How to master a track for Spotify and Apple Music

Mastering for streaming is the final stage that turns a finished mix into a dependable release. It shapes the track’s tonal balance, dynamics, stereo image, peak level and file format so the song translates from studio monitors to earbuds, car systems and phone speakers. The goal is not simply to make the waveform as loud as possible. A strong streaming master should remain clear, energetic and comfortable after a platform processes it.

Spotify, Apple Music and other services use loudness management to reduce the volume of tracks that exceed their chosen reference level. This means a heavily limited master may be turned down without sounding louder than a more dynamic master. Excessive level can also make distortion, harshness and pumping more noticeable after lossy encoding.

Producers working in Sydney, Melbourne, Brisbane, Perth or smaller Australian cities often check masters on several everyday systems before release. A song may be heard through wireless earbuds on a train, a laptop in a share house, a phone speaker in a rehearsal room or a car travelling between suburbs. These listening habits make translation more valuable than impressive meters in one treated studio.

A reliable process begins before the final export. Leave sensible headroom on the mix bus, remove technical problems, confirm that licensed sounds are cleared, and make deliberate decisions about loudness and dynamics. Whether the track is an indie single, a hip-hop release, a dance production or a songwriter demo, careful preparation gives the final master room to work.

Prepare the mix before mastering

Mastering cannot fully repair a crowded or unstable mix. Before opening a limiter, listen for problems that become exaggerated when the track is made louder: muddy low mids, sharp vocal sibilance, piercing cymbals, weak kick definition and bass that changes dramatically between notes. Compare the mix with a few professionally released references at matched playback volume rather than judging the louder file as automatically better.

Keep the stereo master bus from clipping during mixing. A peak level around -6 dBFS is not a compulsory rule, but leaving several decibels of space makes processing more predictable. If the mix already reaches 0 dBFS, reduce the gain before inserting mastering processors; lowering the fader does not undo clipping that has already been printed into the file.

Check the arrangement as well as the frequency spectrum. A vocal buried by synths, a snare that disappears under a chorus or a sub-bass part that consumes all available headroom may need production changes. It is often faster to return to the mix, adjust a few channels and export again than to force a mastering chain to solve a balance problem it was never designed to fix.

Reference tracks should be similar in style and emotional intensity. Use spectrum analysers and loudness meters as evidence, not instructions. A bright electronic master may show more high-frequency energy than an acoustic folk release, while a dense drill or pop mix may have a different crest factor from a live band. Musical intent should guide technical decisions.

Choose loudness without chasing a number

Integrated loudness, measured in LUFS, describes the average level across a complete track. Short-term LUFS shows what is happening over a smaller window, and momentary LUFS reacts even faster. True peak estimates inter-sample peaks that can appear when a digital signal is reconstructed or converted to a streaming codec. These measurements help identify risk, but they do not replace listening.

There is no single loudness setting that guarantees the best result on every service. Many engineers use a moderate master around -14 LUFS integrated as a starting point, then move higher or lower according to genre, arrangement and dynamics. A quieter master can be entirely appropriate if it preserves punch and contrast. A loud master may also work when distortion is controlled and the song’s density suits it.

Watch true peak when setting the final limiter. A ceiling near -1 dBTP is a common practical choice because lossy encoding can create peaks above the original digital samples. Some engineers choose a lower ceiling for very bright, dense or bass-heavy material. If the limiter is reducing several decibels throughout the loudest section, compare it with a less compressed version at matched loudness. Listen for flattened drums, blurred bass transients and a vocal that feels pinned to the front.

Streaming normalisation means listeners can turn a loud track down automatically, but it does not restore dynamics that were removed during mastering. A song with a natural verse-to-chorus lift can feel larger than a constantly crushed track at the same playback level. For an Australian artist competing for attention in crowded Spotify playlists, impact and clarity are usually more useful than chasing a particular loudness screenshot.

Build a mastering chain that serves the song

A simple chain may include corrective equalisation, gentle bus compression, saturation, stereo control, a final limiter and metering. None of these processors is mandatory. Begin with the smallest change that solves a clear problem. A broad cut in muddy low mids may create more space than several narrow filters, while a small high-frequency reduction can make a bright mix easier to encode.

