TŒRN — Menu map

Menu map

The whole settings tree at a glance — every leaf links into the chapter that explains it. This is the navigation map; the per-screen previews live in the screen index.

Root menu — how navigation works

Open the menu by holding a touch pad from the grid. The root list uses red and blue dots on row 16; the encoder 4 hint strip is green to signal "this knob navigates". Submenu families re-colour the same row with the parent menu colour, so you always know which world of settings you are inside.

1 Encoder 1

Short	Exit submenu / back to grid	Exception: AUTO uses short press to run generation.

2 Encoder 2

Turn	Value adjust on subpages	Volumes, enums, spans.

3 Encoder 3

Turn	Alternate value rows	Input selects, reset modes, etc.

4 Encoder 4

Turn	Move between siblings	Watch dots on row 16.
Short	Enter highlighted page	Loads mode or opens nested submenu.

The full tree

Short labels are what the device prints. Indentation is how deep you dive from the root list. Each entry below explains what the page does, every option you can set, and what each value means in practice.

MENU Root list. Reached by holding either front touch pad from the grid. Encoder 4 walks the family items shown as colour dots on row 16; short-pressing encoder 4 enters the highlighted family. Encoder 1 short steps back one level. The menu overlays your pattern — playback never stops while you are here.
- FILE Pattern slot screen — save or load an entire pattern (all 16 pages, every voice, every per-step value, every tone setting) to a numbered file on the SD card. Encoder 4 turn spins through slot numbers 1–999; existing slots glow bright green, empty slots dim blue. Encoder 1 short = load. Encoder 2 short = save (overwrites in place — no confirmation prompt). Encoder 3 short toggles the sidecar: when on, loading also pulls in the matching .txt file next to the pattern (custom palette, MIDI sync offsets, pack hint).
- PACK Sample-pack slot screen — chooses which numbered folder of WAVs fills the eight sample voices (channels 1–8). Synth voices 11, 13, and 14 are unaffected; rows 9, 10, and 12 are empty. Encoder 4 turn spins through packs 0–100; pack 0 is the scratchpad that absorbs any one-off samples you load in the sample browser. Encoder 1 short = load pack (replaces all 8 voices). Encoder 2 short = save pack index so the same pack reopens after the next power cycle.
- WAVE Sample browser for sample voices 1–8. Opens directly when you select WAVE from the root menu, or with a both-touch (left-first) gesture from the grid on a sample row. Encoder 1 turn = trim start 0–100 %. Encoder 2 turn = trim end 1–100 %, always above the trim start. Encoder 2 short = reverse the highlighted sample. Encoder 3 short = load the trimmed file into the voice and autosave. Encoder 4 turn navigates files in the current folder (up to 999); encoder 4 long opens / leaves a folder. Long-press encoder 3 from here jumps into record mode to capture a fresh sample.
- BPM Master tempo and clock-source screen. Three colour states tell you who's leading: cyan + green right-arrow = internal master, green + green left-arrow = external clock locked, red + red left-arrow = external clock missing or jittery. Encoder 2 turn = matrix brightness 3–255 (LEDs only, no audio change). Encoder 2 short toggles the "draw idle" tint between full voice colour and the dimmer base. Encoder 3 turn = clock source INT ↔ EXT (remembered across reboots). Encoder 4 turn = BPM 40–300 when INT; in EXT it nudges a stored offset without overriding the live clock. Tap tempo: on this screen, repeated taps on the appropriate touch pad average into the live BPM.
- VOL Output mixer family. Six pages controlling every place audio leaves the device, plus a high-frequency softener for long sessions.
  - MAIN Headphone output level on the 6.35 mm jack. Encoder 4 turn = 1–10. 1 is whisper-quiet, 10 is full output. The number is shown in cyan; the same encoder press exits to the grid.
  - LOUT Line-out level — independent of MAIN so you can run your phones loud while the mixer feed stays calibrated. Encoder 4 turn = 13–31 (matches the underlying line-driver step). 13 is a low studio-friendly level; 31 is hot. If your mixer's input is clipping, drop this first.
  - PREV Preview loudness — how loud samples sound while you are browsing them in the sample browser. Encoder 4 turn = 0–50. 0 disables auditioning. Independent from MAIN/LOUT so you can browse quietly during a live set.
  - 2-CH Stereo routing of the eight sample voices. Three modes: OFF (mono — every voice in centre), M+P (mono blend with a stereo perception pad), L+R (true wide stereo — voices spread by per-voice pan). Pick L+R when going to a mixer with stereo channels; pick OFF for a quick mono PA.
  - SPKR Internal speaker on/off. Two states: ON (hear everything through the built-in driver — good for the couch) or OFF (mute the speaker even when no headphones are plugged in). The line out stays live in both cases.
  - HFC High-frequency cut. A gentle global treble softener for fatigue-free long sessions or for taming bright headphones. Encoder 4 turn moves through a small set of steps from OFF (flat) up to a noticeable shelf cut. Affects every output equally.
