Three Sisters

Mannequins 10HP
FILTER VCF RESONATOR

Verified sheet

The facts on this sheet have been cross-checked against the sources below.

Sources
Suggest a correction

Triple linked SVF filter. Three tuned peaks (LOW, CENTRE, HIGH) share a master cutoff and spread. Switches between 24dB crossover and three-bandpass formant responses; QUALITY sweeps through notch to self-oscillation.

Patch Ideas · 10

★ Three-band parallel processor
Audio → ALL(IN) · Response: crossover · FREQ noon, SPAN wide · take LOW(OUT) to compressor, HIGH(OUT) to reverb, CENTRE(OUT) to dry mixer · recombine externally for multiband shaping without phase cancellation
▸ Show diagram
Patch diagramPatch diagram with 5 modules and 6 connections. Modules: Three Sisters, Mixer, Audio, Compressor, Reverb. Signals: 6 audio.Three SistersMixerAudioCompressorReverbResponse: crossoverFREQ: noon · SPAN: wideALL INaudioLOW OUTaudioCENTRE OUTaudioHIGH OUTaudioInaudioOutaudioInaudioOutaudioInaudioOutaudioaudio
★ Pinged formant chord
Response: formant · QUALITY ~2 o'clock · trigger burst → ALL(IN) · LOW/CENTRE/HIGH peaks ring as three-note chord · SPAN sets chord interval · FREQ plays the chord via v/8
▸ Show diagram
Patch diagramPatch diagram with 4 modules and 3 connections. Modules: Three Sisters, Trigger, Sequencer, Output. Signals: 2 audio, 1 pitch.Three SistersTriggerSequencerOutputResponse: formantQUALITY: ~2 o'clock (high Q)ALL INaudioFREQ1v/octALL OUTaudioOutaudioPitch1v/octInaudio11. pings the peaksaudiopitch
★ Three-voice sine oscillator
Nothing patched to inputs · QUALITY full CW (all three self-oscillate) · FREQ CV = v/oct from sequencer · SPAN sets interval between voices · LOW/CENTRE/HIGH = three sine voices · stack or mix for instant sine chord
▸ Show diagram
Patch diagramPatch diagram with 5 modules and 4 connections. Modules: Three Sisters, Sequencer, VCA 1, VCA 2, VCA 3. Signals: 3 audio, 1 pitch.Three SistersSequencerVCA 1VCA 2VCA 3QUALITY: full CW (self-oscillating)SPAN: sets chord intervalFREQ1v/octLOW OUTaudioCENTRE OUTaudioHIGH OUTaudioPitch1v/octInaudioInaudioInaudioaudiopitch
Dynamic spectral ducker
Audio → ALL(IN) · Response: crossover · SPAN at noon · QUALITY CCW for notch · LFO → QUALITY CV sweeps between notch and dry · creates a pulsing spectral hole that ducks the resonant frequency region on the beat
▸ Show diagram
Patch diagramPatch diagram with 4 modules and 3 connections. Modules: Three Sisters, LFO, Audio, Output. Signals: 2 audio, 1 cv.Three SistersLFOAudioOutputResponse: crossoverQUALITY: CCW (notch)ALL INaudioQUALITYcvALL OUTaudioOutcvOutaudioInaudio11. pulses notch depthaudiocv
SPAN VCA gate
Audio → CENTRE(IN) · Response: crossover · SPAN at zero = CENTRE passband closes · envelope → SPAN (via attenuverter externally) opens/closes CENTRE as a rough VCA · QUALITY high shapes the gate's tone · percussive plucks with built-in resonance
