Wake — an audio-reactive multi-effect plugin

[noahrbc]

Hi all,

I’m excited to release my first plugin: Wake, an audio-reactive multi-effect.

https://reactive-audio.netlify.app/

Most “shape” plugins (ShaperBox, LFOTool, Glitch, etc.) draw their movement against the tempo grid. Wake is different: it listens to the audio itself and fires its effects on every detected transient. So instead of fighting loose, human, or live-feeling material, it moves with the performance.

For each detected hit, Wake fires two drawable-shape effects:

• Reverse Swell — a reverse-reverb swell that builds into the hit and ends precisely on it (pre-echo via look-ahead). Drawable shape, length, reverb character, mix.

• Volume Shape — a drawable gain curve fired at the transient: ducks, pumps, tremolo, stutters. Sculpt the hit itself.

What makes it more than a one-trick effect:

• Per-onset Dynamics — Wake measures each hit’s loudness and brightness, so louder hits swell bigger and brighter hits stay shorter. The response tracks the actual performance, hit by hit — something grid-based tools structurally can’t do.

• Multiband targeting — each module acts on its own frequency band. Sidechain just the sub on every kick while the cymbals stay crisp, or send airy highs-only risers, all from one instance.

• Surgical triggering — frequency-range trigger filter (fire on kicks only, snares only…), plus probability and humanization for an organic feel.

Onset-activity visualizer, 8 factory presets, full preset recall, global bypass.

It’s a mix-time tool (it uses look-ahead, so it’s not for live tracking — your DAW’s delay compensation handles the latency for mixing).

Formats: VST3 + AU (macOS), plus a Standalone app. macOS build is universal and notarized. Windows (VST3) is coming soon — Mac first.

Price: $20 launch (first 60 days, then $30).

It’s my first release, so I’d genuinely love feedback — on the sound, the workflow, anything. Happy to answer questions in the thread.

Thanks,
Noah / Reactive Audio

[SNBeatz]

Ok I don't mean to be a jerk, but is Shaperbox not able to do this now with the semi-recent implementation of audio detection mode and audio rate/free-hertz based grid? I'm all for devs trying out fun stuff, but that bit in the blurb seems a bit off.

[noahrbc]

Ok I don't mean to be a jerk, but is Shaperbox not able to do this now with the semi-recent implementation of audio detection mode and audio rate/free-hertz based grid? I'm all for devs trying out fun stuff, but that bit in the blurb seems a bit off.

Not a jerk at all — you're right, and thanks for the push. ShaperBox 3's audio-trigger mode does cover transient-triggered shaping (threshold + freq-range detection included), so the "every shaper is tempo-locked" line in my blurb is outdated — that's on me and I'll fix it.

Where Wake actually differs is the reverse swell: for each detected hit it generates a reverse-reverb swell that builds into the transient and resolves on it (a pre-echo via look-ahead) — that's a synthesized effect, not an envelope reshaping the existing audio, and it's the part ShaperBox doesn't have an equivalent for. It also scales each swell by the individual hit's loudness and brightness. The volume-shaping side does overlap with ShaperBox's audio mode — fair point.

If you've got ShaperBox, the honest A/B is the reverse-swell on a drum or vocal — that's the genuinely different bit. Appreciate you keeping me honest.

[SNBeatz]

No worries, honestly not anywhere close to as "bad" as some of the marketing speak that is coming from a ton of devs and companies these days; I respect the honesty and owning it.

As I was reading further through the rest of your description, the brightness and loudness descriptors were piquing my interest for sure; these sorts of things always feel more human than trying to pencil what we, as fallible humans, think sounds natural. And, as far as I'm aware, you're absolutely correct about the reverse swells being unique to what you have here; you can kinda rig the convolution reverb to sorta kinda try to do something similar in Shaperbox but from what you're saying, your product should do it much more organically.

Tangentially speaking on the topic of unique forms of triggering, I think it would be extremely awesome if anyone (you included!) could figure out some sort of knee or hysteresis modifier. Essentially morphing a "standard" transient trigger into a sort of dynamic/interpolated trigger that would theoretically max out the desired intensity of whatever the effect may be (reverb swell, filter cutoff+resonance, etc.) AT the transient peak that a standard detector would pick up on. I don't know if I sound crazy right now or if this already exists in some form haha.

Sorry again for the sideline, but based on how you're describing Wake, it sounds like something you might already be working towards or at least be interested in!

[tumface]

Ok I don't mean to be a jerk, but is Shaperbox not able to do this now with the semi-recent implementation of audio detection mode and audio rate/free-hertz based grid? I'm all for devs trying out fun stuff, but that bit in the blurb seems a bit off.

Not a jerk at all — you're right, and thanks for the push. ShaperBox 3's audio-trigger mode does cover transient-triggered shaping (threshold + freq-range detection included), so the "every shaper is tempo-locked" line in my blurb is outdated — that's on me and I'll fix it.

Where Wake actually differs is the reverse swell: for each detected hit it generates a reverse-reverb swell that builds into the transient and resolves on it (a pre-echo via look-ahead) — that's a synthesized effect, not an envelope reshaping the existing audio, and it's the part ShaperBox doesn't have an equivalent for. It also scales each swell by the individual hit's loudness and brightness. The volume-shaping side does overlap with ShaperBox's audio mode — fair point.

