Module 10 · Mastering Track
Mastering Foundations
Mastering is the last stage between your finished mix and the world hearing your record. It's a different discipline from mixing — fewer tools, smaller moves, higher stakes. This module defines what mastering is, why it exists, the typical chain, and how to prep a mix so it actually masters well.
You finished your mix in Module 8. You've studied pro chains in Module 9. The bounce sounds good in your studio. Why isn't it ready to release?
Because what comes out of your mix bus is a mix, not a master. A mix is a balanced, cohesive song. A master is that mix prepared for the real world — for streaming services that will normalize its loudness, for car stereos and phone speakers and AirPods that will all reproduce it differently, for an album where it has to sit alongside ten other songs without sounding louder or quieter or duller than they do. Mastering is the bridge between "the mix is done" and "the song is ready to release."
Four ideas first about what mastering actually is, then the typical mastering chain, then the prep hygiene that makes everything downstream work.
First, the words
Four concepts that demystify mastering and make the rest of this module land.
Concept 1
Mastering is final polish
Mastering takes a finished stereo mix and prepares it for distribution — controlled loudness, controlled peaks, consistent tone across the album, optimal format for the platform.
Think of a mix as a finished painting. Mastering is the framing, the gallery lighting, and the wall it's hung on.
Mastering operates on the 2-track stereo bounce of your mix — not on the individual instrument tracks. The mastering engineer can't unmute a kick, re-EQ a vocal, or change the snare reverb. They can only shape the whole mix as a single signal. That constraint is the entire point: it forces decisions that affect the whole song coherently. Mastering's tools are limited because its job is limited: small EQ moves, gentle compression, gentle saturation, stereo image refinement, peak limiting, and final loudness optimization. If your mix needs more than that to sound right, the fix is upstream — back in the mix, not in the master.
Concept 2
Mix and master are different jobs
Mixing balances elements inside a song. Mastering balances songs against the world — other releases, playback systems, streaming targets.
Think of a chef plating a single dish (mix) versus a sommelier making sure every dish on the tasting menu pairs with the next (master).
In the mix, you're inside the song — adjusting the balance between vocal and snare, deciding how wet the reverb should be, carving room for the bass. In the master, you zoom out and ask: does this song hold up when it plays after another song? On a phone speaker? At -14 LUFS on Spotify? A great mix can land in mastering and need only 0.5 dB of bus compression and 1 dB of high-shelf air. A flawed mix lands in mastering and reveals every uncorrected mid-range buildup, every overactive low-end. Mastering reveals the mix; it does not rescue it. If the master sounds wrong, mix smarter and try again.
Concept 3
The mastering chain has a typical order
Subtle EQ → gentle compressor → multiband (optional) → saturation → stereo image (optional) → final limiter. Six stages, small moves, big result.
Think of it like a tasting menu's last service — coffee, dessert, mignardise. Each course tiny, perfectly placed, finishing what came before.
The chain you'll see in the visual below is the canonical mastering signal flow. Each stage moves the meter by a hair — total compression usually under 3 dB, total EQ moves usually under 2 dB, total saturation barely audible. The cumulative effect is what mastering engineers call "expensive" — a sense of polish, cohesion, and inevitability that's hard to point to but obvious when missing. Big moves at the master stage almost always mean a problem upstream. If you're tempted to cut 6 dB at 250 Hz on the master, your mix has a mid-range buildup — go fix it in the mix instead.
Concept 4
Prep your mix for mastering
Before you bounce: -6 dB headroom, no master limiter, no master compressor pushed hard, 24-bit WAV, sample rate matching session, dithered if downsampling.
Think of it like sending a manuscript to an editor — clean draft, no track changes, no comments in the margins. Make their job easy.
A mix bounce that arrives in mastering already smashed against a brickwall limiter has nothing left to give the mastering engineer (you, or anyone else). The single most important prep rule: remove or bypass any limiter on your master bus before bouncing the mix. Aim for the loudest peak in the song to hit around -6 dBFS — the average level will sit lower. This is uncomfortable at first. Mixes feel quiet without a limiter. Resist the urge to "make it loud" before mastering — loudness is the master's job, not the mix's. Bounce 24-bit WAV at the session sample rate (44.1 or 48 kHz). Don't apply dither unless you're downsampling bit-depth. The Try-This below is the full prep checklist.
The visual below is the canonical mastering chain. Six stages, in order. Every commercial master you've heard ran through some version of this — the gear varies, the order rarely does.
Six stages, each adding less than 1 dB of polish. Module 11 covers the limiter and LUFS in depth; Module 12 covers when (and when not) to do this yourself.
Mix vs master, side by side
The clearest way to keep them straight is to compare what each stage actually does. Click between the two below. Most beginners blur the line and end up doing mix work in the master (overcorrection) or master work in the mix (loudness wars). Here's the line.
What mastering can fix (and can't)
Things mastering can fix
- Slight tonal imbalances across the whole mix — a bit dark, a bit bright, slightly muddy. ±1–2 dB shelves are the tools.
- Slightly low loudness — pushing the master limiter to hit a streaming target.
