The 30-second version

• Headroom = the gap between your loudest sample peak and 0 dBFS
• Target headroom on your stereo bus: about -6 dBFS peak (so your loudest moment hits -6 dB, not -0.1 dB)
• What it costs you to skip this: mastering arrives with a mix that's already pushed up against the ceiling. The master loses dynamic range, the limiter has nothing to work with, and the final result is flatter than the song should be.

The longer version

When you mix, your stereo bus catches every sound the mix makes — every kick, every snare hit, every chorus swell. The peak meter on your stereo bus shows you the loudest moment of that mix as a sample peak level in dBFS (decibels relative to full scale, where 0 dBFS is the digital ceiling).

If your loudest moment hits -0.1 dBFS, your mix peaks at -0.1 dB. There's 0.1 dB of headroom — almost none. If your loudest moment hits -6 dBFS, your mix peaks at -6 dB. There's 6 dB of headroom — plenty for mastering to work with.

Why mastering needs headroom

Mastering's job is to take the finished mix and shape it for delivery. That shaping involves:

• EQ moves (broad tonal balance adjustments)
• Compression (dynamic shaping for cohesion)
• Saturation (harmonic enrichment for perceived loudness and warmth)
• Stereo widening (where appropriate)
• Limiting (final loudness push and peak control)

Most of those processes add some gain. EQ boosts add a couple dB in their boosted range. Compression and saturation often add gain after the dynamic reduction. The limiter pushes the master to its final loudness target.

All of those gains add up. If your mix arrives at -0.1 dBFS peak with no headroom, mastering has two bad options: (1) turn the mix down before processing (which is a workaround, not a fix), or (2) try to do all the shaping with negative gains everywhere (which is fighting the natural flow of the chain).

Neither option produces a master as good as the one mastering would have produced with proper headroom from the start.

The target — -6 dBFS peak

The widely accepted target for pre-master headroom is about -6 dBFS peak on your stereo bus. Specifically:

• Peak target: -6 dBFS (your loudest sample should hit -6 dB)
• Integrated LUFS: between -16 and -12 LUFS (most genres land in this range when peak is around -6)
• No brick-wall limiter on the mix bus (the limiter would be flattening peaks artificially)

Some engineers prefer -3 dBFS, some prefer -8 dBFS. -6 is the safe middle that works for almost every genre and every mastering chain.

Why this works mathematically

When mastering applies its processing chain, the cumulative gain through the chain typically adds 4–6 dB of "perceived loudness gain" to the master. If your mix arrives at -6 dBFS peak, the master after processing lands somewhere around -1 to -2 dBFS peak before the final limiter pushes it to the streaming-target true-peak ceiling (-1.0 dBTP for most streaming releases).

If your mix arrives at -0.1 dBFS peak, the master after the same processing would land at +5 dB or higher before the limiter — which means the limiter has to do 6 dB of gain reduction just to get back to the ceiling. That's way more limiting than the master needs, and it costs dynamics.

The -6 dBFS target leaves room for the processing chain to do its work naturally.

What headroom is NOT

A few common confusions worth clearing up.

• Headroom is not loudness. A mix with -6 dB headroom can still feel loud — the perceived loudness is set by the average level (LUFS), not the peak level. You can have a mix that peaks at -6 dB AND sits at -12 LUFS integrated, which is plenty loud for mixing reference purposes.
• Headroom is not bit depth. People sometimes conflate them. Bit depth determines dynamic range (the floor — how quiet the quietest meaningful sample can be vs the noise floor). Headroom is about the ceiling. They're different axes.
• Headroom on your master bus is different from headroom inside your mix. You can have channels inside the mix hitting +3 dB on their channel strips that don't show up at the stereo bus because of internal gain staging. The pre-master headroom that matters is the stereo bus output.

How to check your headroom before bouncing

Two paths:

1. Watch the stereo bus peak meter during a full playback of the song. Specifically during the loudest section. The peak meter shows the highest sample peak seen during playback. If it hits -0.5 dBFS during the loudest chorus, your headroom is 0.5 dB — not enough.

1. Use the free Mix Readiness Checklist — drop your bounced file, the tool reads the peak level and flags if you're too hot (above -3 dB) or too quiet (below -12 dB, which is fine but uncommon).

The fix when you don't have enough headroom

You bounce, check, and find your mix is peaking at -0.5 dB. The fix:

1. Check for a bus limiter or compressor doing aggressive gain reduction on your mix bus. If there's one, the easiest fix is to bypass it before bouncing.
2. Pull down the stereo bus output trim by however many dB you need to hit -6 dBFS peak. This is the simplest fix when there's no bus processing causing the issue.
3. Re-bounce and re-check. Confirm the new bounce peaks around -6 dBFS.

A few engineers prefer scaling the entire mix down by adjusting a VCA or group fader rather than just trimming the stereo bus output — same end result, different mechanic.

What if I'm mixing into a mastering plugin chain?

Some workflows have a tasteful mastering chain already on the stereo bus during the mix. That's fine for monitoring, but the bounce that ships for mastering should be from a state where the mastering chain is bypassed (or removed entirely) and only the mix processing remains. The bounce is the "pre-master" — it should look like a finished mix, not a finished master.

The bottom line

Leave 6 dB of headroom on your stereo bus before you bounce. Bypass any brick-wall limiter on the mix bus. Re-bounce, check the peak level in the free Mix Readiness tool, and ship the bounce to mastering knowing the engineer has room to do the actual work.

It's a small change with an outsized impact on the final master.