If you've got ShaperBox, the honest A/B is the reverse-swell on a drum or vocal — that's the genuinely different bit. Appreciate you keeping me honest.

Cableguys SnapBack does the reverse lookahead trigger thing (edit: but it doesn't generate it from reverb, it does it via samples.) Also your post clocks a 100% score for being written by AI. T̶h̶e̶r̶e̶'̶s̶ ̶a̶ ̶d̶e̶d̶i̶c̶a̶t̶e̶d̶ ̶A̶I̶ ̶s̶u̶b̶s̶e̶c̶t̶i̶o̶n̶ ̶i̶n̶ ̶K̶V̶R̶.̶

[Gamma-UT]

Also your post clocks a 100% score for being written by AI. There's a dedicated AI subsection in KVR.

That isn't what the AI subforum is for and you know it.

[tumface]

That isn't what the AI subforum is for and you know it.

I actually didn't know that, but having read the description of the subform again, I now realize it's only for discussion of AI stuff and not for (potentially?) AI generated stuff.

[noahrbc]

No worries, honestly not anywhere close to as "bad" as some of the marketing speak that is coming from a ton of devs and companies these days; I respect the honesty and owning it.

As I was reading further through the rest of your description, the brightness and loudness descriptors were piquing my interest for sure; these sorts of things always feel more human than trying to pencil what we, as fallible humans, think sounds natural. And, as far as I'm aware, you're absolutely correct about the reverse swells being unique to what you have here; you can kinda rig the convolution reverb to sorta kinda try to do something similar in Shaperbox but from what you're saying, your product should do it much more organically.

Tangentially speaking on the topic of unique forms of triggering, I think it would be extremely awesome if anyone (you included!) could figure out some sort of knee or hysteresis modifier. Essentially morphing a "standard" transient trigger into a sort of dynamic/interpolated trigger that would theoretically max out the desired intensity of whatever the effect may be (reverb swell, filter cutoff+resonance, etc.) AT the transient peak that a standard detector would pick up on. I don't know if I sound crazy right now or if this already exists in some form haha.

Sorry again for the sideline, but based on how you're describing Wake, it sounds like something you might already be working towards or at least be interested in!

Thanks! That means a lot, and no need to apologize, this is the kind of tangent I'm happy to get lost in.

You're tracking right with where my head's at. The brightness/loudness stuff is the part I'm most excited about, for exactly the reason you said: it ends up feeling like the effect is responding to the take instead of me guessing at what "natural" should look like.
On the knee/hysteresis idea — you don't sound crazy at all, that's a really nice framing. I see two threads in it:

• A soft-knee trigger. Instead of a hard threshold that's either fired or not, the trigger strength ramps in around the threshold — a glancing transient gives a proportionally smaller, smoother effect, a hard hit gives the full thing. That pairs naturally with the per-onset dynamics already in there (which scales the effect by the hit's loudness); a knee control would basically shape how aggressively that mapping responds near the threshold.

• Peak-aligned interpolation. Landing the effect's maximum exactly on the transient's true peak rather than the threshold-crossing point. Wake already runs on look-ahead, so it has the future samples in hand — it could scan the detection window, find the real peak, and align the swell's apex to it. That's very doable, and honestly a better way to place the swell than what I'm doing now.

So yeah — you've basically described a "trigger response" section I've had half-sketched. Filing this. Appreciate you thinking out loud, this is the good stuff.

[noahrbc]

Another update: Wake 1.1 now has a Spectral swell mode. Alongside the existing reverse-reverb swell, you can now flip the Reverse Swell to Spectral: it freezes the spectrum of each detected hit and resynthesizes the swell from the hit's own timbre, so a vocal blooms into a vocal-colored swell, a synth into a synth-colored one. Same drawable shape, length, and band controls apply. Free update, demo on the site is updated too. Try it on melodic/vocal material; it's a different character from the reverb mode.

[noahrbc]

I'm also working on adding another swell mode that is granular based.

The granular mode would scatter overlapping grains sampled from that snapshot across the swell buffer:

• For each grain: pick a source position in the snapshot, a grain length (say 20–80 ms), copy that slice into the output at a scheduled time, multiplied by a Hann window so its edges taper to zero.

• Grains overlap and accumulate into a cloud that carries the hit's timbre but smeared and textured.

• The existing shape envelope then makes the cloud bloom into the transient.

• Normalize to the hit's RMS for a predictable level (same trick as spectral).

[oobesan]

if you added some pictures and a video of it in action, I think more people would check it out.

[pabloaldunate]

A demo with slower music would be nice, guitar or synths.

[noahrbc]

Wake v1.15 now includes Harmonized swells. The Spectral swell can now bloom into a chord built from each hit's own spectrum — Octave, Fifth, Power, Major, Minor, Octave Stack. A kick blooms into a tonal sub-riser; a vocal hit into a harmonized swell rising into the word. Plus a new Color control that sweeps the swell from dark to bright into the transient. It's the direction I'm taking Wake — less "audio-reactive shaper," more transient-triggered swell instrument. Free update, demo updated, and the launch price is now $20.

Will also be adding more demo videos on different instruments/styles of music to showcase the effect better!

[noahrbc]