Audio Mastering on Your Phone: What It Actually Does

The short answer

Mastering is the final stage that makes a mix sound finished on ordinary playback: even level, balanced frequency range, controlled dynamics, and a ceiling that prevents clipping. It is four or five steps, not a black box — and knowing which step is doing the work tells you what to adjust when it sounds wrong.

The chain

1. EQ (equalization). Adjusts the balance between frequency ranges. The move that matters most is a gentle high-shelf cut — most mixed tracks are bright, and rolling off 1–2 dB above about 8 kHz removes harshness without dulling the mix. The other classic move is a small low-shelf boost for weight.

2. Compression. Evens out the loud and quiet parts so the track holds its level. Gentle ratio and a small amount of reduction — around 1–2 dB — is what a mix bus wants. Heavy compression is a creative choice, not a finishing step.

3. Stereo enhancement. Mild widening using a subtle technique. Heavy widening is one of the most common reasons a master sounds worse than the mix did.

4. Limiting. Holds the peak below 0 dBFS so nothing clips on playback. Small gain reduction, gentle setting. A limiter set to make a track loud by 6 dB is producing distortion and calling it volume.

5. Loudness normalization. The final target level, applied consistently across a release so everything plays at comparable volume.

Why presets are the right starting point

A mastering preset encodes a genre's typical decisions: a rap master and a jazz master differ in EQ tilt, compression speed, and stereo width. Starting from a preset matching your genre gets you 80% of the way in one move, and the remaining 20% is the part that matters — adjusting it against how the track actually sounds.

The failure mode is treating a preset as the finish. Applying one and stopping gives a track that is technically processed but not actually mastered, because mastering decisions are made by comparison, not by defaults.

What mastering cannot fix

This is the useful half of understanding it:

It cannot fix a bad mix. Mastering is 10–20% of a record's sound. A mix with a vocal buried under the instruments stays buried. No amount of EQ moves a buried vocal to where it belongs — the fix is in the mix, not the master.

It cannot restore a clipped recording. If the waveform is already flat-topped, that information is gone.

It cannot fix a bad performance. Timing and tuning are upstream.

It cannot make a demo sound like a studio recording. It can make a good mix sound finished and finished on every playback system. That is the honest scope.

Reference levels

Rough guides, useful mostly for avoiding extremes:

• Streaming platforms normalize to roughly −14 LUFS, so mastering much louder than that buys nothing — the platform turns it back down and may also reduce dynamics
• Peaks should stay below −0.1 dB to avoid inter-sample clipping
• A finished master commonly sits around −9 to −12 LUFS for loud-leaning music

The practical consequence: if you are releasing to Spotify or Apple Music, mastering to −6 LUFS does not make the track sound better. It just makes the platform turn it down.

Doing it on a phone

A mastering chain runs fine on a phone — the operations are simple DSP, not something that needs a desktop. WaveCraft applies the chain to the finished mix with genre-matched presets you can then adjust, so the step is a matter of listening and moving one or two controls, rather than setting up a signal flow.

The honest limit is the same on any device: you cannot hear the result accurately through phone speakers. Export and check on headphones and on a proper system before calling it finished.