The 248t MARF is extremely powerful for percussion because it is not just a sequencer. It is a dual arbitrary function generator, meaning each side can independently move through 16 programmed stages, with per-stage:
That makes it ideal for:
In short: this module excels at composed rhythmic structure that can still mutate.
From the manual, the key percussion-relevant features are:
Those features let you treat the MARF as:
Instead of thinking “16-step drum sequencer,” think:
That means you can build rhythms that feel like:
A very effective starting point:
Example:
Because the two FGs can have different stage timings and loop boundaries, you immediately get phase-shifting percussion.
The Cycle First / Last modifiers are one of the best rhythmic tools here.
Program:
Now you have a 5-against-7 cycle relationship.
Other strong pairings:
Because each stage can also have different durations, this is much more interesting than plain Euclidean overlap.
Use one FG for a relatively stable pulse network and the other for longer-cycle asymmetry. That keeps the groove intelligible while still becoming complex.
The lower slider row sets interval time per stage. This is where the MARF becomes a monster.
Instead of keeping every stage equal, deliberately create uneven phrase lengths.
For example, to suggest a 7/8-like pattern, make a loop of 7 stages with grouped durations like:
This can imply groupings such as:
For more fractured time feels:
You are not limited to notation, either. You can make timings that sit between conventional subdivisions and feel “elastic.”
Each stage can independently emit Pulse 1, Pulse 2, both, or neither.
That means one FG can already do two intertwined trigger patterns.
Example on one Function Generator:
Stage design idea:
This creates rests and ghost stages that shape groove without always sounding a drum.
Patch:
Now every addressed stage contributes to larger rhythmic behavior, even if no drum sounds directly on it.
Very important: not every stage needs a pulse.
Program stages that:
These silent stages are excellent for:
This is one of the easiest ways to get complicated but musical percussion.
The manual states that time can be sourced from internal sliders or external CV via inputs A–D.
This is huge.
You can patch an LFO, random source, envelope, another sequencer, or even the other MARF side into A–D, then program certain stages to use external time source.
Result:
This makes only specific rhythmic cells drift, which is far more interesting than modulating the whole clock.
The module has a global time multiplier with CV attenuverter, scaling stage times from half-time to 4x longer.
For percussion, this is ideal for:
Now FG1 breathes in and out rhythmically while FG2 anchors the groove. This creates sophisticated cross-rhythmic motion.
The All Pulses output emits a pulse every time a new stage is addressed.
This is one of the module’s secret weapons.
Use it to clock:
Because MARF stage lengths are variable, the All Pulses output creates an irregular event clock. This is perfect for non-grid percussion ecosystems.
Now the timbre rotates while the rhythm is already complex.
These are not just utility conditions; they are compositional tools.
A stop stage halts until a start pulse is received.
Use this for:
Example: - patch a gate from another sequencer or manual button into Start - MARF runs until a programmed stop stage - external event releases it into the next phrase
This is excellent for semi-controlled percussion arrangements.
With high pulse at Start input, the stage is held as long as the gate stays high.
Use this to: - stretch a rhythmic cell - create held rests - freeze on an accent - lock into a micro-loop
The FG pauses until voltage above 5V arrives at Start input.
Use a sparse gate source to “authorize” advancement. This can produce: - gated complexity - phrase-dependent fills - polyrhythms that only emerge when another layer lines up
These three modifiers can make the MARF behave more like rhythmic state machine logic than a normal sequencer.
The stage address section includes:
When in Continuous, stage position sweeps through stages using address control and stops the internal clock.
This is amazing for percussive sound design.
Instead of hearing discrete sequence advance, you can use the MARF as a scan-able bank of programmed voltages/times/pulses.
This is not conventional drum sequencing; it becomes voltage-addressed rhythm topology.
Even if you mainly care about triggers, the voltage side is where the groove becomes alive.
The Voltage Output follows the programmed output voltage and modifiers.
Patch it to drum parameters such as:
Then each stage has its own character, not just its own trigger.
A dense rhythm sounds clearer when each event has controlled articulation, not constant full energy.
If a stage is set to Quantize, voltage output becomes 1V/oct scaled. You can also choose key and scale.