- PLAY How the grid feels: visual aids, navigation behaviour, paint mode, pattern chaining and loop length. None of these affect the audio engine — they shape interaction only.
  - FLW "Follow": a gentle whole-grid glow that pulses on each beat. Two states: ON (a soft tint washes the grid in time with playback — useful in dark rooms or for performance feel) or OFF (only the playhead column lights up, much cleaner for editing).
  - PREV Preview policy for the sample browser. Two states: ON (every file under the cursor auto-previews while you turn encoder 4 — fast hunting) or PRSS ("press to preview" — files only play when you short-press encoder 3, useful on stage so a clumsy turn doesn't squirt audio out).
  - VIEW Note rendering density on the grid. Two states: EASY (only the current page's notes — beginner-friendly) or FULL (current page bright, neighbour pages dimly hinted in the background so you can see context — power-user view).
  - PMD Pattern mode — how playback walks from one page to the next. Four states: OFF (loop forever on the current page — DJ mode), ON (march forward to the next page that has notes — quick A→B→C song form), SONG (follow the SONG arranger chain), NEXT (queue a target page and jump to it once at the end of the current loop — live transitions).
  - LOOP Pattern length cap — how many steps of a page actually sound before the page repeats. Encoder 4 turn = OFF (the natural 16 steps) or 1–16 (cut short). Short loops feel groovebox-like (chains rebar on a count of 5 for an odd-meter feel); long loops feel linear.
  - CTRL What encoder 2 does on the draw/single grid when no special mode is active. Two states: PAGE (encoder 2 turn = flip pages 1–16) or VOL (encoder 2 turn = per-row volume slider for the row your cursor is on — great for live mixing). The choice survives reboot.
  - LEDS Tells the firmware which LED panel hardware is installed. Four states: 1 (one panel, default), 1B (one panel mirrored), 2 (two panels stacked = 16×32 grid for double-wide patterns), 2B (two panels mirrored). Pick what physically matches your unit — wrong settings paint pixels off the visible area.
  - PONG Regenerative note trigger in draw mode: a ball bounces across the grid, and every time it hits a step that already has a note, that step fires again (the sample or synth at that cell). OFF / ON from the PONG page. With ON, work in draw mode to see the ball on your pattern. Speed (values 1–99, higher = faster) sets how often the ball moves; when PONG is already on, open the PONG menu page and use encoder 2 turn to dial speed, or turn it back OFF. This is not the idle “eyes” matrix animation (that is separate, after long inactivity when not playing).
  - CRSR Cursor rendering style on the grid. Three states: NORM (thin white ring around the cursor cell — clearest), CHNR ("channel-restricted" — cursor is only drawn on the focused voice in single mode), BIG (chunky 3×3 white halo — easier to read from a distance during live performance).
  - DRAW Paint mode style. Two states: L+R (default — both hands required: encoder 1 picks the row, encoder 4 paints the steps) or R (one-handed — only encoder 4 paints, gated by the MUTE mask below). Use L+R for careful editing; flip to R for fast live painting.
  - MUTE Per-channel paint mask used by the R draw mode. Decides which rows accept new notes while you paint. Each of the 16 channel positions is a toggle: lit = accepts paint, dim = ignored. Use it to constrain a paint pass to "just the kick and snare" so a finger slip can't drop notes onto melody voices.
- REC Recording family: which input you capture, how loud, when it fires, and what happens to existing notes on the voice after a re-record.
  - INPT Audio input source for the recorder. Two states: MIC (built-in / mic-in jack — best for vocals, foley, low-level sources) or LINE (top-edge stereo line in — best for synths, drum machines, modular). Switching also reassigns which gain control (MIC or L-IN below) is in charge.
  - MIC Microphone preamp gain. Encoder 4 turn = 0–64. 0 = silent, ~32 is conversational level, 64 is hot for whisper sampling. Watch the level meter on the record mode screen and back off if it pegs red.
  - L-IN Line-input trim. Coarser than MIC because line gear is already pre-amplified. Encoder 4 turn = 0–15. 0 mutes the line in; 15 boosts a weak source. For most synths, 6–10 is a sensible starting point.
  - TRIG When the recorder actually starts capturing audio. Three states: OFF (manual — you tap a touch pad to start), SENS (threshold — recording arms when the input crosses a level you set on the record mode screen), CNT (count-in — a four-tick metronome before recording rolls). Use SENS when sampling acoustic sources hands-free.
  - CLR Note-clear policy: what happens to existing notes on the voice you just re-recorded into. Five states: OFF (keep the existing notes — fastest way to swap a sound under an arrangement), ON (clear every note on the voice — fresh start), FIX (clear notes but keep velocity / probability / lane volume), ON1 (clear only the current page), CLIC (clear and lay down a single click on step 1 so you can immediately hear the sample at the right tempo).
