arbhar

Instruo 18HP

fw 2.0

Granular processor with six 10-second buffers, two grain engines, 1V/oct pitch tracking, onboard capture and onset detection, and mono or stereo input paths. Built-in microphone and USB sample import turn it into a live recorder as well as an effect.

Patch Ideas · 7

★ Frozen cloud

Capture a slow pad into layer alpha, then play it back as a stable granular cloud with stereo spread.

Walkthrough

Press Layer until alpha's LED lights.
Patch a slow pad → Input. Press Capture, hold ~3 seconds of source, press Capture again to stop.
Set Intensity ~noon (dense continuous grains).
Set Length ~1 o'clock (~500 ms grains) and Grain Window noon (Gaussian, smooth fades).
Set Scan ~noon and Spray ~10 o'clock for a tight but moving cluster.
Set Pitch at unity and Dry/Wet ~3 o'clock (mostly wet).
Patch Output 1 → Mixer L and Output 2 → Mixer R.

Signal out Output 1 / Output 2 — stereo audio (~10 Vpp). Grains are coin-tossed between channels for a wide image.

Listen for The pad becomes a sustained, slowly drifting cloud that doesn't loop identically — grain start positions wander around Scan and pan randomly between L and R.

▸ Show diagram

Onset-driven stutter

Let arbhar listen to a drum loop, fire grains on every detected attack, and clock an external envelope from the Pulse Output.

Walkthrough

Patch a drum loop → Input (also normalises to Onset Input).
Capture ~4 seconds into layer alpha (press Capture, then Capture again).
Set Sensitivity ~2 o'clock so quiet hits still trigger.
Set Hold ~9 o'clock (~50 ms minimum gap between detections).
Set Length ~10 o'clock (~100 ms short grains) and Intensity ~10 o'clock.
Patch Pulse Output → external Envelope (Trig) to fire a VCA in sync with detected hits.
Patch Output 1 → Mixer L, Output 2 → Mixer R, Dry/Wet ~2 o'clock.

Signal out Output 1 / Output 2 — granular stutter mixed with dry. Pulse Output emits a +5 V trigger on each detected onset.

Listen for Every drum hit punches a fresh grain out of the buffer; the external envelope opens in lockstep, so an outside voice ducks or fires along with the rhythm.

▸ Show diagram

Omega wavetable sweep

Link all six layers under omega and sweep a slow LFO across the whole bank as one long wavetable.

Walkthrough

Capture different short sources into layers alpha through zeta (cycle Layer + Capture six times).
Press Layer until omega is selected (all six LEDs lit).
Set Length ~9 o'clock (~50 ms grains), Grain Window noon (Gaussian).
Set Grain Direction noon (50/50 forward/reverse), Pitch at unity.
Patch a slow bipolar LFO (~0.1 Hz) → Scan CV Input.
Set Scan ~noon so the LFO sweeps the full layer chain.
Set Intensity ~1 o'clock and Dry/Wet full CW.
Patch Output 1 → Output L, Output 2 → Output R.

Signal out Output 1 / Output 2 — stereo wavetable-style sweep through all six captured layers.

Listen for A slow, evolving timbre that morphs across all six samples like one long wavetable. As the LFO rises, you cross from layer alpha's material into zeta's; turn Spray up to scramble within each layer.

▸ Show diagram

Dub sound-on-sound smear

Use the Dub knob to layer fresh captures over old ones, smearing a saw drone into a thick wash.

Walkthrough

Press Layer until alpha is selected.
Patch a saw VCO → Input.
Set Dub just past noon (half-old / half-new on each new capture pass).
Press Capture and let ~5 seconds record, then press Capture again.
Repeat capture 3–4 times — each pass blends new audio with the previous content.
Set Intensity ~10 o'clock (low continuous engine), Length ~2 o'clock (long grains).
Set Pitch at unity, Dry/Wet ~2 o'clock.
Patch Output 1 → Output (In).