This means MARF can sequence:
For dense percussion music, a great trick is to keep one drum layer tuned:
That creates a hybrid of rhythm and melody without needing another sequencer.
Per-stage Sloped/Stepped is excellent for percussion, especially when the voltage output is not going to pitch but to timbre.
Use sloped transitions to modulate:
This gives a “morphing machine percussion” effect between events.
Patch Voltage Output to kick pitch FM amount or tom oscillator pitch, with some stages sloped and some stepped. You get a mix of: - hard discrete hits - sliding tuned drum motion - liquid pitch dives between accents
The manual says:
For percussion modulation, Half or Limited is often better than Full because it gives more precision.
Use reduced range when controlling:
This helps keep the patch punchy instead of wild and unfocused.
Goal: Dense interlocking drums with long non-repeating cycle.
Use uneven interval times on both sides. This gives immediate polymeter and evolving phrase overlap.
Goal: Stable groove plus chaotic micro-detail.
This creates a clear backbone with hyperactive top-end percussion.
Goal: Groove that implies changing meter.
Program FG1 as an 8-stage loop but with stage durations grouped: - 3 short - 2 medium - 3 short
Program FG2 as a 6-stage loop with a different grouping.
Then: - kick on FG1 Pulse 1 - snare on FG1 Pulse 2 only on certain long stages - FG2 pulses drive offbeat hats and ghost percussion
Add silent stages and use voltage output to open the VCA more on sparse hits, less on dense ones.
This makes the percussion feel deliberate instead of cluttered.
Goal: phrase-aware fills you can release manually or via logic.
This is powerful live: - regular groove continues until you “approve” the fill - then the sequence advances into a dense, weird section - it returns to main loop later
Very performance-friendly.
The Reference Output gives a downward ramp over the interval time.
Patch it to: - LPG CV - VCA CV - filter cutoff - FM amount - wavefolder symmetry
This is especially useful for: - bongos - woodblock-like pings - toms - plucked noise percussion - LPG-driven hats
Longer stage times produce longer decay-like contours; shorter stages produce tighter ticks.
This can make the MARF function like both sequencer and envelope source at once.
Since stage programming can choose external voltage source A/B/C/D, you can feed different modulation sources and select among them per stage.
For example: - A = random stepped CV - B = envelope - C = LFO - D = fixed offset
Then set certain stages to external voltage source and choose which input is active via the stage slider behavior.
Great for: - changing snare pitch source on only some hits - selecting among several accent curves - alternating between deterministic and random modulation
This is excellent for complex percussion timbral variation.
The MARF can become chaotic fast. A few practical rules help.
If everything is irregular, nothing reads as groove.
Keep one of these relatively stable: - kick pattern - hat pulse stream - phrase length - accent system
Let the rest mutate around it.
Empty stages are crucial. Density comes from contrast.
For percussion, better targets are often: - decay - LPG level - click/transient - tone - distortion amount
Use one FG as phrase architecture and the other as detail engine.
It makes percussive modulation tighter and easier to tune.
If you’re sequencing drum voices or oscillator-based percussion, try these:
That last one is especially Buchla-like and very strong.
Set: - loop lengths with First/Last - rough interval times - stop points if needed
Assign: - Pulse 1 main hits - Pulse 2 counter-hits - silent stages for air
Program voltage rows for: - accent - pitch - decay - timbre
Add: - external time source to a few stages - time multiplier CV - enable or sustain logic - different loop lengths on FG1/FG2
Reserve one or two useful controls: - manual Start release - time multiplier amount - stage address strobe - switching gate source for ART / pulse behavior
That gives complexity you can still perform.
The 248t will shine with:
Especially good pairings: - LPG + Reference out - logic + pulse outs - burst generator + All Pulses - switch + external CV inputs A–D - another sequencer clocked by All Pulses
If you want one practical first patch:
That one patch already gives: - polymeter - asymmetry - accents - fills - evolving overlap - timbral change
For hyper-complex percussion, the 248t MARF is best used as a rhythmic composition engine, not just a trigger sequencer.
Its strengths are:
If you want dense rhythmic music with: - polyrhythms - odd meters - evolving phrase lengths - articulated accents - non-repeating drum systems
the winning strategy is:
That will get you from “sequenced drums” to highly structured rhythmic machinery.