Use compression for movement and cohesion rather than automatic loudness. A slow attack can preserve the initial impact of drums, while a moderate release can help the compressor breathe with the tempo. If the gain reduction meter moves constantly and the groove loses its shape, the mix may need attention before the master. Multiband compression and dynamic equalisation are valuable when a particular range becomes aggressive, but they can also make a track unstable if set without reference to the arrangement.

Stereo widening deserves restraint. Keep deep bass information close to the centre, check mono compatibility and listen for phase cancellation. A wide chorus that collapses on a phone speaker can lose its emotional lift. Mono checks are especially useful for club-oriented productions, where low-end phase problems may become obvious on a single subwoofer or venue system.

Use trustworthy metering at the end of the chain and test the export through an encoder preview when possible. The sample library on Melody Mix can help producers compare the character of different drum, vocal and texture sources before they commit to a final arrangement. Source choices affect mastering: brittle cymbals, noisy loops and overly bright vocal samples may become more tiring after conversion.

Export files that survive platform processing

For a primary delivery master, export an uncompressed WAV or AIFF, normally at 24-bit resolution. Keep the sample rate used by the project when practical, commonly 44.1 or 48 kHz, and avoid unnecessary sample-rate conversion. If a distributor requests a particular specification, follow its current delivery requirements rather than relying on an old tutorial.

Do not add MP3-style encoding before uploading a lossless master. Spotify, Apple Music and distributors may create their own delivery versions, so giving them a clean file reduces one generation of avoidable degradation. Name files clearly with the artist, track title, version and sample rate, then open the exported file from start to finish to check that no silence, clicks, truncation or accidental limiter bypass has occurred.

Create separate versions when the release needs them. An instrumental, clean edit, radio edit, performance version or Dolby Atmos master should be treated as its own deliverable. Do not make a clean version by simply muting random words after the master has been limited; changes to the arrangement can alter peaks and tonal balance, so the alternate version should be checked and mastered independently.

Metadata is part of a professional release workflow. Confirm the spelling of the artist name, featured performers, writers, producer credits, ISRC information, cover artwork details and explicit-content designation. An Australian artist should also maintain accurate records for copyright ownership and performance information. The Copyright Act 1968 protects original musical works and sound recordings, while APRA AMCOS registration and reporting can support royalty administration; neither replaces permission for a sample or vocal recording used in the track.

Test translation and release with confidence

Listen away from the workstation after exporting. Try the master at low volume first, because harsh upper mids and an overpowering vocal often become obvious when the playback level is reduced. Then test it loudly for a short period, checking whether the kick remains controlled and whether the limiter creates audible pumping. Compare the master with the mix and reference tracks at equal perceived volume.

Use several playback environments: studio monitors, closed-back headphones, inexpensive earbuds, a phone speaker and a car. In Australia, commuters in Melbourne and Sydney may hear a release mostly through Bluetooth earbuds and mobile data, while listeners in regional areas may download or cache music before travelling through weaker coverage. These conditions make a stable midrange and intelligible vocal more important than an impressive sub-bass display that disappears on small speakers.

Check the beginning and end of the track carefully. Streaming listeners can skip quickly, so an unintended fade-in, excessive silence or clipped first transient may damage the first impression. Make sure the final reverb tail is not cut short, and compare the gap between tracks if the release is an EP or album. Platform previews can expose problems that were missed during repeated studio playback.

A final quality-control pass should cover loudness, true peak, mono compatibility, clicks, fades, file duration, metadata and alternate versions. It is also worth checking the uploaded result after distribution, because an incorrect file, wrong version or metadata mismatch can occur outside the mastering session. Australian releases may need local tax and royalty records kept alongside the audio paperwork; income from streaming is not a simple fixed payment per play, and rates vary by territory, subscription type and rights ownership.

A good master gives the song enough level to feel present, enough headroom to avoid codec damage and enough dynamics to preserve its musical shape. Practical settings are starting points, not laws: compare at matched loudness, trust repeated listening and choose the version that communicates the song most clearly. Before delivery, export a clean lossless file, inspect it on ordinary Australian listening systems and keep a written record of the final loudness, peak ceiling, version and rights information.