Oscillators

Noise

A multi-color noise source with built-in sample-and-hold — the random voltage at the heart of every modular system.

What is a Noise?

A noise module produces a continuously random signal — a voltage that, sample by sample, is unpredictable. Played as audio, it sounds like a hiss, a rush of wind, or a roar of low rumble depending on its spectrum. Sent as control voltage, it produces erratic, unpatterned modulation. Either way, what makes noise *useful* is that nothing about it repeats: it is the one signal in a synthesizer with no period, no pitch, and no waveform.

The colors of noise — white, pink, red/brown, sometimes blue and violet — are not arbitrary names. They describe the spectrum, by analogy with light. White noise has equal energy at every frequency and sounds bright, hissy, neutral. Pink noise rolls off at -3 dB/octave, giving equal energy *per octave* — the way human hearing weighs frequency — and sounds warm, balanced, natural. Red/brown rolls off harder still and sounds like distant thunder or surf. Choosing a color is choosing how the noise sits in the mix.

Noise has two completely different jobs in modular. As audio: white through a sharply resonant filter is the basis of every wind, ocean, and breath sound; through a short envelope it makes hi-hats, snares, and claps; under a synth voice it adds the attack noise of a real instrument. As CV: noise is what you sample-and-hold to produce *random voltages* — the foundation of generative patches, where a clock fires a S&H that captures a snapshot of noise and freezes it as a pitch, a filter cutoff, or anywhere else you want unpredictability.

Our Noise

Three noise colors out at once — WHT, PINK, RED — so you can mix them or send them to different destinations from a single module. The pink generator uses the Voss-McCartney algorithm with 16 octave-band rows, which gives a much more convincing pink spectrum than the cheaper IIR-filtered white you sometimes see. The red output integrates white through a DC blocker so the signal stays centered around zero rather than drifting off.

Where the module earns its 6 HP is the built-in S&H. Patch a clock to the trigger input and on every rising edge it captures the current noise value and holds it on the S&H output until the next pulse. You get a stepped random CV without needing a separate sample-and-hold module. The source switch picks whether the S&H samples white (uniform random) or pink (less extreme, more musical) — pink usually sounds better for random pitch.

The slew output runs the S&H signal through a one-pole lowpass with adjustable rate, turning the staircase into a smoothly-gliding random voltage. Useful when you want random *movement* rather than random *jumps* — a slowly drifting filter cutoff, a wobbling pitch detune, an unpredictable pan. Both the S&H and the slew output share the same trigger, so they stay in lockstep.

In a patch

For a hi-hat or snare: white noise → VCA → output, with a short ADSR (a few ms attack, ~50 ms decay) gating the VCA. For wind: pink noise → state-variable filter with the cutoff modulated by a slow LFO. For ocean: same patch, but pink → resonant low-pass with a very slow LFO and red/brown noise added underneath for body.

As a CV source, patch the noise output into a S&H clocked by your project's clock. The S&H output is a sequence of random voltages — patch it through a quantizer to a VCO V/Oct input and you have a generative random melody locked to the grid. Patch the same S&H output to a filter cutoff for randomly-stepping timbral changes that align with the rhythm.

A noise source is also an excitation signal for physical-modeling modules. A short noise burst into the input of a Karplus-Strong style resonator, a Rings, or Elements is what makes the model *ring* — the noise is the pluck or strike, the model is the string or bell.

Inputs

Trig (gate) — Trigger input for sample-and-hold. Each rising edge captures a new random value from the selected noise source. Any gate or trigger signal works here.
Clock (gate) — Clock input for sample-and-hold. Works identically to Trig - a rising edge samples a new value. Having two inputs lets you use one for S&H timing and still have the other free.

Outputs

White (audio) — White noise output. Equal energy across all frequencies. Sounds bright and hissy. Use for hi-hats, snares, wind effects, or as a broadband modulation source.
Pink (audio) — Pink noise output (Voss-McCartney algorithm). Energy drops 3dB per octave, so it has more bass and less treble than white noise. Sounds warmer and more natural. Good for ocean/rain textures.
Red (audio) — Red (brown) noise output. Integrated white noise with DC blocking. Very bass-heavy with a deep rumbling character. Good for thunder, earthquakes, or sub-bass modulation.
S&H (cv) — Sample-and-hold output. Outputs a random voltage that changes on each trigger or clock pulse. The voltages are stepped (instant jumps). Perfect for random pitch sequences when patched through a quantizer.
Slew (cv) — Slew-limited S&H output. The same random values as S&H but smoothed by a one-pole lowpass filter. Instead of jumping between values, it glides smoothly. Great for slow, wandering modulation.

Controls

Slew Rate — Controls how fast the slew output tracks the S&H value. Low values (0.001) create very slow, lazy glides between random values. High values (1.0) track almost instantly, approaching the stepped S&H output.
Source — Selects the noise source for sample-and-hold. Off = white noise (fully random, uniform distribution). On = pink noise (values tend to cluster more naturally with less extreme jumps).

Inspired by

A classic noise + S&H combo: simultaneous white / pink / red outputs with a clocked sample-and-hold and a slew-limited variant. The Voss-McCartney pink generator and DC-blocked red integrator are textbook implementations of the analog tradition.

Moog 903 Random Signal Generator
Doepfer A-118 Noise/Random
Analogue Systems RS-110 Noise & Random

← Back to all modules