The word “pop” describes a sound that registers instantly and cleanly, without requiring the listener to lean in. In a platformer, a jump that pops is one the player feels at the moment of input, even when the music is loud and the screen is busy. That is not a property of a single synthesized sweep. It is the result of two layers doing two different jobs, aligned tightly enough that the ear hears them as one event.
The first layer is a transient: a very short, high-frequency event that marks the exact moment of departure. The second is a whoosh: a filtered noise layer that describes the motion after the departure. The transient answers “did it happen?” The whoosh answers “how did it move?” A jump that has only one of the two is either a click or a smear, and neither one pops.
The sections below work through the two layers separately, then through the alignment and frequency separation that let them coexist without masking each other. The parameters are drawn from documented sound-design practice and from the structure of the jump itself, not from a preference for a particular aesthetic.
The Transient: Under 80 ms, High-Pass at 200 Hz, Band-Pass Around 1–4 kHz
The transient is the part of the jump that confirms the input. Its job is not to carry pitch or duration. Its job is to produce a sharp, unambiguous onset that the ear can localize in time. A documented game-SFX layering guide assigns the transient a duration of 0–80 ms and a frequency range of 2–8 kHz, with the role described as “click, snap, high-frequency crack” . A separate source puts the transient band at 2–8 kHz and notes that a high-pass at 200 Hz keeps the body layer from masking the click on small speakers .
For a jump specifically, the transient does not need the full 2–8 kHz range. A narrower band-pass around 1–4 kHz is enough, and it has an advantage: it sits in the region where the ear is most sensitive and where the whoosh’s noise body is usually less dense. A laser transient is documented at 1–4 kHz precisely because that band “cuts through mix without masking dialog” . The jump transient can borrow the same band for the same reason.
The synthesis of the transient is simple. A short burst of white noise through a band-pass filter, with an envelope that has an instant attack and a decay under 50 ms, produces a usable click. A brief sine burst at a high frequency can substitute if the game’s aesthetic is less noisy, but the noise version is more robust: it reads as a contact event rather than as a tone. The key is that the transient must be shorter than the whoosh and higher in frequency, so the two do not compete for the same perceptual space.
The Whoosh: Filtered Noise, 80–200 ms, Centred Around the Midrange
The whoosh is the part of the jump that carries motion. It is not a pitch sweep in the same sense as the classic square-wave jump; it is a filtered noise layer whose spectral centre moves over the duration of the sound. The Krotos whoosh recipe describes the structure directly: “filtered white or pink noise for body, a pitched sweep (synth or stretched sample) for motion, a small transient hit at the end for punctuation” . For a jump, the order is reversed: the transient comes first, and the whoosh follows it.
The frequency range for the whoosh is the subject of some variation across sources, and the differences are useful. The Krotos recipe recommends a high-pass on the noise around 120 Hz and a gentle low-pass on the sweep so it sits under dialogue . A game-SFX formula puts the whoosh band at 500–2000 Hz with a duration of 80–200 ms . A broader whoosh reference suggests a mid-range boost around 1–3 kHz to help the whoosh cut through dialogue and music .
Those numbers point at a practical starting point: a band-pass centred around 800 Hz–2 kHz, with a moderate Q, and a sweep of the centre frequency upward over the whoosh’s duration. The upward sweep is what makes the noise read as motion rather than as a static hiss. A whoosh that does not move spectrally is a texture, not a gesture.
The duration is the other parameter that matters. The same formula puts the whoosh at 80–200 ms . A whoosh shorter than 80 ms tends to fuse with the transient and loses its separate identity. A whoosh longer than roughly 200 ms starts to overlap the character’s actual upward travel in a way that can feel disconnected from the input, especially in a game with short jumps. The duration benchmark guide covers the jump-type ranges; the whoosh layer sits inside the short-hop range, not the full-jump range.
The Offset: The Transient Leads, the Whoosh Follows, and the Gap Is Small
The single most important parameter for making the two layers read as one event is the time offset between them. If the transient and the whoosh start at exactly the same sample, the transient is partially masked by the whoosh’s noise, and the jump loses its click. If the whoosh starts too late, the jump sounds like two events: a click, a pause, and then a whoosh.
A documented wall-jump structure provides a useful reference. The wall-jump sound is described as an impact (3 transients) plus a short whoosh, and the whoosh is described as playing underneath the impact, not after it . That is a different design from the jump, but it establishes the principle: the transient carries the contact, and the whoosh is a supporting layer that begins at or near the same moment.
For a jump, the working relationship is that the transient’s onset is at 0 ms and the whoosh’s onset is within roughly 5–20 ms. The transient is so short (under 80 ms) that the whoosh can begin while the transient is still audible, as long as the whoosh’s initial amplitude is low enough that it does not cover the click. The ear fuses the two because they share an onset envelope, and the whoosh’s spectral movement then carries the sense of motion after the click has passed.
The jump versus landing guide covers the time-domain difference between departure and arrival events; the jump’s two-layer structure is a specific case where the offset between layers is doing perceptual work that a single-layer sound cannot do.
Frequency Separation: The Two Layers Should Not Share the Same Band
The transient and the whoosh will mask each other if they occupy the same frequency region. The transient is high-frequency and short; the whoosh is mid-frequency and longer. The separation is what allows both to be heard.
A practical allocation is: the transient in 2–6 kHz, the whoosh body in 500 Hz–2 kHz, and the low end (below 200 Hz) left mostly empty for the jump’s pitch layer if one is used. The transient’s high-pass at 200 Hz is the standard safeguard: it prevents the whoosh’s low-mid energy from masking the click on a small speaker . A gentle low-pass on the whoosh around 3–4 kHz keeps it from competing with the transient’s upper harmonics.
The over-processing guide covers the phase and masking checks that catch a layer contributing nothing. The transient and whoosh pair is a case where the two layers should pass the polarity test easily: if inverting the whoosh makes the transient clearer, the whoosh was partially cancelling it, and the frequency separation is wrong.
A Working Starting Point for the Two Layers
The table below gives a first-pass configuration for the transient and the whoosh. The values are starting points, not a specification, and they are chosen so that the two layers have a clear separation in time and frequency.
| Property | Transient | Whoosh |
|---|---|---|
| Role | Marks the moment of departure | Describes the motion after departure |
| Duration | Under 80 ms; often 20–50 ms | 80–200 ms |
| Frequency band | 2–6 kHz, high-pass at 200 Hz | 500 Hz–2 kHz, low-pass around 3–4 kHz |
| Onset | 0 ms | 5–20 ms after the transient |
| Envelope | Instant attack, fast decay | Short attack, longer decay; spectral centre sweeps upward |
The configuration is not a recipe for a specific aesthetic. It is a structure that can be tuned toward a retro square-wave jump, a noisy arcade jump, or a softer characterful jump by changing the whoosh’s filter centre and the transient’s noise colour. The square wave and pitch sweep guide covers the pitched layer that can sit underneath this pair; the transient-and-whoosh structure is the part that makes the jump read as an event before the pitch has finished moving.
A jump that pops is not a louder jump. It is a jump where the click and the motion are both present, separated in frequency and aligned in time. The transient confirms the input; the whoosh carries the movement; and the small gap between them is what lets the ear hear both.