Instrument Engines

Each of the 64 instrument slots can use one sound engine, chosen with the TYPE field. Nine engines are available: Sampler, Sub, Wavetable, Macro, Multi, Reso, Chip, Drum, and MIDI. The available instruments range from sample playback to professional synthesis DSP, plus output to external MIDI hardware.

How the sound engine works

Each channel has one voice, plus 3 extra "chord" voices for paraphonic CHD chords, and there's a shared pool of 16 voices used by all Drum instruments across every channel — so up to 16 drum hits can ring at once regardless of which channels trigger them.

Every voice shares one lowpass filter and one ADSR envelope shape:

• Filter: a single resonant lowpass, roughly 20 Hz–14 kHz. It's bypassed entirely when CUTOFF is at its maximum (FF) and RES is at 0; adjusting either parameter engages it. Touching the filter on a Sampler or Drum instrument also sums the output to mono, since the filter itself is mono, so keep CUTOFF at FF and RES at 0 on stereo samples you want to keep stereo.
• Envelope: attack, decay, and release each run 0–4 seconds depending on the ADSR value; not every engine uses it (see the per-engine notes below).
• Panning is equal-power: 00 = hard left, 80 = centre, FF = hard right.

A master limiter sits on the final mix to catch peaks, so you generally don't need to worry about clipping from stacking several loud instruments.

Shared parameters

Sampler, Sub, Wavetable, Macro, Multi, Reso, and Chip all have: VOLUME, PAN, CUTOFF, RES, SEND A, SEND B, SEMI (coarse transpose, ±24 semitones), SCALE, ROOT, CHOKE.

Drum is the exception: no CUTOFF, RES or SEMI. FINE (sub-semitone tuning) only exists on Sampler and Wavetable.

MIDI has none of these as it has no local voice to apply them to (see MIDI Out).

Every instrument also has two LFOs and four MOD slots on page 2 of the Instrument screen. See LFOs and MOD Commands for more information on these parameters.

Sampler

Plays WAV files. Fields: SAMPLE (file slot), EDIT (opens the waveform editor — dive in with R), LOOP (NONE / FWD / PING / ONCE, default ONCE), START / LOOPST / END (frame offsets — END at 0 means "play to the end of the file"), SLCMOD / SLCRT (slice mode — see Slices), FINE (sub-semitone tuning, ±100 cents). NONE and ONCE behave identically — both stop at the end marker.

Sampler has no amplitude envelope: a note-OFF or the end of a one-shot cuts the voice immediately rather than fading, in every loop mode.

Sub

Simple 3-oscillator subtractive synthesiser. Per oscillator: WAVE (Saw / Square / Tri / Sine / Noise / Off), SEMI (±48), FINE, LEVEL. Saw and Square are band-limited; Tri and Noise are not. Shared filter (CUTOFF/RES) and ADSR envelope.

Wavetable

Scans 2048-sample single-cycle frames, up to 256 per file (longer files are truncated to whole frames and summed to mono on load). WT POS selects the frame (0–255, interpolated between neighbouring frames as it moves).

Slots 0–4 are built-in sine/square/saw/tri/pulse tables, and each of those is a single frame, so WT POS has no audible effect on them. Load your own multi-frame table to hear it scan.

Wavetable has a similar ADSR envelope to Sub. FINE tunes the pitch by ±100 cents. R on the WTABLE field opens the wavetable browser.

Macro

48 algorithmic oscillator shapes (SHAPE field), including everything from classic waveshapes through FM, physical models, and granular textures. TIMBRE and COLOR morph each shape. The same ADSR envelope and filter as the other synths. Built on top of a modern classic DSP algorithm. Yeah, you know the one.

Multi

24 synthesiser models in one engine, selected by MODEL and shaped with TIMBRE, HARM, and MORPH (in that order on screen): two virtual-analog models, phase distortion, three 6-op FM variants, one 2-op FM, wave terrain, string machine, chiptune, waveshaping, grains, additive, wavetable, chords, speech, swarm, noise, particles, strings, modal, bass drum, snare, and hi-hat. On the percussive models, ENV D (decay) also drives the model's own internal decay time as well as the amplitude envelope. More classic DSP from the same place as Macro.

Reso

Modal / string resonator. MODEL selects one of 6 physical models (MODAL, SYMPS, STRNG, FMVOX, SYMPQ, STRRV). Parameters: STRUCT (harmonics), BRIGHT (brightness), DAMP (damping), POS (excitation position). Monophonic. You guessed it - more modern classic DSP.

Chip

4-channel chiptune oscillator for you video game nerds. Each channel (A, B, C, D) plays the instrument pitch with an independent waveform:

Channels B, C, and D also have a SEMI offset (±48 semitones relative to channel A) and an independent V (volume) level. Channel A always plays at the root pitch. All four channels share a single ADSR amplitude envelope and the common CUTOFF, VOLUME, PAN, SEND A, SEND B, SEMI, SCALE, CHOKE parameters. it's worth noting that Chip's envelope decays and releases considerably more slowly than the shared ADSR used elsewhere for the same A/D/S/R values.

Drum

A polyphonic drum sampler with 8 independent pads, drawn from the shared 16-voice pool described toward the top of this page. The left column shows the 8 pads (notes C-4 through G-4) and the right column holds shared parameters (VOLUME, PAN, SEND A, SEND B, SCALE, ROOT, CHOKE).

Each pad has:

• Sample: the WAV file assigned to this pad. A+←→ scrolls through available samples. R (or A-tap) opens the sample browser. The pad name scrolls when the cursor is on that row.
• Choke group: A+↑↓ sets the choke group (0 = none, 1–8). Pads in the same group cut each other when triggered. Choke groups are per-pad, independent of the instrument-level CHOKE.

Pads always play at their sample's natural pitch. Chain transpose and phrase transpose have no effect, and notes outside C-4–G-4 simply don't trigger.

Because Drum pulls from a separate voice pool rather than the per-channel voice, most phrase FX commands (PIT, OFS, REV, SLC, GLI, CHD) have no effect on it. This means pads are fixed-pitch, one-shot triggers by design. VOL, PAN, SDA, SDB, RTG, KIL, and note-OFF do work, releasing or adjusting every pad voice currently ringing on that channel.

Midi

Sends notes to external MIDI hardware. Fields:

• CHAN: MIDI output channel, 1–16.
• BANK: bank select MSB (CC 0), 1–128, or --- to skip.
• PROG: program change, 1–128, or --- to skip. Allows different phrase steps to trigger different presets on the receiving synth.
• SCALE / ROOT: scale quantize mode and root, same as other engines. Controls which notes the phrase editor steps through.

Chain-level transpose, the CHD chord FX command, and VOL all work: transpose shifts the outgoing MIDI note, CHD sends additional Note Ons for each chord tone, and VOL sets the Note On velocity (0x00–0xFF maps to MIDI velocity 0–127). Everything else (PAN, filter, sends, and most other phrase FX) has no effect. See MIDI Out for full setup, bank/program selection, and other info on MIDI functionality.