Signal out Output 1 — mono mix of dry saw plus smeared granular layers (Output 2 unpatched normalises into Output 1).

Listen for Each capture pass thickens the buffer instead of replacing it, so the saw becomes a layered, blurred drone with the original still audible underneath the wet grains.

▸ Show diagram

★ Strike percussion accents

Fire one short pitched grain per sequencer step with the strike engine, sitting on top of a dry rhythm.

Walkthrough

Capture ~3 seconds of any tonal source into layer alpha.
Turn Intensity full CCW to disable the continuous engine — only strikes will sound.
Set Length ~10 o'clock (~80 ms), Grain Window noon.
Set Pitch at −12 semitones (one octave down from unity).
Set Pitch Deviation ~2 o'clock (CW = quantised intervals per grain).
Patch a sequencer Gate → Strike Input. Each gate fires one accent grain.
Set Dry/Wet ~2 o'clock so the dry rhythm stays audible.
Patch Output 1 → Output (In).

Signal out Output 1 — dry source plus one short pitched grain per Strike trigger. Output 2 unpatched mirrors Output 1.

Listen for A pitched percussive accent on every step, transposed an octave down with quantised pitch jumps from Pitch Deviation. Continuous engine stays silent, so the rhythm is purely strike-driven.

▸ Show diagram

On-board mic field recorder

Use the panel microphone with no input cable to capture room sound, then play it back through the granular engine.

Walkthrough

Leave Input and Onset Input unpatched — the on-module microphone normalises into the capture path.
Press Layer until alpha is selected.
Set Input Level ~11 o'clock and tap near the module to test the mic LED responds.
Press Capture, speak / play / scrape near the module for ~5 seconds, press Capture to stop.
Set Length ~12 o'clock (~750 ms), Intensity ~1 o'clock, Spray ~11 o'clock.
Set Scan slowly turning by hand or patch a slow LFO → Scan CV Input.
Set Dry/Wet full CW (wet only) and patch Output 1 / Output 2 → Mixer L / R.

Signal out Output 1 / Output 2 — granular playback of the mic recording. No external source, no cables on the input side.

Listen for Whatever you recorded acoustically (voice, keys, a tap on the case) comes back as a granular cloud you can scrub through with Scan. Useful as a self-contained sketchpad with no other modules patched in.

★ 1V/oct melodic granular voice

Capture a held tone, then play arbhar as a pitched instrument tracked by a sequencer's 1V/oct.

Walkthrough

Patch a sustained VCO drone → Input.
Capture ~2 seconds into layer alpha, then press Capture again to stop.
Patch a sequencer 1V/oct → 1V/Octave Input (tracks grain transposition).
Patch the sequencer Gate → Strike Input (one grain per note).
Turn Intensity full CCW so only strikes sound.
Set Pitch at unity (your sequencer drives pitch from 0 V).
Set Length ~11 o'clock (~150 ms), Grain Window noon.
Set Dry/Wet full CW, Output 1 → Mixer L, Output 2 → Mixer R.

Signal out Output 1 / Output 2 — melodic granular voice. One pitched grain per sequencer gate, transposed by 1V/oct.

Listen for A plucky, vocal-ish lead that follows your sequence note for note. Sweep Scan to change the timbre source from inside the captured drone; raise Pitch Deviation CW for harmonised stacks.

▸ Show diagram

Behaviors

Two grain engines Intensity above minimum or Strike input

The continuous engine runs from Length and Intensity, while the strike engine fires on Capture Button or Strike Input. Together they can generate up to 88 grains polyphonically.

Layered capture Capture Button with Layer selected

arbhar records into one of six 10-second layers. Accumulative capture can pause and resume within a layer, turning one buffer into a composed collage.

Scan and follow Layer omega or follow mode

In omega, the Scan knob addresses all layers through position. In follow mode, the playback head tracks the record head, keeping grains anchored to captured motion.

