Eight complete voices. One instrument.

Octet is modelled on the Oberheim SEM voice card used in the Four Voice and Eight Voice. A complete synthesizer voice sits behind every note, 8 times over.

That card predates the 4-pole filters of later polysynths. Its state-variable architecture gives Octet a continuous low-pass / notch / high-pass path. Band-pass has its own selection.

Measured targets with limitsSee the proof Voice and playing specificationRead the specifications

Instrument panelOctet

The Octet panel: modulation and two oscillators; mixer with state-variable filter; two envelopes and output

01 The instrument

A complete circuit for every note.

Eight related voices give a chord its internal movement. Each voice carries a complete oscillator-to-amplifier path. The patch seed supplies stable component variation.

VOICE

Each voice keeps its character

Pitch and cutoff vary by voice. Envelope timing has its own spread; amplifier behaviour does too. The same patch seed restores those differences and their slow drift.

FILTER

One control crosses 3 shapes

Low-pass moves through a deep central notch into high-pass. Band-pass remains a separate selection.

Resonance acts on the band around cutoff.

PLAY

Play it wide

Hard sync and pulse-width movement animate a single note. The filter can sound vocal. Stable unison groups open the voice further, while polyphony turns the same motion into chords or pads.

02 Factory sounds

Ten starting points in 2 renders

Night Choir and Glass Current play the same original C - Am - Dm - G - C progression at 96 BPM. Both renders use 30% sustain with a 0.40-second release. The note gate lasts 0.86 seconds; listening level is peak-normalized to -3 dBFS.

10 factory starting points

Hearth Brass · Glass Current · Night Choir · Copper Keys · Hollow Reed · Velvet Pulse · Split Horizon · Northbound Sync · Low Lantern · Static Bloom

Night Choir · chord progression

Glass Current · chord progression

03 Instrument overview

Eight voices. One clear specification.

This is the current Octet specification. It covers the voice circuit first, then the polyphonic engine around it.

Voices: 8 voices by default. The voice pool accepts 1–16.
Oscillators: 2 discrete saw-core oscillators per voice. Each produces saw or variable-width pulse. Oscillator 2 adds hard sync; pulse-width modulation reaches both.
Filter: 2-pole OTA state-variable filter. Mode travels continuously through low-pass / notch / high-pass. Band-pass is separate, with resonance around cutoff.
Envelopes: 2 resistor-capacitor envelopes per voice. Each has attack / decay / sustain / release.
Modulation: One LFO per voice with sine / triangle / sawtooth / square shapes. Sample-and-hold supplies repeatable stepped noise.
Playing modes: Poly / unison / configurable poly-unison groups. Allocation is round-robin; quiet-release voices are stolen first. Glide follows the selected mode.
Output: Stereo main output with program-level pan and repeatable per-voice character
Current format: CLAP instrument. Available through Bitsnacks Vending Machine.
Program state: One shared program across the voice pool. Patch recall is deterministic. The bank contains 10 categorized starting points.

04 Proof

Results arrive with their limits

See the oscillator targets and filter-equation conformance. Rate checks cover the circuit timing. Separate sections state the current calibration and performance limits.

Measured behaviour and current limitsOpen the Octet proof Foundation and specialist seriesCompare the instruments