Mastering & Mix Delivery

What Mastering Does

Mastering serves three purposes:

1. Final tonal balance. A mastering engineer listens with fresh ears in a calibrated room and makes broad EQ adjustments to address tonal issues the mix engineer may have stopped hearing after hours of work.

2. Loudness and dynamics. Mastering brings the mix to competitive loudness for the target medium (streaming, vinyl, CD) while preserving as much dynamic range as possible. This is primarily done with limiting and compression.

3. Consistency across an album. If you’re releasing multiple songs, mastering ensures they sit together tonally and in loudness — the listener shouldn’t need to adjust their volume between tracks.

Preparing a Mix for Mastering

Headroom

The headroom conversation is a little overblown. Yes, a mix peaking at -3 to -6 dBFS gives the mastering engineer room to work. But any mastering engineer worth their salt is more than capable of lowering your mix by a few dB if they need headroom. As long as you aren’t clipping — check your true peak levels and make sure they don’t go above -0.1 dBTP — it’s a non-issue. If your mix hits the red line, you’re not going to know whether it was actually clipping or not, so err on the side of leaving a little space.

No Limiting on the Master

Remove any limiter, maximizer, or heavy compression from your master bus before sending to mastering. If you’ve been mixing into a limiter (which many engineers do for monitoring purposes), bypass it when you print the final mix. The mastering engineer will apply their own limiting with better tools and more headroom.

If you have a gentle bus compressor on the master (1–2 dB of glue compression), that’s usually fine to leave in — it’s part of the mix sound. But heavy limiting, maximizers, and loudness processing should come off.

File Format

Print your final mix as a WAV or AIFF at the session’s native sample rate and bit depth. If you tracked at 48 kHz / 24-bit, deliver at 48 kHz / 24-bit. Don’t dither or truncate to 16-bit — the mastering engineer will handle that conversion.

Include a clean, unprocessed vocal-up version and a vocal-down version (vocal 3 dB up, vocal 3 dB down). The vocal tends to be one of the things that gets adjusted in revisions, and printing these alternates can save you running off a new mix. You might find that the vocal sits differently when it goes through a final pass of bus compression, and having a 3 dB down version ready to swap in for a verse can save hours. Include instrumental and a cappella if requested. Label everything clearly.

Notes for the Mastering Engineer

Tell them what you want:

• Reference tracks for loudness and tonal character
• Target medium (streaming, vinyl, CD)
• Any specific concerns (“I’m worried about the low end,” “The vocal might be too bright”)
• Whether this is a single or part of an album

Good communication prevents wasted rounds of revisions.

Stereo vs. Stem Mastering

Stereo Mastering

You deliver a single stereo mix. The mastering engineer works with that stereo file — they can EQ, compress, and limit the overall mix, but they can’t adjust individual elements (turn up the vocal, reduce the kick, etc.).

This is the standard approach and works well when the mix is solid.

Stem Mastering

You deliver separate stems (submixes): drums, bass, guitars, keyboards, vocals, effects. The mastering engineer can process and balance each stem independently before combining them.

Stem mastering gives the mastering engineer more control, which is helpful if:

• The mix has balance issues the mix engineer couldn’t fully resolve
• The artist wants the mastering engineer to make creative adjustments
• The song is destined for multiple formats (a club mix and a radio edit from the same stems)

The downside: it’s more expensive, takes longer, and if the mix engineer’s bus processing depends on all elements playing together (sidechain compression, bus glue), the stems won’t sound the same in isolation.

Basic Self-Mastering

If you trust your room and your perspective, you can master your own material. A second set of ears is half of what you’re paying for with a professional mastering engineer, but self-mastering has its place — especially for delivering client mixes. If you’re giving mixes to a client, they don’t understand that your mix is going to come in quiet. You may need to do some rough mastering so they don’t feel like you don’t know what you’re doing. It’s always better to be honest and explain, but I typically run off an unmastered version and a mastered version alongside it and ask the client which is most useful.

The Mastering Chain

A simple mastering chain, in order:

1. EQ: Broad, gentle adjustments. A slight high shelf boost for air. A slight low-end cut if the mix is boomy. No surgical cuts — those should have been done in the mix.

2. Compression: A gentle stereo compressor for glue. Low ratio (1.5:1–2:1), slow attack (30ms+), auto or slow release, 1–2 dB of gain reduction. The mix should breathe slightly with the compressor engaged.

