Mastering Revisited — Dynamics in Context

The Mastering Chain

A typical mastering chain processes the stereo mix through a series of effects, each doing a specific job. The order matters — each stage feeds the next.

1. Corrective EQ

First pass: fix problems. If the mix has too much low-mid buildup (common in home studios with untreated rooms), cut it. If the top end is dull, a gentle shelf boost opens it up. If there’s a harsh resonance at 3 kHz, notch it out.

This isn’t the same as mix EQ. You’re working on the entire stereo signal — every instrument, every vocal, every element is affected by every move. The cuts and boosts should be small: 1-2 dB is a significant mastering move. If you’re reaching for 6 dB cuts, the mix needs more work before mastering.

A high-pass filter at 20-30 Hz is almost always appropriate. Frequencies below the audible range eat headroom without contributing anything you can hear. Removing them gives the limiter less to work with, which means less distortion and more loudness for the same effort.

2. Compression

Master bus compression serves two purposes: it controls the macro-dynamics (the difference between the loudest and quietest sections of the song) and it “glues” the mix together — adding a subtle cohesion that makes all the elements feel like they belong in the same space.

Settings for master bus compression are conservative. Ratio of 1.5:1 to 3:1. Slow attack (let the transients through). Auto or medium release. Gain reduction of 1-3 dB on the loudest peaks. If you’re seeing more than 4-5 dB of gain reduction on the master bus, you’re compressing too hard.

The character of the compressor matters here more than the settings. Different compressor designs (VCA, optical, tube) impart different feels — Chapter 19 covers this. For mastering, transparent VCA-style compressors are the most common starting point.

3. Enhancement EQ (Optional)

Second EQ pass, after compression: subtle enhancements. A high shelf to add “air.” A gentle low shelf for warmth. These are the “pretty” moves from Chapter 9 — small adjustments that enhance what’s already working.

Not every master needs this step. If the mix sounds good after corrective EQ and compression, don’t add processing for the sake of it.

4. Limiting

The limiter is the final stage before the output. It’s a compressor with an infinite ratio and a very fast attack — nothing above the threshold gets through. Its job: bring the average level up to the target loudness while preventing the peaks from exceeding the ceiling.

Ceiling: Set this at -0.2 dBFS (or -0.3 for extra safety). This prevents intersample peaks — clipping that can occur during digital-to-analog conversion — from causing distortion on playback systems.

Gain/Input: The amount of level you’re driving into the limiter. This is the loudness control. More input gain = louder master, but also more limiting (more dynamic range lost). Less input gain = quieter, more dynamic master.

The art of limiting is finding the balance: loud enough for the platform targets (Chapter 9), dynamic enough that the music breathes. Watch the gain reduction meter — if the limiter is consistently reducing more than 3-4 dB, you’re probably pushing too hard. The music will sound squashed, fatigued, and flat.

5. Metering

The final check, not a processing stage. Your metering should tell you:

• Peak level: Should not exceed -0.2 dBFS (your limiter ceiling handles this)
• Integrated LUFS: Your overall loudness target. -14 for Spotify/YouTube, -16 for Apple Music
• Short-term LUFS: Shows how loudness varies across the song. The chorus should be louder than the verse — if they’re the same, you’ve over-limited
• True peak: Accounts for intersample peaks that might clip on conversion. Should stay below your ceiling

When to Master Yourself vs. Hire Someone

Master your own work when you’re learning, when you’re making demos, or when the project budget doesn’t justify a mastering engineer. The practice is invaluable — you learn what mastering can and can’t do, which makes you a better mixer.

Hire a mastering engineer when the project matters commercially, when you’ve lost perspective on the mix (you’ve heard it too many times), or when you want the objectivity of fresh ears in a calibrated room. A good mastering engineer hears things you’ve gone deaf to.

The worst approach: mastering your own work under deadline pressure with no reference tracks and no monitoring calibration. That’s when you make the mistakes you can’t hear until the track is released.

Gain Staging the Master Bus

Before you put anything on the master bus, check the level feeding it. If your mix is peaking at 0 dBFS, you have no headroom for master bus processing. Pull all your faders down proportionally (or use a trim plugin at the top of the master chain) until the mix peaks at -6 to -3 dBFS. Now the EQ, compressor, and limiter have room to work.

Every plugin in the mastering chain should maintain this headroom through to the limiter. If the corrective EQ adds 2 dB, either compensate with the plugin’s output control or account for it at the next stage. The limiter should be the only thing that brings the final level up to the ceiling.

What to Practice

1. Build a mastering chain. On a finished mix, add corrective EQ, a gentle compressor (2:1, slow attack, 1-2 dB gain reduction), and a limiter (ceiling at -0.2 dBFS). Adjust the limiter input until the LUFS meter reads your target. Compare the mastered version to the unmastered mix at matched levels.
2. Push the limiter too far, on purpose. Drive 8-10 dB into the limiter and listen to what happens. The transients get crushed, the music sounds flat and lifeless, and a subtle distortion appears. Now back off until the life returns. This teaches you where the limit is.
3. Compare your master to a commercial release. Load a reference track (Chapter 10 method), level-match it, and A/B against your mastered version. Where does it fall short? Those differences tell you what to work on — in the mastering chain and in future mixes.