Aim for stereo mix peaks between −3 dBFS and −6 dBFS, because that gives a mastering engineer room to work without turning your file down first. The caveat matters just as much: deliver an unclipped, un-limited mix, and check true peak on export so nothing overshoots once it is converted or encoded. Everything below shows you how.
TL;DR:
- Set loudest sources, often the kick or vocal, near −12 dBFS and keep the mix bus around −10 dBFS during mixing to prevent summing problems.
- Use clip gain before compressors or saturators to control the level entering each effect, rather than relying on makeup gain after heavy processing.
- A mix bus limiter can hide balance problems and smear transients, so bypass it for export and use it only to preview loudness.
- Deliver a stereo WAV or AIFF at the session’s sample rate, and lower reference tracks to your mix level for fair tonal comparisons.
- Lowering the master fader can correct output clipping, but it cannot repair distortion already created on individual tracks or buses; fix those sources directly.
Table of Contents
- What headroom is and which meters to trust
- How much headroom to leave: the rule of thumb and the mastering nuance
- How to create and check headroom in your DAW
- Common mistakes that reduce headroom and how to fix them
- Preparing files for mastering: export settings and deliverables
- Standards and evidence behind these recommendations
- A quick pre-send checklist
- Why we teach these checks at AubioMix
- Get a second opinion on your headroom before mastering
- FAQ
- Sources
What headroom is and which meters to trust
Headroom is the gap between the loudest peak in your mix and 0 dBFS, the digital ceiling. If your highest peak sits at −5 dBFS, you have 5 dB of headroom. It sounds simple, but a few different meters can report quite different things about the same file, so it helps to know what each one tells you.
- Sample peak: the highest individual sample value in the file. It is what most DAW meters show by default, and it is fast and useful for a first check.
- True peak (dBTP): an estimate of the peak level of the signal once it is converted back to analogue or encoded, which can be higher than any single sample.
- LUFS: a loudness unit based on K-weighting. The ITU-R BS.1770 family and AES loudness materials describe momentary, short-term and integrated measurements, and each tells you something different about how loud a track feels.
- Crest factor: the difference between peak level and average level. A high crest factor means punchy, dynamic material, and a low one means a squashed mix.
Inter-sample peaks are the reason true-peak metering exists. Standard sample-peak meters are blind to them, so audio that appears to peak below 0 dBFS can still produce inter-sample peaks above +1.0 dBTP. That overshoot can cause problems when a file is converted to MP3 or AAC. If you want quick definitions of any of these terms, our audio metrics glossary covers LUFS, true peak, crest factor and LRA in plain language, and our guide to LUFS for producers goes deeper on loudness.
One more thing worth knowing: modern DAWs run on 32-bit floating-point engines, which means there is a lot of internal headroom. That does not mean you can ignore levels, but it does mean that pulling down the master fader can remove output clipping without harming the audio, provided your individual tracks were gain-staged sensibly.
How much headroom to leave: the rule of thumb and the mastering nuance
The number most producers hear is −3 to −6 dBFS on the stereo master fader. Practitioners at SoundCloud's help centre give the same guidance: keep your loudest peaks in that window and deliver an unclipped file for mastering. The reason is practical. A mastering engineer will typically want to apply EQ, compression and limiting, and a file that already touches the ceiling forces them to turn it down before they can do anything useful.
A commonly recommended target for mix bus peaks is between −3 dBFS and −6 dBFS, which gives mastering engineers a safe working buffer to apply EQ, compression and limiting without first reducing the file's gain.
The same source warns against sending files that peak at −0.1 dBFS, because that leaves no room for mastering-grade processing at all. Think of it like handing someone a suitcase that is already zipped to bursting and asking them to add a coat.
Now for the nuance, because the range is a rule of thumb, not a law. Mastering engineers generally care less about an exact peak number and more about receiving a file that has not been clipped or squashed by a limiter. A mix with healthy dynamics is usually easier to master than one that has been flattened on the mix bus. So the number is a safety margin, and the real goal is preserved dynamic range. If your mix has a few stray transients that poke above −3, that is rarely a disaster. If the whole mix has been pushed into a limiter to hit a loudness target, that is a much bigger problem.
Plenty of mastering engineers will also apply gain right at the start of their chain, which is why a lower peak level costs you nothing. Turning a quieter file up is lossless in a floating-point workflow, while turning a file down cannot undo clipping damage. That asymmetry is the whole argument for leaving space.
