How loud should mix be before mastering?

These match what our mastering order form asks for: peaks no louder than -6 dB, no clipping, no normalization, compressor, EQ or limiter on the master bus, and a stereo 24-bit WAV at 44.1 or 48 kHz.

What dB should my mix be before mastering?

Most mastering engineers ask for digital peaks somewhere between -9 and -6 dBFS. That gap below 0 dBFS is the headroom the mastering engineer works in.

EQ boosts, compression and saturation all add level. With headroom in the file, the engineer can apply them without first turning your song down or running into clipping.

Why is headroom important for a mastering engineer?

Headroom prevents clipping. A mix that arrives at -0.1 dBFS leaves no space for any processing that raises the level, so every boost has to be paid for with a gain cut first, or with distortion.

The second reason is inter-sample peaks. They don’t show on a normal sample-peak meter, but they still use up headroom.

How do inter-sample peaks affect your audio headroom?

A peak meter in most DAWs reads the highest sample value. The waveform that comes out of a converter can go higher than that.

What are inter-sample peaks?

When a digital-to-analog converter rebuilds the waveform from the samples, the smooth curve it draws between two samples can rise above both of them. If those samples are already close to 0 dBFS, the reconstructed peak goes over full scale. These overs are called inter-sample peaks, or true peaks.

So a file that never shows a sample above 0 dBFS can still distort on playback through speakers or headphones.

Are inter-sample peaks possible in any digital audio?

Yes. They can occur at any sample rate and in any format, WAV included. Lossy encoding makes them worse: MP3 and AAC encoders change the waveform, and the decoded file often peaks higher than the source. That’s why streaming services ask for extra true-peak headroom on the final master.

How inter-sample peaks steal headroom you leave for mastering engineers

True peaks can sit 1 dB or even 2 dB above the sample peaks, and more after lossy encoding. With peaks around -6 dBFS, that still leaves several dB of clean headroom for mastering.

What should the true peak be before mastering?

A true-peak meter (as defined in ITU-R BS.1770) oversamples the signal and shows the peak including those inter-sample overs.

For a mix, true peak isn’t a number you need to hit. If the sample peaks are around -6 dBFS, the true peak will be well below 0 dBTP. It matters on the final master. Spotify recommends -1 dBTP, while most commercial masters, ours included, use a ceiling around -0.3 dB. Either way, that is set in mastering, not in the mix.

Can a mix be too quiet for mastering?

In theory yes, in practice almost never.

Accessible dynamic range depends on audio bit depth

A 24-bit file has a theoretical dynamic range of about 144 dB, and a 32-bit float file far more. A mix peaking at -12 dBFS loses nothing that matters.

It only becomes a problem if you bounce at 16-bit, which has about 96 dB of range. Then very quiet detail sits closer to the noise floor. That’s one more reason to send 24-bit.

It’s better to mix quiet than too loud

Turning a quiet 24-bit mix up costs nothing. Undoing a limiter is impossible. If you’re unsure, leave more headroom, not less.

How do I know if my mix is loud enough?

What this question usually means is: will the mix still sound right once it’s mastered loud?

A good way to check is to put a limiter on the master bus while you mix, purely for monitoring. Push it to the loudness you expect for your genre (for example -14 LUFS for a dynamic release or -8 LUFS for a loud one) and listen to what changes. Switch it off before you bounce.

If the mix falls apart under the limiter (the kick and snare disappear, the vocal jumps forward, the low end blurs), fix that in the mix. The goal is a mix that sounds right both with and without that monitoring limiter.

What LUFS should my mix have before the mastering process?

There is no LUFS target for a mix. Mixes that peak around -6 dBFS usually read somewhere between -25 and -19 LUFS integrated, depending on how dense the arrangement is. If you hit the peak target, the LUFS reading takes care of itself.

Streaming loudness targets such as -14 LUFS apply to the finished master. Platforms use them to play every track at a similar level, so listeners don’t have to reach for the volume between songs.

What is LUFS measure?

LUFS stands for Loudness Units relative to Full Scale. It measures perceived loudness: the meter weights frequencies the way the ear does, and integrated LUFS averages the whole song while ignoring silence.

You can also measure it across a whole album, which is how album normalization on streaming services works.

How do you create headroom in a mix?

Set it at the start. Many engineers build the mix from the drums, with the drum bus peaking around -12 to -9 dBFS. Bring everything else in around that, and the full mix usually lands close to -6 dBFS.

If the finished mix is too hot or too quiet, pull the master fader down or put a clean gain plugin first on the master bus. In a 32-bit float session, that changes nothing but the level.

Should I normalize my mix before mastering?

No. Normalizing usually pushes the peak to 0 dBFS, which removes the headroom you just built. Our mastering order form asks you not to normalize for that reason.

If the peaks are off, change the level with the master fader and bounce again.

Mixed it yourself and it’s ready? Our online mastering is $50 per song, 2–3 days per song, 6 revision rounds included. If the mix still needs work, mixing and mastering starts at $133 per song. Not sure the mix is ready? An engineer will give you free feedback on your mix before you pay for anything.