~/music-production/foundations mixing
Foundations

Mixing

How the finished parts are heard: level, pan, and tone.

Mixing is the step after the parts exist. It does not change what any part plays; it decides how the finished parts are heard together. This reading steps back from any one DAW to name the moves a mix is built from, then takes up the three you reach for first: level, pan, and tone.

Up to now your work has been creation: choosing sounds, playing parts, arranging them into a piece. Creation decides what happens in the music. Mixing decides how the finished parts are heard, and it does not change what any part plays. It sets how loud each part is against the others (level), where each sits from left to right (pan), how near or far it feels (space, mostly from reverb), how parts share the frequency range so they do not mask each other (tone, from EQ), and how steady each part holds its level (dynamics, from compression).

PAN AND DEPTH FAR back · reverb NEAR front · dry L R
Pan sets a sound left to right; reverb sets its depth near to far. Every part gets a place in this field.
THE MIX AS A PIZZA TOPPINGS on top: vocals, solos, fills CHEESE granular middle: rhythmic guitar SAUCE the glue in the middle: pads DOUGH the sustaining base: drums, bass
A mix stacks like a pizza: a base that sustains, a middle that glues and fills, and the bright things on top.

Dynamics, how steady each part holds its level, has its own reading. The three this one takes up are the moves a mix starts from: level, pan, and tone, each in turn.

1 · Track levels

The volume of one track against the others. The first thing a mix balances. Too loud and a track overwhelms everything else; too quiet and you don't hear it at all. Most mixing decisions start here.

Every DAW puts this on the channel as the track's volume fader (sometimes labeled gain). Pull it down to make the part quieter, push it up to make it louder.

Gain staging: the two-part habit

Good levels come down to two things, in this order:

  1. Nothing peaks. No track should hit 0 dB on the meters. When a signal reaches 0, the waveform clips flat at the top and the audio distorts. Once a track has clipped, the only fix is to lower its level and play it back again. Keep peaks comfortably below 0, with the loudest moments sitting around −6 dB. That gap between your loudest peak and 0 is headroom.
  2. Tracks are balanced against each other. Once nothing is clipping, the question is whether every element sits at the right level relative to the rest. The main sound of a moment should be loudest, supporting sounds quieter. As the piece moves through sections, the balance can shift: the foreground in one section can step back in the next.
Gain staging: two contrasting mixes shown as level meters. Healthy mix has all tracks peaking around -7 to -9 dB, all in the green zone. Problem mix has one clipping track, two underused but balanced tracks, and one creeping into the orange zone. HEALTHY MIX NEEDS WORK 0 -3 -6 -12 -18 -24 -36 CLIP HOT TARGET T1 T2 T3 T4 All peaks sit below −6. Headroom intact. CLIP! −28 T1 T2 T3 T4 T1 distorts. T4 too hot. T3 quiet, but balance is the issue.
Two contrasting mixes shown on level meters. The black tick on each bar marks the loudest peak that track has hit. Below −6 dB shows green; the orange and red zones above −6 indicate signal getting hot or clipping.

The black tick on each bar marks the highest peak that track has reached. In the healthy mix on the left, every tick sits well below −6 dB, in the green zone, and the four tracks sit reasonably close to each other, so nothing overwhelms or disappears. In the mix that needs work, T1 has hit clip and is distorting (its bar fills the green, orange, and red zones because its peak reached all the way to 0), T4 is creeping into the orange hot zone, T2 sits comfortably in the green, and T3 is in the green but so quiet, at −28 dB, that it will be hard to hear once everything plays together.

Watch the master meter as the piece plays. If it turns red, something is clipping: pull down whichever track is the culprit, usually the loudest element, play again, and check.

The reliable way to set levels is two passes. First, solo each track on its own and check it against the meter: is anything clipping, is anything so quiet it will vanish? Set each track's fader so its peaks sit below clipping. Then un-solo and play the whole thing. Now the question is balance: with everything together, can you hear each element at the loudness you want? Move faders in small amounts, three to six dB at a time, until nothing peaks anywhere and every part you mean to be heard can be heard.

Soloing to check one track at a time is a habit worth building. It comes back constantly: any time you work on a single part's level or tone, soloing lets you hear what you are doing without the other parts in the way. And small adjustments matter more than large ones. A three-to-six dB change is usually enough to bring a part forward or push it back. If you find yourself making fifteen dB moves, something else is wrong: the source may be too loud, or two parts are fighting for the same frequency range, which is a tone problem rather than a level one.

The master track is the channel every track sums into, and its fader sets how loud the whole mix plays back. Leave it alone while you balance individual parts. Reach for it only when the balance is right but the mix as a whole sits too loud or too quiet.

2 · Stereo pan

The position of a track in the stereo field, from full left to full right. Pan is what gives a mix space and width: instead of every sound stacked in the center, parts occupy different points across the stereo image, the way instruments occupy different points on a stage.

Every channel has a pan control, usually right by the fader, running from full left to full right. Centered is the default: equal in both speakers.

Why two ears matter

You localize sound using two cues: interaural time difference (sound from the left reaches the left ear a fraction of a millisecond before the right) and interaural intensity difference (sound from the left is slightly louder in the left ear, because the head shadows the right). Together these tiny differences let your brain place a sound in space.