There is also the distribution side to think about. Streaming and download services re-encode your master into lossy formats, and that process can add overshoots. Guidance for internet audio streaming recommends that maximum true peak should not exceed −1 dBTP at the codec input of lossy-encoded streams. Your mix is not the final master, so you will not be sitting at −1 dBTP at this stage. But it shows why a true-peak safety margin matters, and why mastering engineers like to receive files that are comfortably below the ceiling. For more on how this plays out on specific platforms, see our piece on streaming loudness targets.
Finally, resist the temptation to treat headroom as a loudness decision. Leaving headroom does not make your mix quiet in any way that matters, because loudness is something that gets added at the mastering stage. Your job at the mix stage is to make the balance, tone and dynamics right, then hand over a file with room to breathe. A mix that sounds good at a lower level will sound good once it is turned up, and a mix that only sounds good when it is loud is usually hiding something.
How to create and check headroom in your DAW
Creating headroom is mostly about decisions you make early, not fixes you apply at the end. Work through these steps in order and you will rarely need to rescue a mix at the last minute.
- Gain stage from the start. Set conservative peaks on your loudest element, usually the kick or the vocal, and build everything else around it. Gain staging guidance suggests channel peaks around −12 dBFS as a common reference point, and keeping mix bus peaks around −10 dBFS during the mix avoids summing problems and leaves room for processing.
- Use clip gain before inserts. Adjust region or clip gain to set the level going into a compressor or saturator, rather than adding make-up gain after heavy processing. This keeps each stage from pushing the next one too hot.
- Do not mix through a limiter on the mix bus. Professional advice is to avoid mastering-style limiters while mixing, because they can mask technical issues and cause transient smearing. If you want to preview how your mix feels at loudness, do it, but bypass the limiter before you bounce.
- High-pass where it makes sense. Rolling off inaudible low-end rumble on tracks that do not need it frees up energy that would otherwise eat into your headroom. Be gentle, and use your ears so you do not thin out the elements that should carry weight.
- Meter after the mix bus. Place your LUFS and true-peak meters on the master output, and confirm the values before you export. Check them on the loudest section of the song, not just the intro.
- Pull the master fader down if needed. If everything sounds right but the peaks are too hot, lower the master fader rather than rebalancing the whole mix.
Pro Tip: Loop the loudest chorus while you watch the true-peak meter. Peaks that only show up when the whole arrangement is playing are the ones that catch you out.
Our walkthrough on metering for streaming shows how we set up these checks in a typical producer workflow, and it is worth a look if you have never run a true-peak meter before.
A word on fader strategy, since it trips people up. A common mistake is to start with every fader at unity and then wonder why the mix bus is slamming into the ceiling. Starting lower costs you nothing. If you set your drums and bass at modest levels first, the rest of the arrangement falls into place around them, and you rarely have to claw back level later. The same logic applies to busses: if a drum bus is already hot before it reaches the mix bus, no amount of master-fader trimming will make the processing in between sound better.
It also helps to remember that analogue-style plugins respond to the level you feed them. Many emulations are designed around a specific operating level, so feeding them very hot signals can make them distort in ways that were never part of your plan. Gain staging properly means those plugins behave the way they were designed to, which is another quiet benefit of leaving space.
Common mistakes that reduce headroom and how to fix them
Most headroom problems come from a handful of habits. The good news is that each one has a quick remedy.
- Mix bus limiting that hides problems. A limiter on the master can make a harsh or unbalanced mix feel controlled, but the issues are still there. Remove it, fix the offending sources, and bring it back only for previewing.
- Comparing against mastered references at full level. A finished commercial track is already loud, so your quieter mix will sound weak next to it. Lower the reference until it matches your mix level, then compare tone and balance fairly.
- Exporting without a true-peak check. A sample-peak meter can say −1 dBFS while the true peak sits above 0. Use a true-peak meter, and if you find overshoots, pull the master fader down a little and bounce again. Our article on inter-sample peaks explains why a −1 dBTP margin is worth keeping.
- Trusting meters over ears. Numbers tell you whether a file is safe, but not whether it sounds good. Listen for transient clarity and dynamics, because a mix can pass every meter and still feel flat.
- Fixing everything with the master fader. Pulling the fader down is fine for output clipping, but it will not rescue a mix where individual channels are already distorting. Go back to gain staging at track level.
If you spot yourself doing any of these, do not panic. They are among the most common reasons mixes arrive at mastering with too little room, and each one is straightforward to correct once you know where to look.
Preparing files for mastering: export settings and deliverables
A clean handover saves everyone time. Here is what we suggest you prepare before you send a mix anywhere, whether to a mastering engineer or to an online analyser.
- Export a stereo WAV or AIFF. Use the same sample rate as your session and an appropriate bit depth supported by your DAW. Avoid MP3 and other lossy formats for the delivery file.