Onset-driven capture audio into Onset Input

Onset analysis can arm capture and also generate triggers or gates from detected attacks. The Pulse Output becomes a usable event stream for external timing.

Pitch deviation Pitch Deviation knob

Counterclockwise adds random pitch drift. Clockwise adds quantized interval offsets, so each grain can move in a repeatable harmonic family instead of pure noise.

Stereo and mono paths Output 2 patched or left open

With only Output 1 patched, Output 2 normalizes to mono. Patching Output 2 opens stereo imaging and lets each side take independent grain distribution.

Audio capture from the panel mic no cable at Input or Onset Input

The onboard condenser microphone normalizes into the capture path. The module becomes a self-contained field recorder without external cabling.

Controls

Global	Input Level	Sets input gain before the limiter. Around 10 o'clock is unity; clockwise drives the preamp. unity at ~10:00 · analogue limiting on high levels
Global	Dry/Wet	Crossfades between the dry input and the granular engines. equal-power crossfade
Global	Output Level	Sets the level at both outputs. post-granular output trim
Global	Capture	Starts and stops recording to the selected layer. press to record · press again to stop
Global	Shift	Accesses secondary functions such as capture modes, track-and-hold, and mode selection. holds secondary state · used with Capture and Layer
Global	Layer	Selects one of the six layers, or omega to link them through Scan. alpha through zeta · omega = scan all layers
Continuous engine	Intensity	Sets grain density and asynchronous versus synchronous grain generation. centre = max grains · extremes disable continuous engine
Global	Strike	Generates a single accent grain from the strike engine. button and trigger input both fire a grain
Shared grain engine	Grain Window	Shapes the grain envelope from square to Gaussian to sawtooth. CCW: square · noon: Gaussian · CW: descending saw
Shared grain engine	Grain Direction	Sets the probability of forward versus reverse playback. CCW: forward · noon: 50/50 · CW: reverse
Shared grain engine	Length	Sets grain duration. ~4 ms to 3 s · center around 750 ms
Shared grain engine	Scan	Sets the playback position inside the active layer. sets grain start position at grain onset
Shared grain engine	Spray	Adds random offset around Scan. CCW: tight cluster · CW: fully random layer access
Shared grain engine	Pitch	Sets the base grain transposition. applied at grain onset
Shared grain engine	Pitch Deviation	Adds per-grain random or quantized pitch offsets. CCW: unquantized random · CW: quantized intervals
Onset analysis	Sensitivity	Sets onset detector sensitivity or Input 2 level in stereo mode. shared with onset analysis threshold
Onset analysis	Hold	Sets the minimum time between onset detections or recorded events. 5 ms to 10 s
Capture	Dub	Sets sound-on-sound attenuation for overdub capture. CCW: replace · noon: half-old/half-new · CW: full layering

I/O

IN · 8

Input audio AUDIO

AC-coupled audio input for capture and playback processing.
Onset Input audio AUDIO

Audio input for onset detection and stereo capture analysis.
NORM → Input
Capture Pulse Input pulse / gate GATE

Gate or trigger input that starts, stops, or retriggers capture.
Intensity CV Input bipolar CV CV

CV over grain density and amplitude distribution.
Length CV Input bipolar CV CV

CV over grain duration.
Scan CV Input bipolar CV CV

CV over layer position.
Strike Input pulse / gate GATE

Trigger or gate input for the strike granular engine.
1V/Octave Input 1V/oct

Pitch tracking input for grain transposition.

OUT · 3

Output 1 audio AUDIO

Audio output. Grains are coin-tossed between both outputs by default.
Output 2 audio AUDIO

Second audio output. Unpatched, it normalizes to Output 1 for mono playback.
NORM → Output 1
Pulse Output +5V trigger / gate GATE

Trigger or gate output from the onset engine.

Patch Ideas · 7

Behaviors

Controls

I/O

▸ IN · 8

◂ OUT · 3

IN · 8

OUT · 3