▸ Show diagram
Patch diagramPatch diagram with 4 modules and 3 connections. Modules: Three Sisters, Envelope, Audio, Output. Signals: 2 audio, 1 cv.Three SistersEnvelopeAudioOutputResponse: crossoverSPAN: starts at 0 (closed)CENTRE INaudioFREQcvCENTRE OUTaudioOutcvOutaudioInaudio11. open gate via FREQ since SPAN has no CVaudiocv
Vowel morphing
Noise or sawtooth → ALL(IN) · Response: formant · QUALITY high · slow LFO → FREQ CV sweeps all three formants in parallel · SPAN set for vowel-like /a/–/i/ spacing · ALL(OUT) = morphing formant voice
▸ Show diagram
Patch diagramPatch diagram with 4 modules and 3 connections. Modules: Three Sisters, LFO, VCO, Output. Signals: 2 audio, 1 cv.Three SistersLFOVCOOutputResponse: formantQUALITY: highALL INaudioFREQcvALL OUTaudioOutcvSawaudioInaudioaudiocv
Audio-rate FM sines
QUALITY full CW (self-oscillating) · VCO saw → FM · FM attenuverter open ~2 o'clock · FREQ sets carrier pitch · three FM sines with slightly different peaks produce metallic cross-modulated tones · SPAN detunes voices for beating
▸ Show diagram
Patch diagramPatch diagram with 3 modules and 2 connections. Modules: Three Sisters, VCO, Output. Signals: 2 audio.Three SistersVCOOutputQUALITY: full CWFM attenuverter: ~2 o'clockFMaudioALL OUTaudioOutaudioInaudioaudio
Resonant drum body
Response: formant · QUALITY just below self-oscillation · short trigger → ALL(IN) · FREQ sets pitch · SPAN narrow for tom-like unified ping, wide for metallic bell · AD envelope → QUALITY CV for damping
▸ Show diagram
Patch diagramPatch diagram with 4 modules and 3 connections. Modules: Three Sisters, Envelope, Trigger, Output. Signals: 1 audio, 1 cv, 1 trigger.Three SistersEnvelopeTriggerOutputResponse: formantQUALITY: ~2-3 o'clockSPAN: narrow for tom, wide for bellALL INtrigQUALITYcvALL OUTaudioOutcvOuttrigInaudio11. dampingaudiocvtrigger
Mid/side-style crossover split
Stereo sum → ALL(IN) · crossover mode · SPAN wide · LOW(OUT) + HIGH(OUT) → external sum as 'sides' · CENTRE(OUT) → 'mid' · independent processing of spectral extremes vs midrange
▸ Show diagram
Patch diagramPatch diagram with 3 modules and 4 connections. Modules: Three Sisters, Mixer, Audio. Signals: 4 audio.Three SistersMixerAudioResponse: crossoverSPAN: wideALL INaudioLOW OUTaudioHIGH OUTaudioCENTRE OUTaudioSides AaudioSides BaudioMidaudioOutaudioaudio
Notch-to-peak morph
Audio → ALL(IN) · QUALITY knob around noon · bipolar LFO → QUALITY CV · signal morphs between deep spectral notches (CCW excursion) and resonant peaks (CW excursion) at the same filter frequencies · vowel-like wah with negative resonance
▸ Show diagram
Patch diagramPatch diagram with 4 modules and 3 connections. Modules: Three Sisters, LFO, Audio, Output. Signals: 2 audio, 1 cv.Three SistersLFOAudioOutputQUALITY: noonALL INaudioQUALITYcvALL OUTaudioOutcvOutaudioInaudio11. notch ↔ peak morphaudiocv

