Tidewaves

What Tidewaves is for

Tidewaves is a broadband glue compressor — the one dynamics processor in the Smash Happy Audio mastering line, built around a simple rule: zero added colour. The audio path is never filtered or EQ'd; every character change comes from what the detector listens to and how the gain-reduction envelope moves, never from anything added to the signal itself.

At Swell 0 the plugin is a bit-exact, sample-for-sample passthrough — engaging it is always a deliberate choice, not a baked-in colouration.

The interface & The Tide

A full-width scrolling waterline panel — The Tide — runs across the top of the window, showing roughly 26 seconds of gain-reduction history. The waterline pulls down as the glue grips, exactly like a tide pulling out, with a glowing meniscus that brightens the harder it's working. Six knob sections sit below it: Swell, Current, Undertow, Lean, Depth, and Makeup.

Swell & Ratio

Swell is the main compression amount — it moves the threshold. At 0% the plugin takes a bit-exact idle path; there is no separate on/off switch. Ratio is a six-step hardware-style switch (1.2, 1.5, 2, 2.5, 3, 4 : 1) rather than a continuous dial, so you always know exactly which ratio you're on.

Current — ballistics

Attack and Release live here as two lit vertical faders. Switch on AUTO and Tidewaves derives both from the material itself: attack follows the smoothed crest factor (transient-heavy material gets a slower attack so punch survives), and release blends a fast and a slow time constant depending on how sustained the gain reduction has been — brief overshoots spring back quickly, sustained reduction lets go slowly. The live, auto-derived values are shown right on the readout so you always know what the compressor is actually doing, not just what a knob implies.

Undertow — detector character

All three share the exact same clean gain-computer core — Undertow changes only the detector and ballistics behaviour, never the tonal colour.

Lean & X-Over

Both shape what the detector hears — never the audio path itself. Lean is a tilt on the detector's own listening: positive values make it more sensitive to highs, negative to lows, changing what triggers compression without EQ'ing the output. X-Over is a high-pass filter on the sidechain (20–500 Hz, 20 = off) — raise it to stop a heavy low end from driving the gain reduction on its own.

Depth — stereo link

Controls how tightly the left and right detectors are linked. 100% is a single shared detector — the safest, most image-stable glue setting. Lower values let each channel compress more independently, which can add width and movement at the cost of some stereo stability. The section meter shows live left/right divergence.

Makeup & Mix

Makeup (−6 to +12 dB) is the only output trim — there's no automatic loudness matching, so what you hear is genuinely what the compression is doing, not a level illusion. It's smoothed internally so fast moves and automation glide rather than step. Mix blends processed against dry for parallel-style glue.

Undo & A/B compare

Specifications

Troubleshooting

Swell is up but I hear no compression

Check Ratio isn't parked at the lowest step and that X-Over isn't set so high it's excluding the material driving the mix. Watch The Tide waterline — if it isn't moving, the detector isn't seeing anything worth reducing.

The sound feels squashed on transients

Try AUTO on Current, or manually lengthen Attack — a faster attack catches more of the initial transient. VARI-MU's gentle bloom can also feel different from VCA's precision; try switching Undertow.

Output feels louder or quieter after compression, not just "glued"

Makeup is a manual trim with no auto-level matching by design — that's so you can judge the compression on its own merits. Dial Makeup to match perceived loudness once you're happy with the amount of glue.