When the same signal goes equally to both speakers, there is no time or intensity difference between your ears, and the brain reads the sound as straight ahead, between the speakers. That is a centered sound. Pan to one side and you change the proportion sent to each speaker, which changes the intensity difference at your ears, which moves the sound across the perceived space. Stereo width is built entirely on this trick.

The stereo field

Imagine looking down at a stage from above, two speakers in front of you and you in the audience. The space between and around the speakers is the stereo field, and different kinds of sound tend to live in different parts of it.

Top-down view of the stereo field showing typical placement zones: center for foundational and low-frequency sounds, slight L/R for supporting elements, wider spread for textures, hard L/R for special effects. L R YOU CENTER foundation, focal, low frequencies SLIGHT SLIGHT supporting percussion secondary voices WIDER WIDER pads, textures ambient, layers HARD L HARD R FX, ping-pong FX, ping-pong
The stereo field, viewed from above. The listener is at the bottom; the speakers (L and R) are in front. Different kinds of sounds tend to live at different positions across the field.

Where things tend to live

Pop and electronic mixing has settled conventions over decades about which sounds belong where. They are not rules, but they encode useful perceptual realities:

The rough translation: the most foundational or lowest-pitched element of a moment usually wants the center, and supporting or textural material can spread out. On headphones the effect is clearest, so decide what role each part plays, foundation, support, or texture, and pan accordingly.

A small pan, 20 to 40 percent, is often more effective than a hard one, because it suggests a real space rather than an artificial split. And if something with a lot of low-frequency content suddenly feels lopsided when you pan it, trust that: bring it back to center.

3 · EQ

EQ, short for equalization, lets you boost or cut specific frequency ranges in a sound. Think of it as a way to shape timbre, the color of a sound: the same pitch on a flute and on a violin has a different timbre because the harmonics above the fundamental are distributed differently. Another name for that distribution is the spectrum.

EQ reshapes the spectrum: a recording with too much rumble in the low end you want to thin out, a voice that sounds dull and wants brightening up top. It is one of the most-used tools in mixing and one of the easiest to overuse. In a DAW the EQ sits on the channel as a device, shaping the signal as the piece plays without altering the underlying file, so you can change it freely, take after take.

You add an EQ to a track the way you add any device, and it opens a graph. Here is what one looks like:

A typical EQ curve. Frequency runs along the horizontal axis from 20 Hz on the left to 20,000 Hz on the right (logarithmic spacing, as in real EQ dialogs). Amplitude in dB runs along the vertical axis. The curve shows a low cut at the left, a roughly flat middle, a gentle boost around 4 kHz, and a roll-off in the highest frequencies. +12 +6 0 -6 -12 AMPLITUDE (dB) 20 100 500 1k 5k 10k 20k FREQUENCY (Hz) low cut remove rumble flat at 0 dB no change boost at 4 kHz add brightness high roll-off soften the air
A typical EQ curve. Frequency runs left to right (low to high) on a log scale, the way real EQ dialogs are laid out; amplitude runs up and down (boost above the dashed center line, cut below it). The curve traces what the EQ does at every frequency.

An EQ opens with a graph like this: frequency on the horizontal axis, low on the left and high on the right, and amplitude in dB on the vertical axis, boost above the center line and cut below it. You shape a curve by adding control points and dragging them up or down. Where the curve sits at 0 dB, that frequency passes through unchanged; where it bows up, that frequency is boosted; where it bows down, it is cut.

You build that curve from a few standard shapes, and an EQ like Ableton's EQ Eight is just several of them stacked. The ones you reach for:

In a mix, two sounds that crowd the same frequency range mask each other, and the result turns muddy. A small cut in one sound, right where the other one lives, opens room for both. Carving space like this is what keeps a busy mix clear.

Listen for this: every EQ change reshapes the character of a sound. Cuts, curves that go below 0, tend to sound natural; boosts, curves that go above 0, are more obvious and easier to overdo. A small change at the right frequency often does more than a large change at the wrong one.

Space, in practice

Depth gets only a glance here, in the pan-and-depth field above. The tools that set how near or far a part sits, reverb and delay, are worked hands-on in Mixing in Ableton.

Go deeper

If shaping how the finished parts are heard is the part that grabbed you, there is a whole course in it. MUS 485, Advanced Sound Design I: Sampling, Editing, and Mixing, takes mixing well past this first pass.


Read more

Sound on Sound’s technique archive goes deeper on each move here, and on mastering, the stage that comes after the mix is done.

Mixing

Mixing Essentials The basics, start to finish Mix Mistakes: audio examples Common mix problems, by ear

EQ

What Is An Equaliser? How EQ works, and the types Using EQ EQ in practice EQ: A New Perspective Why EQ moves work on the ear

Mastering

What Is Mastering? Mastering Essentials · Part 1 The Three Ms of Mastering Mastering Essentials · Part 2 How Loud Should You Master? Mastering Essentials · Part 3 EQ: Balance, Don’t Match Mastering Essentials · Part 4 Compression & Limiting Mastering Essentials · Part 5 Final Delivery, Requirements & Specs Mastering Essentials · Part 6