- Keep true peak safely below the ceiling. Aim for your −3 to −6 dBFS peak range, and confirm with a true-peak meter that nothing overshoots. Bear in mind the distribution guidance of −1 dBTP at the codec input, and leave comfortable room beneath it.
- Turn off mix bus limiting and dither. Leave the file unclipped and un-limited. Mastering will handle final dither when the bit depth is reduced.
- Include your references, lowered to mix level. If you want the mastering engineer to hear what you are aiming for, supply the tracks, but match their level to your mix so the comparison is honest.
- Add short notes. Mention your intent, any problem sections, and anything you would like left alone. A few lines save a lot of back-and-forth.
Standards and evidence behind these recommendations
None of this is guesswork. The loudness and peak guidance comes from a small group of standards bodies and long-standing practice in the mastering world.
Distribution guidance recommends that maximum true peak should not exceed −1 dBTP at the codec input of lossy-encoded streams.
That recommendation comes from the AES technical document on streaming loudness, which also notes that distribution loudness tolerances shape the peak and loudness choices for different genres. Alongside it, the ITU-R BS.1770 family and related AES loudness material define K-weighted loudness measurement, and explain why momentary, short-term and integrated readings each have a role in characterising content for distribution.
The practitioner side lines up with the standards. Sound On Sound's engineering advice points the same way on limiters, and the inter-sample peak concern is well documented in loudness-normalisation references. Put together, they explain why a peak window of −3 to −6 dBFS, no mix bus limiting and a true-peak check are such consistent advice across the industry.
A quick pre-send checklist
Run through these checks before you send a mix off.
- Peak headroom sits between −3 and −6 dBFS.
- True peak is below your ceiling, with no inter-sample overshoots.
- Integrated LUFS is recorded for your own reference.
- No limiter is active on the mix bus.
- Reference tracks are lowered to match your mix level.
- Mono and phase check done, with nothing collapsing.
- File is a stereo WAV or AIFF at the right bit depth.
- Short mastering notes are written.
| Check | Target |
|---|---|
| Sample peak | −3 to −6 dBFS |
| True peak | Below −1 dBTP at the codec input, with a safe margin |
| Mix bus limiter | Bypassed or removed |
| File format | Stereo WAV or AIFF |
If you find clipping or an inter-sample overshoot, lower the master fader, re-check the meter, and bounce again. If the clipping is coming from a specific track or bus, go back and fix it there rather than hiding it at the end.
Why we teach these checks at AubioMix
At AubioMix, we help producers prepare their mixes for mastering. Most of the time, the issue was not talent but a missing habit, such as checking true peak or removing a mix bus limiter before bouncing.
Our view is that headroom is a symptom of a good workflow, not a number to chase. If your gain staging is sensible, your sources are clean and you are not leaning on a limiter, the right peak level tends to fall out naturally. That is why we pair measured benchmarks with practical, plain-language checks rather than a single magic figure. The measurements tell you where you are, and the steps tell you what to do about it.
We would encourage you to test this yourself. Take a recent mix, run it through the checks above, and see how it compares with published benchmarks for your genre. You may find that a mix you thought was finished has more room for improvement than you expected, and that is a good discovery to make before mastering, not after.
— AubioMix
Get a second opinion on your headroom before mastering
If you want an expert check before you hand a mix over, we offer a 1 Mix Review for £4.99. You upload your file and receive visual measurements and written, actionable steps across mixing and mastering prep, including loudness, clipping, phase, stereo width and how your mix translates across playback devices. It is a quick way to confirm your peaks and true-peak levels are safe, and to spot anything that would trip up a mastering engineer.
You can see every option on our pricing page, including the 1 Reference Pass for £2.99 and our monthly Professional plan at £19.99 per month. When you are ready, upload a mix and see what the report finds before you send it on.
FAQ
How much headroom should my mix have?
Aim for stereo master peaks between −3 dBFS and −6 dBFS. That window gives a mastering engineer room to process without first turning your file down, as practitioner guidance on mastering delivery recommends.
Is it better to mix quiet or loud?
Mix at a moderate level and leave the loudness to mastering. A mix that peaks lower but keeps its dynamics is easier to master than one squashed by a limiter. You can always turn a clean file up, but you cannot undo clipping.
What is the 1 dB rule in mixing music?
There is no single agreed rule by that name, so treat any claim of one with caution. A related, well-documented idea is to keep a true-peak margin, with distribution guidance recommending no more than −1 dBTP at the codec input of lossy-encoded streams.
Do I need a true-peak meter for headroom?
Yes, if you are preparing files for release or mastering. Sample-peak meters can miss inter-sample peaks, so a true-peak meter is the safer way to confirm nothing overshoots once your audio is converted or encoded.