How to Prepare Your Mix for Mastering — What to Check Before You Export

The quality of a master is bounded by the quality of the mix it comes from. That's obvious. What's less obvious is that even a great mix can arrive at a mastering engineer in a state that limits what's possible — not because of the music, but because of how it was exported.

This article covers the preparation side: what your mix needs to look like before it leaves your DAW, what format to export in, what to include with the file, and what to avoid. It assumes the mix itself is done and you're preparing for the final step before distribution.

Leave headroom on the master bus

The most important preparation step is this: make sure your mix bounces without the master bus clipping, and with enough headroom that the mastering engineer has room to work.

The practical target is a mix that peaks no higher than −3 to −6 dBFS on the loudest moment. Some engineers prefer −6 dBFS as a floor. Somewhere in that range is fine. The specific number matters less than the principle: the waveform should not be hitting 0 dBFS or clipping at any point.

Here's why headroom matters. Mastering involves EQ, compression, and limiting. All three can increase the peak level of the signal. If your mix arrives at 0 dBFS, there is no room to add anything without introducing digital clipping. The mastering engineer either has to start by turning your file down (which is fine technically but wastes dynamic range) or they're immediately working around a ceiling you've created unnecessarily.

A mix exported at −4 dBFS peak gives the mastering chain something to actually work with from the start.

Remove master bus processing before you export

This is the step that most artists get wrong, and the one that causes the most avoidable problems.

Many producers have processing on the master bus during mixing — a limiter to control peaks, a multiband compressor for density, a saturation plugin for warmth, an EQ for overall tonal shaping. This is completely normal and useful during the mixing process. It gives you a rough sense of how the final product might sound and helps you make better decisions on individual tracks.

Before you export for mastering, bypass or remove all of it.

The mastering engineer's job is to apply professional mastering processing to your mix. If your mix already has a brick wall limiter on the master bus, the file arrives with crushed transients, a flattened dynamic envelope, and a waveform that can't be undone. The mastering session now starts from an already-damaged signal. Whatever the mastering engineer applies goes on top of that damage, not underneath it.

The argument against removing master bus processing is usually "but without it, the mix sounds completely different and I can't hear how it feels." That's fine. Your job as the mixer is done. Export with the master bus clean. The mastering engineer will handle the rest from there.

If your mix needs the limiter on to sound right, the mix may need more work before it's ready for mastering. A well-balanced mix should have enough dynamic integrity to survive without a limiter holding it together.

The one exception: if your mix is specifically for a format like video game audio or film where the delivery requirements differ, discuss this with the mastering engineer before exporting.

Export settings that actually matter

Keep these as simple as possible:

• Format: WAV or AIFF. Not MP3, not AAC, not OGG. Compressed formats discard audio information that can't be restored. Always export the final mix as an uncompressed file.
• Bit depth: 24-bit. If you mixed in a 32-bit float session, export at 24-bit — most mastering tools work natively in 24-bit and the file will be smaller. Do not export at 16-bit; this discards low-level information and introduces dither artifacts that should only happen at the final mastering stage.
• Sample rate: Match your session's native sample rate. If your session runs at 44.1 kHz, export at 44.1 kHz. If it runs at 48 kHz, export at 48 kHz. Do not upsample or downsample during this export — the mastering engineer will handle any necessary conversion for the delivery format.
• Dither: Off. Dither is applied at the final stage of mastering when reducing bit depth for delivery. Applying it at the mix export stage and then again during mastering introduces unnecessary noise.
• Normalize: Off. Do not use your DAW's "normalize" export option. Let the file export at whatever level the mix naturally produces after you've set headroom as described above.

What to actually send

The primary deliverable is the final stereo mix bounce described above. Beyond that, what you send depends on the service.

For stereo mastering (the most common case), you send a single stereo WAV file — the complete final mix.

For mix and master services, you send the stems or the multitrack session, not a stereo bounce. The engineer needs to access individual elements to balance them properly. How stems are organized varies by engineer — ask before you prepare them. Common approaches include grouping by instrument type (drums stem, bass stem, instruments stem, vocals stem) or sending individual tracks. The key is that every stem together, summed at unity gain, should sound like the complete mix.

In both cases, include the following with your file:

• The original project tempo and time signature
• The key of the track, if known
• The intended streaming platforms (affects loudness targeting)
• Any timing or delivery deadlines
• Notes on anything specific you want preserved — a particular vocal character, a specific low end feel, an element that shouldn't be changed

This context helps more than most artists expect. "I want it to feel wide but controlled" is more useful than "make it sound good."

Send a reference track

A reference track is a commercially released song that you want your master to be in the same sonic territory as. It doesn't have to be the same genre. It just has to give the mastering engineer a target for the overall tonal balance, width, and energy level you're after.

Good references are tracks you've listened to extensively on the speakers or headphones you mix on — tracks where you understand the relationship between what you hear in your monitoring environment and how they actually sound. "This song sounds massive on everything" is useful. "I heard this once and thought it was bright" is not.

Send one to three references. More than that creates conflicting targets that are harder to work with than no reference at all.

You don't need to send the actual audio file. The track name and artist is sufficient. The mastering engineer will have access to the same commercial release you're referencing.

The most common preparation mistakes

In order of frequency:

1. Leaving the limiter on the master bus. Already covered, but worth repeating because it's the mistake that causes the most frustration on both sides.
2. Exporting at 0 dBFS or with clipping. This can't be fixed after the fact. Clip indicators on the master bus mean the file is already damaged.
3. Sending an MP3 and calling it a final mix. The quality loss from even a high-bitrate MP3 is audible in a professional mastering context. Always send uncompressed.
4. Over-compressing individual tracks to the point where there's no dynamic range left. The mastering engineer can't restore dynamic range that was removed during mixing. If the snare has zero transient punch after a heavily overdriven compressor, no amount of mastering EQ or limiting will bring it back.
5. Not sending context. A file with no information about genre, references, platform, or deadline is harder to master well than one with a paragraph of notes.

Quick reference checklist

That's the full preparation. None of this is particularly technical — the hardest part is remembering to turn the limiter off before you bounce. Do that, and the mastering session starts from a clean signal that can actually be improved.