Mastering is the most misunderstood step in music production. It gets mythologized as a secret process that only experienced engineers with golden ears and $50,000 of outboard gear can do. That's mostly not true anymore.
Modern mastering is more accessible than ever, and understanding what it actually is — and what it isn't — will make you better at mixing too. This guide covers the whole process from mix export settings to streaming-ready loudness targets.
What Is Mastering, Really?
Mastering is the final processing step applied to a stereo mix before release. Its goals are:
- Tonal balance — Make sure the overall frequency balance is right (not too bassy, not too harsh)
- Loudness — Bring the track up to a competitive level for its genre without over-compressing it
- Consistency — If you're mastering an album, make sure all tracks sound cohesive at the same volume
- Translation — Ensure the track sounds good on all playback systems (phones, laptops, club speakers, earbuds)
Mastering is not a way to fix a bad mix. If your mix has problems — a vocal buried in the arrangement, a kick drum that's fighting the bass, instruments that are poorly balanced — no amount of mastering will fix them. Master a great mix; don't master a mediocre one hoping it transforms.
Before You Master: The Mix Must Be Ready
Export your mix before mastering with these settings:
- Format: WAV or AIFF (lossless, not MP3)
- Bit depth: 32-bit float preferred; 24-bit is fine
- Sample rate: Same as your project (44.1 kHz or 48 kHz)
- Headroom: Your mix should peak at -3 to -6 dBFS — not 0 dBFS. The master needs headroom to work with.
- No limiting on the master bus: Remove any limiter or loudness maximizer from your mix bus before exporting. Mastering adds its own.
If your mix is peaking at 0 dBFS and the meters are slamming the ceiling, you haven't left headroom for mastering. Turn down your mix bus by 3–6 dB and re-export before mastering.
The Mastering Signal Chain
A typical mastering chain has 4–5 stages. Here's the standard approach:
-
EQ — Tonal CorrectionA gentle, transparent EQ to correct any overall frequency imbalances. Low-cut below 20–30 Hz to remove sub-bass rumble. High-pass is gentler here than in mixing — you want to keep warmth. Use wide, gentle moves of ±1.5 to ±3 dB maximum. No surgical cuts like you'd use in mixing. This is broad shaping.
-
Compression — Glue and DensityA mastering compressor with a slow attack (30–60 ms), moderate release (200–500 ms), low ratio (1.5:1 to 2:1), and gentle gain reduction (1–3 dB max). The goal is to "glue" the mix together and add density — not to audibly pump or squash it. Use a mix knob (parallel compression) if available. If you can hear the compression clearly, pull back.
-
Stereo Imaging (Optional)Subtle widening of the stereo image if the mix sounds narrow. Narrow, not wide, in the low end (below 200 Hz, keep mono — mono-compatible bass is essential). Mid-side processing can add subtle air to the sides. Optional stage — skip if your mix already has good width.
-
Saturation / Harmonic Exciter (Optional)Subtle tape or tube saturation adds warmth and harmonic richness to digital mixes. Use very gently — just enough to add some analog character without obvious coloration. Particularly useful for mixes that sound too "clean" or digital.
-
Limiter — Final LoudnessThe last plugin in the chain. A transparent limiter (FabFilter Pro-L 2, iZotope Ozone Maximizer, or your DAW's stock limiter) brings the track up to target loudness. Set the ceiling at -1.0 dBTP (True Peak) for streaming. Reduce the input gain until the limiter's gain reduction is 3–6 dB for most genres. See loudness targets below.
Loudness Targets for Streaming
Every streaming platform has a loudness normalization target — it measures your track's LUFS (Loudness Units relative to Full Scale) and adjusts its playback level to match. If your master is louder than the target, it gets turned down. If it's quieter, it may get turned up.
The implication: there's no benefit to making your master crushingly loud. If Spotify targets -14 LUFS and you master to -8 LUFS, it just gets turned down — and you've sacrificed dynamic range for nothing. Master to the target, not above it.
| Platform | Target LUFS | True Peak Ceiling | Notes |
|---|---|---|---|
| Spotify | -14 LUFS | -1.0 dBTP | Normalizes to -14; masters above this get turned down |
| Apple Music | -16 LUFS | -1.0 dBTP | Slightly quieter target; classical/jazz benefit |
| YouTube | -14 LUFS | -1.0 dBTP | Music videos same as Spotify |
| Tidal | -14 LUFS | -1.0 dBTP | Lossless playback — quality matters more here |
| SoundCloud | -14 LUFS | -1.0 dBTP | Less consistent normalization; aim for -14 |
| Club/DJ Use | -9 to -6 LUFS | -0.3 dBTP | Louder is expected in club context; no normalization |
Genre context: Classical and jazz typically master to -18 to -20 LUFS to preserve dynamic range. EDM and hip-hop typically hit -8 to -10 LUFS because the genre expectation is loud, punchy sound and these tracks are often played in contexts without normalization.
For most genres releasing on streaming: master to -14 LUFS integrated with a -1.0 dBTP true peak ceiling. This is the sweet spot that sounds competitive on Spotify and Apple Music without over-limiting.
How to Use a Reference Track
A reference track is a commercially released song in the same genre that you want to emulate in tone and loudness. Using one while mastering keeps your decisions anchored to what actually sounds competitive on streaming platforms.
Import your reference track into your mastering session at the same loudness level as your master (use a reference plugin or manually level-match). A/B between your master and the reference frequently. You're listening for:
- Is the low end balanced similarly? (Is your bass or kick heavier or lighter?)
- Is the vocal presence in the same range?
- Does your master feel dynamically similar, or is it more compressed/dynamic than the reference?
- Is the "air" and high-end brightness in the same territory?
The goal is not to sound identical to your reference — it's to be in the same ballpark so your track doesn't stand out as obviously under or over-processed.
How to Know When Your Master Is Done
This is the hardest part. Here's a checklist:
- ✅ The track sounds good on headphones, laptop speakers, phone speakers, and a car stereo
- ✅ The loudness is at your target (-14 LUFS or genre-appropriate)
- ✅ The true peak doesn't exceed -1.0 dBTP
- ✅ The A/B comparison with your reference track doesn't reveal obvious tonal imbalances
- ✅ The mono sum sounds clear and punchy (check by collapsing to mono)
- ✅ You've listened to it fresh the next morning and still feel good about it
Never finalize a master in the same session you started it. Come back the next morning with fresh ears. If something bothers you after sleeping on it, fix it. If it still sounds right, you're done.
Once your track is mastered, make sure your mix was solid before you got here. If you want to revisit the mixing stage, read the full vocal mixing guide — the techniques there apply to every element of a mix, not just vocals.