3. Limiter: The final brick-wall limiter that catches peaks and brings the overall level to the target loudness. Set the ceiling to -1.0 dBFS (or -0.3 dBFS for streaming) and push the input gain until you reach your target loudness.

LUFS and Loudness Standards

Different platforms have different targets:

• Spotify: -14 LUFS (integrated)
• Apple Music: -16 LUFS
• YouTube: -14 LUFS
• CD/Vinyl: Variable, but typically -10 to -12 LUFS

If your master is louder than the platform’s target, the platform will turn it down. If it’s quieter, the platform will turn it up (on most services). There are different schools of thought on this. Some argue that mastering loud still has advantages — with certain styles of music, loud is pretty, and that extra bit of aggressive crunch is part of the sound.

Think of mastering in three stages: perfect (getting rid of imperfections), pretty (last sculpting to make it translate and shine), and loud (bringing it to competitive level). What you’ll find is that when you start making it louder, you begin to compromise pretty. The balance between loud and pretty is really what you’re paying for when you hire a mastering engineer. Separating those processes in your mind is a healthy way to approach it.

Aim for the dynamic range that serves the song. A jazz quartet at -16 LUFS will sound more musical than the same performance crushed to -8 LUFS.

Advanced Mastering Techniques

Mid-Side Processing

Separate the stereo signal into mid (center content) and side (stereo difference) components. Process each independently. A common mastering move: boost the sides slightly for a wider master, or cut the low end from the sides to tighten the bass in mono.

A word of caution: proceed carefully with mid-side processing. If you don’t know what you’re doing, you stand to do more harm than good. Mid-side gives you control that feels powerful, but it’s easy to create problems that are hard to diagnose.

Dynamic EQ and Multiband Compression

Both allow frequency-dependent dynamics processing, but they’re different tools. A dynamic EQ reduces a specific frequency only when it exceeds a threshold — more transparent, less coloring. A multiband compressor splits the signal into frequency bands and compresses each independently — more control, but can change the tonal balance if overused.

In mastering, dynamic EQ is usually preferred for problem-solving (taming a harsh frequency that appears on certain passages). Multiband compression is used for broader shaping (controlling a boomy low end or taming an overly dynamic high end).

Archiving and Future Proofing

There’s an important distinction between a backup and an archive. A backup is a working copy — something you can get back to if things go wrong while you’re still in the middle of the project. An archive is when you pack it up, put a bow on it, and stick it on your shelf for safekeeping. People usually get in trouble when they conflate the two. Either way, redundancy is the only way to ensure either will work.

When the project is done:

• Archive the full session at the native sample rate and bit depth
• Archive the printed stems
• Archive the mastered files in all delivered formats
• Keep notes on the session: processing chain, plugin settings, any issues

Here’s the thing you need to accept: assume you will not be able to open this session the next time you need it. In ten years, SchmoTools 11 is not going to authorize your old session. The people who made your plugins may not be in business. There may be no one to issue a response to the challenge when you try to authorize. The one thing you can count on is that this session is not permanent. Your best defense is stems — audio files that any DAW on the planet can open. It’s an arduous process, but you’re doing it with the idea that this song will be a success. Because if you need this material years from now, that’s going to be a good problem to have.

The Final Thought

After eighteen chapters of tools and techniques, the most important thing is this: the mix serves the song. Not your ego, not your plugin collection, not the technical requirements of the delivery spec. The song.

A great mix makes you feel something. It might be technically imperfect — a little too much reverb, a slightly boomy low end, a vocal that’s a dB too loud — and still be undeniably right because it captures the emotion of the performance.

The technical skills in this guide are the foundation. They let you hear problems, diagnose them, and fix them efficiently. But the art of mixing is knowing when to stop fixing and start listening. When the song feels right, it’s done. Everything after that is fidgeting.

Trust your ears. Trust your room. Take breaks. And when in doubt, reference a record you love.

What to Practice

• Prepare a mix for mastering: remove the master limiter, check headroom (-3 to -6 dBFS peaks), print at native sample rate. Package it with notes for a hypothetical mastering engineer.
• Set up a basic mastering chain (EQ → compressor → limiter) on a finished mix. Target -14 LUFS integrated with a -1.0 dBFS true peak ceiling. Listen to what the limiting does to the transients at different input gain levels.
• Master the same mix to three different loudness targets: -14 LUFS, -10 LUFS, and -7 LUFS. Listen to each version and note where the dynamics start to suffer.
• A/B your self-mastered version against a commercially released reference track at matched loudness. Note the differences in tonal balance, width, and dynamic feel.