- MIDI Everything about exchanging clock and notes over the TRS MIDI ports. The five subpages cover note mapping, transport, what to send, what to receive, and fine sync offsets.
  - CH How incoming MIDI notes are mapped onto the device's voices. Three states: YPOS (incoming MIDI channel 1 plays voice row 1, channel 2 plays row 2, etc. — fastest for a controller using channels), MIDI N (notes go onto whichever voice is currently selected on the grid — single-voice playing), KEYS (synth voices follow a "keys" mapping — incoming pitch becomes the synth note; sample voices use velocity for triggers).
  - TRAN MIDI transport — start, stop, continue messages. Three states: OFF (ignore and never send), GET (listen for MIDI start/stop so an external sequencer can run TŒRN), SEND (transmit your transport so external gear follows). Combine with the clock source on the BPM screen for full master/slave.
  - SEND What TŒRN emits on the MIDI out jack. Three states: OFF (silent — useful when MIDI cable is plugged in only for return), CLK (clock pulses only — the simplest sync setup), NOTE (notes only — drive an external multitimbral synth), BOTH (clock + notes). Pick CLK when slaving a drum machine; pick NOTE when using TŒRN as a sequencer for a separate tone module.
  - RCVE Whether incoming MIDI notes actually trigger the internal voices. Two states: ON (notes coming in light up the internal sample / synth voices according to the CH mapping above) or OFF (incoming notes are ignored — TŒRN becomes a pure sequencer/recorder).
  - SYNC Per-direction sync offset to compensate for real-world latency. Two adjustable lines: SNC1 shifts the send path, SNC2 shifts the receive path. Encoder 1 turn picks which line you edit, encoder 2 turn changes the offset in milliseconds, range −127 … +127 ms. Use when you can hear an audible flam between TŒRN and another clocked device: nudge in ±2 ms steps until the downbeats click as one.
- SONG 64-slot arranger. Each slot points at one of the pattern's 16 pages; with PMD = SONG, playback walks the chain and wraps to the start. Encoder 4 turn = song slot 1–64. Encoder 2 turn = page to drop into the highlighted slot (1–16 or empty). Encoder 2 short = clear slot. Encoder 4 short = insert empty slot. Empty slots are skipped during playback. While playing you see two strips: the page strip on row 16 (which page is sounding) and a moving green dot on row 1 (which song slot you are on).
- NEW Pattern generator — seeds a fresh pattern based on a genre template plus a length, then drops you back on the grid with the result live in RAM. Encoder 2 turn = genre (HOUSE, TECHNO, DNB, AMBIENT, HIPHOP, BREAKS, BLANK). Encoder 3 turn = length in pages (1–16). Encoder 4 short = generate & load (overwrites current pattern — save first if you care). Encoder 1 short = cancel.
- ETC Everything that didn't fit anywhere else: firmware info, generative fill, ring-light toggle, palette selection, battery, and the reset ladder.
  - INFO Firmware version and scrolling credits. Read-only. Encoder 4 turn scrolls the text; the build version appears on the first row so you can confirm before reporting a bug.
  - AUTO Generative fill across a page span — extends what's already on a voice into more pages, in the style of what you already wrote. Encoder 2 turn picks the span in pages, encoder 3 turn picks the source voice the algorithm learns from. Encoder 4 short = run generation (writes into pages immediately after the source). Encoder 1 short = back. Re-running with the same settings produces fresh output every time.
  - LGHT Optional external ring-LED strip (mounted around the chassis) on or off. Two states: ON (ring mirrors the matrix's main accent colour for visual flair) or OFF (saves battery; also useful on dark stages where you don't want extra glow).
  - COLR Colour palette / scheme — swaps the entire voice palette and the dim "base" tint behind it. Three built-in slots: 0 (factory — red/orange/yellow/green/cyan/blue/magenta/pink), 1 (blue-ish gradient), 2 (high-contrast red/violet). A fourth slot becomes available when you save a custom palette via the colour-scheme web tool (saved as scheme.txt on the card).
  - BATT Battery state of charge as a percentage. Read-only number, refreshed every few seconds. Accurate enough for "is this enough juice for the set?" — when it drops below ~10 %, plug in USB-C. The number is shown in green > 50 %, yellow 20–50 %, red < 20 %.
  - RSET Reset ladder — three increasingly destructive recoveries. Always wait for the device to settle before judging the result. Three targets: SD (rescan the card — useful after you swapped files via USB; non-destructive), EFX (reset every tone control back to factory across all voices — patterns and sample assignments survive), FULL (wipe the entire session: clears notes, velocities, probabilities, sets BPM to 100, rescans SD. Equivalent to a blank slate — only FILE-saved patterns can come back).

For every screen's actual LED layout — and a flat searchable thumbnail grid — open the Screen index.