The principle

Mix-bus processing that shapes the character of the mix stays. Mix-bus processing that flattens dynamics or pushes loudness comes off.

The character moves (bus comp, bus EQ, bus saturation, parallel-comp blends) are part of how the mix sounds. They're decisions you made during the mix. Mastering inherits them and works with them.

The loudness moves (brick-wall limiter, aggressive maximizer, anything doing more than a couple dB of gain reduction in service of pushing the mix louder) overlap with what mastering does. When they're on the bus AND mastering applies its own limiter, the two compete and the result is over-processed.

What to leave on the mix bus

Bus compression (1-3 dB gain reduction max)

Light bus compression — a gentle slow-attack/medium-release move averaging 1-3 dB of gain reduction on the loudest sections — is glue. It helps the mix feel cohesive, smooths the relationship between instruments, and gives the bus a slight "settled" quality.

Mastering will often add its own bus compression (or multiband dynamics) on top of yours, but that's fine — yours is doing a different job. Yours is glue; mastering's is loudness and translation.

Keep on the bus.

Bus EQ (broad tonal moves)

A wide low shelf, a broad mid scoop, a high air boost — these are character moves that shape the mix's tonal signature. They're part of why the mix sounds like the mix.

Mastering may make additional tonal moves, but those will be broader still (full-spectrum balance) and informed by what the mix is doing. Leaving the bus EQ in place tells mastering "this is the tonal character I want." Mastering shapes from there.

Keep on the bus.

Bus saturation / harmonic enhancement

Tape emulation, tube saturation, console saturation — these add harmonic complexity that contributes to the mix's perceived warmth, weight, and density. Most modern mix-bus chains include some saturation, and most modern mastering chains include additional saturation tuned for loudness preservation through codec encoding.

The two layers work together. Leave the bus saturation in.

Keep on the bus.

Stereo widening or imaging plugins

If your mix-bus widener is doing tasteful work (not over-widening, not introducing phase issues), leave it. It's part of the mix's stereo character. Mastering may make additional imaging moves but won't compete with reasonable bus widening.

Keep on the bus — but check mono compatibility before bouncing.

(The free Mono Compatibility Checker verifies your widening hasn't introduced fold-down issues. If it has, dial the widener back before bouncing.)

Parallel compression on the bus

If you've routed a parallel compression bus that blends back into the mix, the parallel blend is part of the mix. Leave it.

Keep on the bus.

What to take OFF the mix bus

Brick-wall limiter (the most common culprit)

The brick-wall limiter on the mix bus is the single most common pre-master issue I see. The reasoning is usually: "I want to hear what it sounds like at streaming loudness while I mix." Fine — leave the limiter on while you mix, take it off when you bounce for mastering.

The limiter doing 3-6 dB of gain reduction is replacing the work mastering will do. The bounce that arrives for mastering with the limiter still on is a pre-mastered file, and the master that comes back can only build on what's already been done. Less to work with means a smaller master.

Take off before bouncing.

Aggressive maximizer / "loudness" plugin

Same logic as the limiter. Any plugin whose primary job is "make this louder" — Ozone Maximizer in aggressive mode, Pro-L 2 with significant gain reduction, Newfangled Elevate doing the same — overlaps with mastering's job.

Take off before bouncing.

Multiband limiting / multiband maximization

Multiband dynamics that are doing surgical work (controlling a specific frequency range that misbehaves on certain instruments) can stay. Multiband dynamics that are doing "louder, please" work on the master bus should come off.

If you're unsure: a multiband plugin doing under 2 dB of gain reduction per band, with the intent to fix a tonal problem, is fine. A multiband plugin doing 4-6 dB of gain reduction across multiple bands to add density is mastering work.

Take off if it's doing loudness work; keep if it's doing surgical fixing.

True-peak limiting set for delivery

Some mix-bus chains include a true-peak limiter as the final element, set to a streaming-target ceiling (e.g., -1.0 dBTP). If that's on your bus, take it off — true-peak limiting is the last step of mastering, not the last step of mixing.

Take off before bouncing.

Any plugin that's there "for streaming"

If you added a plugin to your mix bus because it's the "streaming master" emulation, or it's a one-knob "make it streaming-ready" tool — take it off. That work belongs to mastering and your mastering chain will be more nuanced than any one-knob plugin.

Take off before bouncing.

The fast checklist

Before you bounce for mastering, your mix bus should have:

• ✓ Bus compression doing 1-3 dB gain reduction (glue compression)
• ✓ Bus EQ doing broad tonal shaping
• ✓ Bus saturation (tape, tube, console)
• ✓ Parallel compression blends
• ✓ Stereo imaging plugins (verified mono-compatible)

And should NOT have:

• ✗ Brick-wall limiter
• ✗ Aggressive maximizer
• ✗ "Loudness" plugins of any kind
• ✗ Multiband dynamics doing loudness work
• ✗ True-peak limiter set for delivery
• ✗ Any "streaming master" emulation plugin

How to verify before sending

Bounce the mix. Drop the bounced file into the free Mix Readiness Checklist. The tool flags:

• Bus-limiter signature (catches if the bounce has been brick-walled)
• True peak (catches if you've pre-limited for streaming delivery)
• Peak headroom (confirms you have the -6 dBFS headroom mastering needs)
• Channel balance (catches accidental mono summing)
• File format compatibility

If those all pass and you bypassed the loudness plugins per the checklist above, your bounce is ready.

What changes when you do this right

A mastering engineer working from a bounce with proper bus character processing AND proper headroom can:

• Apply broader tonal moves without fighting pre-applied EQ
• Set their own limiter chain optimally for the song's character
• Make the loudness decisions in service of the song instead of working around pre-applied loudness
• Deliver a master that's more cohesive and translates better across systems

The difference is audible. Compare two masters of the same song — one mastered from a clean bounce, one mastered from a brick-walled pre-master. The clean-bounce master has more depth, more dynamic motion, more "the song breathing." The brick-walled bounce master sounds smaller no matter what mastering does.