Behaviors

Crossover mode Response switch: crossover

LOW becomes 4-pole (24dB) lowpass, HIGH becomes 4-pole highpass, CENTRE becomes a bandpass whose edges are LOW and HIGH cutoffs. SPAN controls the CENTRE passband width — at zero SPAN CENTRE closes entirely, acting as a SPAN-controlled VCA.

Formant mode Response switch: formant

All three paths become 12dB/oct bandpass peaks. LOW and HIGH track FREQ ± SPAN; CENTRE tracks FREQ alone and is immune to SPAN. Sum via ALL(OUT) builds vowel-like resonant stacks; ideal for pinging with triggers.

Anti-resonance / notch QUALITY knob CCW of noon

CCW half mixes each filter's complementary (inverted, 180° out of phase) output back into the signal. Full CCW produces deep notches at the peak frequencies — the inverse of the CW resonant response. Useful as a spectral duck or for subtractive timbre sculpting.

Self-oscillation QUALITY past ~3 o'clock

Each of the three peaks sings its own pitch. With nothing patched, tapping LOW/CENTRE/HIGH yields three sine voices tracked by FREQ v/8. In formant mode the three peaks can be pinged by triggers at ALL(IN) for additive percussive voices.

v/8 pitch tracking FREQ CV input

FREQ jack is calibrated to 1V/octave over roughly five octaves. With QUALITY high enough to self-oscillate, Three Sisters becomes a three-voice sine oscillator. Intervals between the three peaks follow SPAN, making chord-like voicings from a single v/oct.

ALL(IN) / ALL(OUT) parallel processing Patch to ALL jacks

ALL(IN) fans the same signal into all three filters; ALL(OUT) is a passive equal mix of the three outputs. Patching ALL(IN) and taking individual outs splits one signal into three spectral bands; patching individual INs and taking ALL(OUT) is a three-channel spectral mixer.

Audio-rate FM Audio source into FM with attenuverter open

The FM jack is capable of audio-rate modulation, not just slow sweeps. Self-oscillating Three Sisters plus an FM carrier creates classic FM timbres. The FM attenuverter at full CCW produces symmetric negative-going modulation for through-zero-style character.

Stackable ALL mix Stackable on ALL(OUT)

Officially supported passive mixing: stackable ALL(OUT) with other Three Sisters outputs or other modules on the same bus. No active mixer required; voltage summing occurs on the cable.

Controls

Global FREQ Master cutoff. Shifts all three filter peaks in parallel across ~5 octaves.
v/8 tracking · adds to FREQ CV jack
Global SPAN Pushes LOW down and HIGH up around FREQ center. In crossover mode widens/narrows CENTRE passband; in formant mode CENTRE is unaffected.
Narrow: peaks converge · Wide: peaks separate across spectrum
Global QUALITY Bipolar resonance. CW increases Q to self-oscillation; CCW mixes inverted complementary filter outputs producing notch / anti-resonance.
CCW: notch · noon: dry · CW: resonate · ~3 o'clock: self-oscillates
Global FM attenuverter Grey knob. Scales and inverts FM jack before summing to FREQ. Accepts audio-rate modulation.
CCW: invert · noon: zero · CW: unity
Global Response switch Selects filter topology. CROSSOVER splits band into 24dB LP / BP / HP; FORMANT turns all three paths into 12dB bandpass for vowel-like spectra.
Two-position toggle · crossover | formant

I/O

IN · 7

  • LOW (IN) ±5V audio AUDIO
    Signal into the LOW filter path only
  • CENTRE (IN) ±5V audio AUDIO
    Signal into the CENTRE filter path only
  • HIGH (IN) ±5V audio AUDIO
    Signal into the HIGH filter path only
  • ALL (IN) ±5V audio AUDIO
    Signal routed to all three filter paths in parallel
  • FREQ ±5V · v/8
    1V/oct cutoff pitch tracking input
  • FM ±5V CV
    Attenuverted modulation summed to FREQ. AC/DC coupled, audio-rate capable.
  • QUALITY ±5V CV
    CV sums with QUALITY knob. ±5V at noon ≈ full sweep.

OUT · 4

  • LOW (OUT)
    LOW path. Crossover: 24dB lowpass at FREQ−SPAN. Formant: 12dB bandpass.
  • CENTRE (OUT) GATE
    CENTRE path. Crossover: bandpass bounded by LOW/HIGH cutoffs (VCA-like gate). Formant: 12dB bandpass at FREQ.
  • HIGH (OUT)
    HIGH path. Crossover: 24dB highpass at FREQ+SPAN. Formant: 12dB bandpass.
  • ALL (OUT)
    Equal-level passive mix of LOW/CENTRE/HIGH. Can clip with loud inputs or self-oscillating peaks.