud something is, panning decides where it is — and two sounds in different places can both be heard clearly at volumes where, stacked in the centre, they'd be mud. Panning is free separation. It costs nothing, damages nothing, and most beginners barely use it.
The centre is expensive real estate. Only a few things can occupy it before it congests, and they're the things that carry the song's weight:
Centre
Kick, bass, snare, lead vocal. The spine of the record. These are the loudest elements and the ones the ear expects to come from in front of you. Everything below ~120 Hz belongs here regardless.
Wide (hard left / hard right)
Doubled guitars, pads, stereo keys, room mics. Big, sustained, textural things. Hard panning is bolder than it sounds and almost always better than the timid 20% that leaves something neither centred nor separated.
In between
Toms, percussion, backing vocals, counter-melodies. Placed to fill the space between the spine and the edges — and to keep two similar parts from landing on top of each other.
The mirror trick
Two parts of similar character? Pan them to opposite sides at equal amounts. The mix stays balanced, and both become individually audible. This is why doubled guitars are tracked twice and panned hard rather than copied and pasted.
Some producers work strictly LCR — every element is either dead centre, hard left, or hard right, with nothing in between. It sounds like a harsh restriction and it produces remarkably clear mixes, because it forces a decision instead of allowing everything to hover vaguely near the middle. Try it once as an exercise even if you don't keep it.
The Low End
More amateur mixes are undone below 200 Hz than everywhere else combined. The low end holds most of a mix's energy — a small level change down there moves the whole track's loudness — and it's simultaneously the region you can least trust in an untreated room or on small speakers. It rewards rules over ears.
Rule one: decide who owns the sub. The kick and the bass both want the region below about 100 Hz, and if you let them both have it they'll sum into a formless boom that eats your headroom and pumps unpredictably. Pick one.
| Approach | How | Suits |
|---|---|---|
| Kick owns the sub | Bass gets a gentle high-pass around 50–70 Hz; kick keeps everything below | Hip-hop, electronic, metal — anything where the kick is the punch |
| Bass owns the sub | Kick is high-passed and lives on its 80–150 Hz thump and 2–4 kHz click | Rock, funk, soul — anything where the bassline is the hook |
| They trade | Sidechain the bass to duck a few dB each time the kick hits | Dance and modern pop, where both must feel huge |
| Carve | Cut on one exactly where you boost on the other (e.g. kick +3 @ 60, bass −3 @ 60) | When the arrangement can't change and both must stay full |
Rule two: keep the bottom mono. Below roughly 100–120 Hz, stereo information does you no favours — your ears can barely localise it, vinyl cutting hates it, and club systems sum it to mono anyway. Worse, wide low frequencies are exactly what cancels when a mix is folded down. Collapse everything under that point to the centre and the low end gets tighter and louder for free.
Rule three: high-pass what doesn't need it — but not reflexively. A vocal, a hi-hat, a shaker and an acoustic guitar all carry rumble and proximity build-up they contribute nothing with. Rolling it off cleans up enormous amounts of room. But high-passing everything at 100 Hz by habit is how mixes end up thin: many instruments carry real body between 80 and 200 Hz. Sweep the filter up until you hear the sound get smaller, then back off.
Subtraction — The Arrangement Is the Mix
When a mix sounds cluttered, the instinct is to EQ. Usually the real diagnosis is simpler and less flattering: too many things are playing at once, in the same register. No amount of EQ separates two parts that are genuinely occupying the same frequencies at the same time — it only makes both of them worse. The fix isn't a plugin. It's arrangement.
Three ways to solve it upstream, all better than EQ:
Mute it
The pad and the rhythm guitar are both playing sustained mid-range chords. You do not need both. Delete one and the mix opens up instantly — and the song usually gets better, not worse.
Move it
Same part, different octave, or a different voicing. Two instruments an octave apart stop competing entirely. This is what voicing is for.
Stagger it
Have them play in the gaps rather than simultaneously. Call and response instead of two parts shouting over each other. Costs nothing, sounds arranged rather than mixed.
