The question sounds like a mixing preference, but it is actually a scheduling problem. A sidechain duck is a temporary reduction in the music’s level, triggered by another signal and released when that signal stops. That works well for dialogue, which arrives in long bursts and leaves gaps. It works badly for a jump, which fires hundreds of times in a session and is often only 100–150 ms long.
A full-band duck on every jump produces a music bed that pumps continuously, because the release from the previous jump has not finished before the next one arrives. The jump becomes audible, but the soundtrack becomes unstable. The alternative is not to abandon ducking. It is to decide which part of the music needs to move, and to use the priority hierarchy and the frequency band to make that decision before touching a compressor.
The sections below work through three decisions in order: whether the jump should be allowed to move the music at all, whether the movement should be full-band or frequency-specific, and what the sidechain parameters should be if a duck is used.
First Decision: Does the Jump Belong Above the Music in the Priority Hierarchy?
Ducking is a mechanism that enforces a priority decision. It should not be used to discover one. The priority hierarchy has to be set before any sidechain is configured, because the duck is only doing its job if it is moving the music for a sound that the design has already decided is more important.
The documented game audio hierarchy puts voice on top, player SFX below it, enemy SFX below that, music below both, and ambience at the bottom . The jump is a player SFX, which means it sits above music and below dialogue. That ordering is the permission for the duck: the jump is allowed to move the music, and the music is not allowed to bury the jump.
The failure mode is a hierarchy where music and SFX are set at the same level because both “sound good” in isolation. That is not a hierarchy; it is a tie, and a tie is resolved by frequency masking rather than by design. The audio mix hierarchy guide covers the bus structure that makes the hierarchy survive into the game; the jump-specific point is that the jump belongs above the music on the bus tree, not beside it.
Once the hierarchy is set, the question is not “should the jump duck the music” but “how much of the music should move.” A jump that fires every few seconds does not need the same depth of duck as a dialogue line that fires once a minute. The jump mix guide covers the static-carve alternative for cases where a dynamic duck would pump; the sidechain question is specifically about when the music should move during the jump.
Second Decision: Full-Band Duck or Frequency-Specific Duck?
The default sidechain duck is full-band: the entire music bus drops by a fixed amount for the duration of the trigger. For dialogue, that is usually the right choice, because dialogue occupies a broad midrange band and the music needs to move out of the way across that whole range.
For a jump, full-band ducking is usually the wrong choice, because the jump does not occupy the whole spectrum. A jump’s identity sits in a relatively narrow band — typically the attack and upper harmonics in the 2–5 kHz region, with the body below that . If the music is ducked across the entire spectrum, the low end and the high end move even though the jump is not competing with them. The result is a music bed that feels like it is breathing on every jump, rather than a mix where the jump has a place.
The frequency-specific alternative is to duck only the band the jump occupies. A dynamic EQ or a multiband sidechain compressor can reduce the music by a few dB in the 2–5 kHz range while leaving the low end and the high end untouched. The jump becomes audible because the band it needs is opened, and the music does not lose its weight or its air.
The documented frequency-specific approach for game audio is explicitly framed this way: “reduce mid-range where voices live, preserve bass/treble” . The same logic applies to a jump, with the band shifted up slightly. A multiband ducker that operates on three bands — low, mid, high — can be configured so that only the mid band responds to the jump, while the low and high bands pass through unaffected.
The trade-off is complexity. A full-band duck is one compressor and one sidechain send. A frequency-specific duck is a multiband processor or a dynamic EQ, and it has to be tuned so that the band it affects actually matches the jump’s identity band. The frequency range guide covers the band placement in more detail; the sidechain point is that the duck’s band and the jump’s band have to be the same band, or the duck is moving music the jump does not need moved.
Third Decision: If a Duck Is Used, What Should the Parameters Be?
The parameters that matter are attack, release, depth, and threshold. The documented ranges for game audio ducking are broader than what a jump needs, because most guidance is written for dialogue. The table below gives the general ranges and the jump-specific adjustments.
| Parameter | General game ducking range | Jump-specific adjustment |
|---|---|---|
| Attack | 0.1–0.3 s for dialogue; 1–10 ms for instant ducking[reference:1] | 5–20 ms: fast enough to engage before the jump’s transient, short enough not to smear it |
| Release | 0.3–0.8 s for dialogue; 50–200 ms for fast recovery[reference:3] | 50–150 ms: recovers before the next jump, so the duck does not accumulate into a permanent dip |
| Depth | −12 dB typical for dialogue; −6 to −12 dB gentle dip[reference:5] | 3–6 dB: enough to open the band without the music sounding muted |
| Ratio | 4:1 typical; 10:1+ produces an obvious dip[reference:6] | 2:1–4:1: a softer ratio that tracks the jump’s level without hard ducking |
| Threshold | −30 dBFS default; −40 dBFS triggers easily[reference:7] | Set so the jump’s transient crosses it but the music alone does not; typically −20 to −30 dBFS |
The parameter that is easiest to get wrong is the release. A release of 300 ms is appropriate for dialogue, because dialogue lines are separated by pauses. A jump fires every few hundred milliseconds during movement, and a 300 ms release means the music never fully returns before the next duck. The result is a music bed that is permanently 3–6 dB down during platforming sections, which is the opposite of the intended effect: the jump is audible, but the music has lost its energy for the whole section.
The jump-specific release should be short enough that the music returns between jumps. A 50–150 ms release, combined with a 5–20 ms attack, keeps the duck tied to the individual jump rather than to the section. The ducking and sidechain guide covers the full parameter set; the jump-specific point is that the timing is set by the jump’s repetition rate, not by the music’s phrasing.
The Alternative That Avoids Ducking Entirely: Give the Jump a Band the Music Has Left Open
There is a solution that does not require a sidechain at all, and for some platformers it is the better one. If the jump’s identity band is chosen so that it sits in a region the music is not using, the jump cuts through without the music moving.
The documented version of this approach is to create a hole in the frequency spectrum of the music so the effect can be heard . That is a static EQ cut, not a dynamic duck, and it has the advantage of being predictable: the music sounds the same whether the jump is playing or not, and the jump has a permanent place in the spectrum.
The disadvantage is that it costs music energy all the time, not just during the jump. For a platformer where the jump fires constantly, that cost is usually acceptable, because the jump and the music are both always present. For a game where the jump is rare and the music is the main event, a static carve is the wrong tool.
The sound effects versus music guide covers the static-carve versus dynamic-duck trade-off in more detail. The jump-specific decision is that a static carve is the better default for a high-frequency event, because it avoids the pumping that a dynamic duck produces when the trigger is too frequent.
The Check: Does the Music Sound Stable Across a Platforming Section?
The test for a jump duck is not whether the jump is audible. It is whether the music still sounds like music after thirty seconds of continuous jumping. A duck that is working correctly is inaudible as a duck: the jump is present, and the music does not appear to be moving.
The specific failure to listen for is a music bed that seems to “breathe” with the jump rhythm. That is a duck that is too deep, too slow to release, or too broad in frequency. The fix is not to remove the duck; it is to narrow its band, shorten its release, and reduce its depth until the music stops moving audibly.
If the music still sounds unstable after those adjustments, the correct answer may be that the jump should not duck the music at all. A jump that is placed in an open band does not need a duck, and a music bed that never moves is a better outcome than a jump that is audible at the cost of a pumping soundtrack.