How to prepare your mix for mastering

A good master starts with a clean export. Leave headroom, take the limiter off the mix bus, and bounce a 24-bit WAV at your session sample rate. Do that and the engineer works on your mix, not around its problems. Here is the full checklist.

1. Leave headroom. Aim for peaks around -6 dBFS and an average level that is comfortable, not crushed. Mastering needs room to work. A mix that already touches 0 dBFS leaves nothing to shape.
2. Take processing off the mix bus. Remove any limiter or heavy clipper on the master channel. Gentle bus compression or EQ that is part of your sound can stay, but bounce a version without the limiter. If you love your bus chain, send both: one with, one without.
3. Do not normalise, dither, or fade to silence unnaturally. Leave normalisation off. Let mastering handle final level and dither. Keep natural start and end tails rather than hard cuts.
4. Export the right file. Bounce a WAV or AIFF at 24-bit (or 32-bit float) at the session sample rate (44.1 kHz or higher). Do not convert sample rate on the way out, and do not send an MP3.
5. Name it clearly. Use artist, track title and a version note, for example DeadQuiet_TrackName_v1_nolimiter.wav.
6. Flag what you want. A one-line note helps: a reference track you love, any known issues, and the alternates you need (clean, instrumental, radio, vinyl).

Alternates to plan for

If you will need a clean (no explicit lyrics), an instrumental, a radio edit or a vinyl-prepped version, mention it up front. It is easier to line these up from the start than to chase them later.

Once your bounce is ready, send us the record and pick your engineer. If you want to hear a streaming-ready master instantly first, drop it into DQ Master Rack.

Written by the Dead Quiet engineers. Meet the engineers.

How loud should your master be?

For most streaming services, aim for about -14 LUFS integrated with true peaks no higher than -1 dBTP. LUFS measures perceived loudness over the whole track, and streaming platforms turn every track towards their own target, so chasing a louder number no longer wins. Here is what that means in practice.

Why loudness normalisation changed the game

Services play all tracks at a similar loudness by turning quiet tracks up and loud ones down. So a track crushed to -8 LUFS does not sound louder than one at -14 LUFS on Spotify. It arrives with less dynamic range and less punch. Normalisation makes dynamics more valuable than raw volume.

Targets by platform

These are the figures commonly cited by mastering engineers. Treat them as a guide, not gospel, and note that YouTube and Tidal are reported inconsistently.

Leave true-peak headroom

Keep peaks at -1 dBTP or lower. Lossy encoding (AAC, Ogg Vorbis) can push peaks higher after you upload, so a master that reads 0 dBFS on a sample-peak meter can clip audibly once it is streamed. A true-peak limiter set to -1 dB protects against this.

In practice

Master so the track sounds right first, then check it lands near -14 LUFS integrated with peaks at -1 dBTP. Do not master to the number at the cost of the feel. A track with controlled dynamics and clean peaks will translate better across phones, cars and club systems than one squashed for a loudness contest it cannot win.

You can check your integrated LUFS, RMS and true peak with a meter like Meeter. To hear your track mastered to a streaming-ready level for free before you pay, run it through DQ Master Rack, or bring us the record for a full analogue master.

Written by the Dead Quiet engineers. Meet the engineers.

Sources: Fora Soft LUFS targets per platform 2026, SoundBridge mastering for streaming 2026, Mastering The Mix.

AI mastering vs a mastering engineer

AI mastering is fast and cheap. A human master costs more and takes longer. The honest answer to which you need is that it depends on the record, the budget and how finished the mix already is. Here is how we think about it, as the engineers who built both an AI mastering tool and the studio chain behind it.

What AI mastering actually does

An AI master analyses your track and applies EQ, compression and limiting to hit a target curve and loudness. The good ones get you a clean, loud, competitive master in seconds for a few pounds. For demos, club tests, reference bounces and release deadlines, that is often all a track needs. What it cannot do is hear what you meant. It does not know the vocal should sit forward, that the drop should feel wider than the verse, or that the sub is fighting the kick in a way that only shows up on a big system.

What a mastering engineer adds

A person brings intent and context. They catch a mix problem before it gets baked into the master, make calls across a whole EP so it sounds like one body of work, and shape the record for where it will actually play. On an analogue chain, that character goes in through real hardware rather than a model of it. The trade is time and money, and it is worth it when the release matters.

A simple way to choose

• Demo, rough or deadline tonight: AI master. Hear it finished, pay only to export.
• A single you care about, or a mix you are unsure of: a human master, or at least a human ear on it first.
• An EP or album that needs to feel consistent: a human master every time.

You do not have to pick once and for all

This is why we built DQ Master Rack on our own mastering chain: you drop a track, hear the full master for free, and only pay to export. If the record then deserves the studio, the step up to a human analogue master is there. Same sound philosophy, two speeds.

If you are deciding on a specific track, hear it through DQ Master Rack first, then bring us the record if it needs the room.

Written by the Dead Quiet engineers. Meet the engineers.