- Slight peak inconsistency — taming a few stray transients with the limiter.
- Album cohesion — making 10 songs feel like they belong on the same record by matching their tonal balance and loudness.
- Stereo image refinement — making a mix feel a bit wider or tighter using M/S processing on highs only (mono lows always).
Things mastering can't fix
- A bad balance. If the vocal is too loud or the kick is too quiet, no amount of mastering will rescue that — the relationship is baked into the bounce.
- Mid-range buildup. Mud at 250 Hz across the whole mix isn't a master fix — it's a mix fix. Notching 6 dB on the master damages everything else with that frequency.
- Phase issues. Two tracks fighting in phase produce cancellation. The master can't separate them.
- Distortion in the source. Audible clipping, over-compressed individual tracks, distorted recordings — none of this comes back in mastering. Fix it upstream.
- A boring arrangement. Loudness ≠ excitement. If the song's energy curve is flat, no master makes it interesting. (Module 8's "static balance" stage is where dynamic excitement is built.)
The pattern: mastering refines what's already there. It doesn't add what's missing. Send a great mix to mastering and it gets 5% better. Send a flawed mix to mastering and the flaws stay flawed. Mix first; master second.
A note on order
The mastering chain is shown in canonical order, but skilled engineers do swap stages — putting saturation before compression for more grit, or splitting the EQ into "before compressor" and "after compressor" passes. The principle (Module 9 idea 1: respect the order) still applies: each position changes what the next plugin sees. When you're learning, follow the canonical order. When you can hear what each stage is doing, experiment.
Try this · 30 minutes
Prep a mix for mastering
Open a finished mix from your library. Even better — use the mix you built in Module 8's capstone exercise. Walk through the prep checklist before bouncing.
- Bypass any limiter on the master bus. If you have a brickwall, Maximizer, Adaptive Limiter, or anything similar on the master output, turn it off. This is the move that feels wrong but is correct.
- Bypass any compressor pushed hard on the master. A bus comp doing 1–2 dB of glue (Module 9, Greg Wells) is fine and can stay. A comp doing 4+ dB of GR on the master should come off. The master needs headroom.
- Check the loudest peak. Play the loudest section of the song. Watch the master meter. The peak should be hitting around -6 dBFS. If it's hotter than -3 dBFS or quieter than -10 dBFS, adjust the master fader (not individual tracks) until peaks land near -6 dBFS.
- Set the bounce settings. 24-bit WAV. Sample rate matching session (44.1 or 48 kHz). Stereo interleaved. No dither if staying at the same bit depth/rate. Dither only when downsampling bit depth.
- Add 1 second of silence at the start and end of the bounce. This gives the limiter a clean attack window and prevents click artifacts on playback systems.
- Bounce offline (faster than real time) if possible — modern DAWs are sample-accurate either way, but offline bounce is fewer chances for buffer issues.
- Listen to the bounce. Play the WAV file in a different app than your DAW (QuickTime, VLC, Foobar). Confirm it sounds like the mix and contains no clicks, dropouts, or weirdness. This step catches more bugs than people expect.
- Name the file with version & date. "Songname_mix_v3_2024-04-30.wav" — so when you have v4 and v5 next week, you know what's what.
This bounce is the file you'd hand a mastering engineer if you were hiring one. It's also the file you'll feed your own mastering chain in the modules ahead.
"The job of a master is to translate. The job of a mix is to mean something. Don't confuse the two."— common mastering adage
In your DAW
Mastering happens in a separate session from the mix. Open a new project, import the bounced WAV onto a single stereo track, route it through the mastering chain, and bounce again to a new file. Never master inside the mix session — you'll be tempted to keep tweaking individual tracks, and the line between mix and master will blur.
Stock plugins that map onto the mastering chain in any major DAW:
- Logic Pro: Channel EQ → Compressor (Vintage VCA, 2:1) → Multipressor (if needed) → Tape Delay (with delay 0 ms, just for tape coloration) → Stereo Spread or Direction Mixer → Adaptive Limiter.
- Ableton Live: EQ Eight → Glue Compressor → Multiband Dynamics → Saturator (tape mode, very subtle) → Utility (width control, mono-below-120 Hz) → Limiter (true peak mode).
- Pro Tools: EQ III 7-Band → BF-76 or Pro Compressor → Multi-Band Dynamics → Lo-Fi (subtle saturation) → Stereo Position → Maxim Limiter.
- Reaper: ReaEQ → ReaComp → ReaXcomp → Saturation JS plugin (or free Klanghelm IVGI) → ReaJS Width → ReaLimit.
- FL Studio: Parametric EQ 2 → Fruity Compressor → Maximus (multiband) → Soft Clipper → Stereo Shaper → Soundgoodizer / Limiter.
- GarageBand: Channel EQ → Compressor (Studio FET) → Multipressor → Bitcrusher (off — use just for taste in subtler stages) → Adaptive Limiter. GarageBand's mastering chain is real and works for streaming-grade output.
Module 11 covers loudness and the limiter in detail. Module 12 covers the workflow specific to self-mastering — the part that's different when you mixed the song too.