What Files Should I Send for Music Mastering?

A beautiful master can only begin with the material it is given. If you are asking, “what files should I send for mastering?”, the short answer is: send the best version of your finished mix, with enough room for the music to breathe. The longer answer matters because a few calm decisions at export can protect the intimacy of a vocal, the body of an acoustic guitar, and the natural rise and fall of a song.

For folk, acoustic and singer-songwriter work, the goal is rarely to make everything louder or shinier. It is to let the finished record translate clearly while keeping its character intact. The song is the subject.

What files should I send for mastering?

Send a stereo WAV or AIFF file of each final mix. WAV is the most common choice and will suit almost every mastering workflow. Your file should be exported at the same sample rate used in your session, ideally at 24-bit or 32-bit float.

If you recorded and mixed at 48 kHz, send 48 kHz. If the project is at 44.1 kHz, send 44.1 kHz. There is no benefit in converting a 44.1 kHz session to 96 kHz before mastering. That does not add detail, and it can introduce unnecessary processing. Keeping the file native to the session gives the mastering engineer the clearest, most honest starting point.

A stereo interleaved file is usually right for a stereo mix. In practical terms, that means one file containing both left and right channels, rather than separate mono files. If your track is deliberately mono, send it as mono and say so. A centred vocal and guitar arrangement may be intended to feel direct and close, rather than artificially widened.

Export the full song from the exact start you want heard to the natural end of the final reverb, delay or room decay. Do not trim away the last breath of a guitar or the quiet tail after a final lyric. Those small details often carry the emotional weight of a recording.

Leave headroom, but do not chase a number

Your mix does not need to peak at a particular figure to be ready for mastering. Advice such as “leave exactly -6 dB” is a useful safeguard for some people, but it is not a law. A clean mix peaking at -1 dB can be perfectly workable if it has not been clipped or aggressively limited. Equally, a mix peaking at -8 dB is not automatically better.

What matters is that there is no accidental clipping and that the mix has not been pushed into a loudness processor simply to make it feel finished. If your master channel is red, or if the limiter is constantly taking bites out of transients, pull it back and export a more open version.

Acoustic music often depends on transient detail: the first touch of a picked string, the consonants at the front of a vocal, the gentle lift of a brushed snare. Heavy limiting can flatten these moments before mastering has a chance to make a considered decision. Leave the mix alive.

If you have used gentle bus compression or subtle saturation because it is part of the sound, keep it. Mastering is not about stripping a mix back to a technically pure state. It is about receiving the version that represents your artistic intention. The question is whether a processor is shaping the song, or merely trying to make it louder.

Turn off loudness processing you do not want printed

Before exporting, check the master bus for limiters, clippers, maximisers and loudness plug-ins. If they were only there to help you compare your mix with commercial releases, disable them. A reference level is useful while mixing, but it is not usually something you want baked into the file.

There are exceptions. Sometimes the sound of a limiter is integral to an indie arrangement, or a soft clipper gives a drum loop its particular edge. If removing it changes the feeling of the track, send the processed mix. It can also be helpful to send an unprocessed alternative, clearly labelled, so there is a point of comparison.

Avoid normalising the file on export. Avoid converting it to MP3, AAC or another compressed format. Those formats are fine for sharing a demo, but they are not suitable sources for a final master. The mastering file should be the highest-quality version available.

Send the final mix, not the whole session

For standard stereo mastering, the engineer needs a finished stereo mix rather than your full multitrack session. This keeps the role of mastering clear: refining the finished picture, not rebuilding it from its individual parts.

If you know there is a problem in the mix - perhaps the vocal disappears during the final chorus, or a low guitar note becomes distracting on headphones - address it in the mix if possible. Mastering can improve balance in broad, musical ways, but it cannot always solve a moment that needs an automated vocal ride or a different arrangement decision.

That said, there is no need to hold a song back because you are worried it is not perfect. A good mastering conversation makes room for questions. Mention anything you are unsure about, particularly if the issue appears in one precise section. Clear notes are more useful than apologising for the work.

Include the context around the track

The audio file is central, but a few supporting details help the master serve the release rather than just the waveform. For a single, tell the engineer where the song is headed: streaming only, a digital download, CD, vinyl, or a combination. Different formats may call for different considerations.

For an EP or album, provide the intended track order and the exact titles. Sequence changes how a record is heard. A quiet fingerpicked song after a full-band track may need a different sense of level and tone than it would as a standalone single. If you have thoughts about gaps between songs, fades or where one track should begin in relation to another, include them early.

Reference tracks can be useful too, provided they describe a feeling rather than become a demand for imitation. You might say that you like the vocal closeness of one record, the low-end restraint of another, or the way a third leaves space around an acoustic guitar. Two or three carefully chosen references are enough. The aim is to understand your listening perspective while keeping your own song at the centre.

A short note about the music can be surprisingly valuable. Is the vocal meant to feel almost whispered and fragile? Is the room sound part of the performance? Is a slightly dark guitar tone deliberate? These are artistic choices, not defects to be automatically corrected.

Name and organise your files clearly

Simple, consistent file names prevent uncertainty later, especially across a multi-track release. Include the track number, artist name, song title and version. For example: `01ArtistNameSongTitleFinalMix48k24.wav`.

Avoid labels such as `finalfinalNEW2.wav`. They are understandable in the heat of finishing a release, but they make it harder to know which version is genuinely approved. If you send alternatives, say exactly what differs: `VocalUp`, `NoLimiter`, `RadioEdit` or `Instrumental`.

For most projects, this small delivery checklist is enough:

• One final WAV or AIFF mix per song, exported at the session's native sample rate and bit depth.

• A version without unintended limiting, clipping or normalisation.

• Track titles, sequence and any required spacing for an EP or album.

• A few reference tracks and brief notes on the sound or feeling you want preserved.

• Clearly labelled alternate versions, if they are needed for release or comparison.

Should I send a loud reference version as well?

Often, yes. If you have been mixing through a limiter and that louder playback has informed your decisions, send two files: the open mix for mastering and the limited reference version. Label them clearly. The reference helps communicate what you have been hearing, while the open mix gives the mastering process room to work.

Do not worry if the open version sounds quieter when you first compare it. Louder nearly always feels more exciting at first. Level-match the two versions before judging their tone, punch or emotional impact. You may find that the quieter file contains more depth, movement and space.

For artists working with Tom Lavender Audio, the same principle applies throughout: send the version that tells the truth about the song, alongside any notes that help explain your intention. Technical preparation matters, but it is there to support listening first.

The best delivery is not the most complicated one. It is a clean, clearly named final mix with room around it, accompanied by enough context for someone else to hear what you heard when the song first felt complete.