There's a useful discipline here: when a section feels crowded, try removing something before adding anything. Records that sound big are usually records where few things happen at once — the space is what makes the loud parts land. If your chorus doesn't feel like an arrival, the problem may be that your verse is already too full for the chorus to be bigger than.
Practical version of this rule for a stuck mix: solo the two elements you think are fighting and listen to just those two. If they sound bad together on their own, no processing will fix it in context — change what one of them plays.
Automation — A Mix That Moves
A static mix is a compromise: one fader position that has to work for the whole song. But the verse and the chorus are different pieces of music with different densities, and the level that suits one won't suit the other. Automation is how a mix stops being an average and starts being right at every moment.
This is also the honest alternative to over-compression. If a vocal disappears in the loud sections and jumps out in the quiet ones, you can flatten it with 8 dB of gain reduction and lose all its life — or you can draw the fader up where it's buried and down where it's too much, and keep the dynamics. Engineers call this riding the fader, and it's what compression is a crude automatic imitation of.
| Automate | To do what |
|---|---|
| Volume | Ride the lead so every word sits; lift a part for one bar where it matters; drop the pad under the vocal |
| Send level | More reverb on the last word of a phrase, on a breakdown, or on the section you want to feel vast |
| Filter cutoff | Open a filter across a build; strip the low end from a verse so the chorus's arrival feels enormous |
| Pan | Move a delay or a texture across the field for a moment of motion — sparingly, it draws attention |
How to Listen
You can know every technique on this page and still make bad mixes, because every decision you make is only as good as what you're hearing. Your room, your speakers, your volume and your fatigue level are all part of the signal chain, and unlike your plugins they're lying to you constantly.
The room is the biggest variable
Untreated rooms have standing waves that make some bass notes disappear and others boom, at specific frequencies, depending on where your head is. You will mix against those errors and bake the opposite problem into the file. Even a little absorption at the first reflection points and behind the speakers changes more than any monitor upgrade.
Headphones aren't a workaround, but they help
Headphones skip the room entirely, which is why they're excellent for detail, edits, noise and reverb tails — and misleading for stereo width and low end, because nothing is crossfeeding between your ears. Use both. If you only have headphones, make your width and bass decisions conservatively and check them somewhere else.
Pick one level and keep it
Loudness changes the frequency response of your ears — bass and treble seem to vanish as you turn down. Mix at a consistent, moderate level (conversation volume) and your judgements become comparable across days. Turn it up briefly to check excitement; turn it down to check balance.
Fatigue is real and fast
Your ears adapt within minutes and your judgement degrades within an hour. The three-hour mix is almost always worse than the forty-five-minute one. Take breaks, and treat the first ten seconds after returning as the most honest listening you'll get all day.
Reference tracks — the only real calibration
A reference track is a finished, professionally mixed record in a similar style, loaded into your session so you can A/B against it. It's the single most effective way to escape your room's lies, because it was made in a good room and it tells you what "correct" sounds like here, right now, on these speakers.
There is one rule that makes the difference between a reference track helping and actively harming: level-match it. A mastered record is far louder than your unmastered mix, and louder always sounds better — so if you A/B without matching, your mix will lose every time for reasons that have nothing to do with your mixing. Pull the reference down until it's subjectively the same loudness as your mix, then compare. Suddenly you're hearing balance and tone instead of volume.
The same rule applies to your own plugins. Before you decide a compressor "improved" a channel, match the bypassed and engaged levels and switch again. A great deal of what people believe processing does for them is just the 2 dB it added.
Check it everywhere
Before you call a mix done, play it on a phone speaker, in a car, on cheap earbuds, and in another room while you're not looking at the screen. Each one exposes something the others hide — the phone tells you whether the vocal and the snare survive without any bass at all, the car tells you about the low end, and the other room tells you about the balance, because distance and walls strip away everything except what's actually loud.
Get the balance right with faders, place things with the pan pot, decide who owns the low end, and take something away before you add anything. Do that on a mix you can trust what you're hearing on, and the processing you have left to do will be small. Then — and only then — it's ready to master.