Skip to content
eddie_dodd edited this page Sep 9, 2026 · 1 revision

Phaser

10 HP. Four, six or eight first-order allpass sections in series, swept from 120 Hz upwards over six octaves at full depth, with feedback around the chain.

Nothing tagged Phaser in the Rack library has been rebuilt since 2025, which is most of the reason this module exists.

Controls

Control Range Default
Rate 0.02 to 8 Hz 0.3 Hz
Depth how far up the sweep travels 70%
Feedback 0 to maximum resonance 40%
Mix dry to wet 50%
Stages 4, 6 or 8 6

Ports

Port Direction
Rate, Depth, Feedback, Mix CV in, one per knob, each with an attenuverter
IN L, IN R audio in. R is normalled to L
OUT L, OUT R audio out

Bypass routes IN L to OUT L and IN R to OUT R.

What the stage count actually changes

Each pair of allpass sections produces one notch, so four stages give two notches, six give three and eight give four.

Four is the familiar sound, and it is what most classic hardware phasers used. Eight is deeper and more obviously synthetic, because the additional notches crowd the spectrum closely enough that the ear stops hearing individual moving notches and starts hearing a single swept character. Six sits between them and is the default for that reason.

If you are trying to make a phaser sound like a phaser, use four. If you want it to sound like a modular effect, use eight.

Why feedback is trimmed, and why I am telling you

An allpass chain has unity gain at every frequency. That is the definition of allpass, and it means a feedback loop wrapped around one has nothing damping it. Left alone, high feedback produces a resonant peak around ten times unity.

That is not an effect, it is a level problem. So the signal entering the loop is attenuated as feedback rises, which holds the peak near twice unity across the whole range of the knob. Below roughly half feedback the trim does nothing measurable, so the normal working range of the control is untouched.

This is the one place in the plugin where something happens that the panel does not show you, which sits awkwardly against the idea that these modules do exactly what their labels say. I decided a hidden gain trim was the lesser problem compared with a feedback knob whose top third is unusable. Writing it down here is the compromise.

Where it misbehaves

At maximum feedback the resonance is strong enough to ring on transient material, which reads as a metallic tail. That is inherent to a resonant phaser and not something the trim removes.

The sweep starts at 120 Hz. On bass-heavy material the lowest notch may sit below most of the content, so the effect will sound weaker than it does on a mid-range source. Increasing depth does not help, because depth extends the top of the sweep rather than lowering the bottom.

Clone this wiki locally