Dithering When Mastering 24-bit to 16-bit Explained

If you have searched this before, you have probably found three articles that contradict each other.

This covers dithering when mastering 24-bit to 16-bit the way it actually works in a session: the mechanism, the numbers that matter, and the decisions that follow from them.

Nothing below is theory for its own sake. Every number here is one you can check on a meter, and every step is one you can run on the next track you open.

Key takeaways

Mastering makes a good mix translate, it does not repair a bad one.
-1 dBTP ceiling, and measure true peak with oversampling.
2 to 4 dB of limiting is a working maximum before punch suffers.
Ask for an unlimited mix if the one you received is already squashed.

What mastering is actually for

Mastering is the last set of decisions made about a record before it goes out. It is the stage where the track stops being a mix you are still working on and becomes a finished object that has to hold up next to everything else in a playlist.

Three jobs, in order of importance: make the tonal balance translate on every system, control dynamics so the record holds attention without fatiguing, and deliver files that meet the technical requirements of wherever they are going.

Loudness is a consequence of doing those three things well. It is not the goal, and since normalisation it is not even a competitive advantage.

The mastering chain in the order each stage assumes.

The chain, and why the order matters

Stage	Purpose	Typical amount
Corrective EQ	Remove rumble and resonances	High-pass 20 to 30 Hz, narrow cuts of 1 to 3 dB
Compression	Glue and dynamic control	1 to 2 dB of gain reduction at 2:1
Tonal EQ	Shape the overall balance	Broad shelves, rarely over 1.5 dB
Saturation	Density and harmonic weight	Subtle enough that bypassing it sounds thin, not different
Stereo imaging	Width above the crossover	Bass mono below 100 to 120 Hz
Limiting	Ceiling and final loudness	2 to 4 dB maximum before it costs you punch
Dither	Bit depth reduction only	TPDF, last in the chain, 24 to 16-bit only

Corrective work comes before anything that responds to level. Compressing a mix with a 40 Hz rumble in it means the rumble is driving your gain reduction.

The numbers that matter

True peak ceiling: -1 dBTP. Sample peak is not the same measurement and will let inter-sample overs through.
Integrated loudness: -14 LUFS covers most streaming. Go quieter for acoustic and classical, louder only if the genre demands it.
Loudness range: 4 to 9 LU for most modern music. Under 3 usually means over-limited.
Crest factor: aim to keep at least 8 to 10 dB between peak and RMS on a full mix.
Headroom in the mix you receive: 3 to 6 dB. More is fine, none is a problem.

If a mix arrives already limited, the honest answer is to ask for an unlimited version. No mastering chain recovers transients that were removed upstream.

Knowing when it is finished

It holds up level-matched against a commercial reference in the same genre.
It translates on monitors, headphones, a phone speaker and in a car.
The low end is still there in mono.
The AAC encode has no clipping and no audible artefacts.
You have stopped making changes that improve it and started making changes that just make it different.

That last one is the real test. There is a point where further processing is motion rather than progress, and recognising it is most of the craft.

How it translates on real systems

Almost nobody hears your record on the system you made it on. The point of every decision above is that it survives the journey.

System	What it exposes	What to check
Phone speaker	No low end at all below roughly 500 Hz	Does the bass line still read from its harmonics?
Earbuds	Exaggerated width and sub, hyped top	Is anything sibilant or fatiguing?
Laptop	Thin midrange-only playback	Is the vocal still intelligible?
Car	Boomy low end and road noise	Does the low end turn to mush at 80 to 120 Hz?
Club system	Everything below 40 Hz, loud	Is the sub mono and controlled?

If it works on a phone and in a car, it works almost everywhere else. Those two are the honest tests.

A ten-minute version of this

If you have one evening and not one week, this is the order that gets the most improvement for the least time.

Two minutes: listen all the way through without touching anything, and write down the three things that bother you.
Two minutes: fix the loudest problem on the list, with the simplest tool that will do it.
Two minutes: check in mono and on a phone speaker. Fix anything that falls apart.
Two minutes: level-match against a reference and note the tonal difference, not the loudness difference.
Two minutes: make one broad corrective move based on that comparison, then stop.

Three specific problems solved beats twenty small adjustments that cancel out. The written list is what stops the session drifting.

What MaestroNet does with this

MaestroNet analyses the source before it processes anything: spectral balance against a genre reference curve, crest factor, stereo correlation across frequency, transient density, and where the energy actually sits.

The targets move with the style. The chain does not. That matters, because a system that swaps in a different EQ curve regardless of what you sent will fight a mix that was already correct.

Loudness targeting to the platform standard, measured integrated across the whole track.
True peak limiting with oversampling, so the AAC encode stays clean.
Bass management below the crossover, keeping the low end mono and centred.
Tonal correction referenced to the selected style, applied in broad strokes rather than narrow ones.

What it will not do is decide that the second chorus should feel bigger than the first. That is an interpretive call, and it stays yours.

Frequently asked questions

What is the most important thing to get right with dithering when mastering 24-bit to 16-bit?

Tonal balance and translation. A track that holds up on a phone, in a car and on monitors is finished. One that only works on your monitors is not.

What loudness should I target?

-14 LUFS integrated with a -1 dBTP ceiling covers almost all streaming. Go quieter for acoustic and classical material, louder only where the genre genuinely demands it.

How much headroom should the mix have?

Three to six decibels, with no limiter on the mix bus. That gives the mastering stage something to work with.

How do I know when it is finished?

When it holds up level-matched against a commercial reference, translates on several systems, keeps its low end in mono, and your changes have stopped improving it.