Processor Conduit

s3g Processor Conduit is a mono-or-stereo live-input, stereo-output vocal distortion in which a virtual actuator sends the voice through a virtual material and virtual contact pickups receive it elsewhere. The entire transducer → material → contact-mic path is DSP. It requires no physical transducer, contact microphone, hydrophone, auxiliary channel, or external feedback routing.

The material is the filter: a selectable pedal can damage the voice before the actuator, after the piezo pickup, or on the moving-microphone return. The virtual actuator injects the voice into a propagation path and eight dispersive modes before the nonlinear virtual pickup recovers it. A true octave-down path can turn that recovered body into a smeared, slowed-vocal mass. The design is oriented toward vocal harsh noise and unstable physical color rather than realistic room simulation.

Processor Conduit interface showing stereo input selection, material body controls, pedal placement, octave and moving-PA controls, and feedback containment history — The material channel, movable pedal, PA and microphone motion, stereo input choice, and containment history remain visible as one live-vocal performance surface.

Workflow

Insert s3g Processor Conduit on a live-microphone track and begin with a low monitoring level. The plug-in advertises stereo-in/stereo-out by default; a mono-in/stereo-out CLAP configuration remains available for hosts that support alternate port configurations.
Route the stereo output to the intended speakers or fold-down destination, then establish a conservative audible monitor path.
Set INPUT LISTEN to CHANNEL 1, CHANNEL 2, or SUM MONO when one selected source should drive a single virtual body. Leave it at its default STEREO setting to pass both input channels through two complete Conduit bodies.
Choose a MATERIAL, set SIZE and TENSION, then use DAMPING to decide how long the virtual body answers.
Choose a PEDAL, set POSITION, and raise PEDAL MIX and DRIVE. Use DRIVER to overload the sending transducer and CONTACT to harden the receiving pickup.
Move PICKUP through nodes and cancellations while speaking or sustaining a vowel.
Raise OCT DOWN and DRAG for slowed-down vocal weight. Add FEEDBACK deliberately, then use PA DRIVE and THROW to make the virtual microphone lose, bend, and reacquire the PA.
Use CHAMBER with a vessel, shell, ceramic, or cavity material; set WIDTH, MIX, and OUT; use PANIC when the body must stop immediately. The factory PRESET menu and RANDOM change character without moving INPUT, main MIX, or OUT.

Virtual Physical Channel

INPUT LISTEN selects channel 1, channel 2, an equal-gain mono sum, or independent stereo passage before the selected pedal insertion and actuator.
The selected live input reaches the pedal insertion and actuator, then travels through the virtual material's propagation path and modal body.
Two wide-band piezo pickups recover the body; the optional post-piezo pedal, octave-down grains, chamber, and stereo image shape that recovered signal.
The feedback witness drives the nonlinear PA and moving microphone; the optional post-mic pedal precedes the governed return to the actuator.

POSITION places the selected pedal at one of three actual circuit points. PRE DRIVER distorts the voice before the bandwidth-limited asymmetric actuator. PIEZO > PA independently distorts the two recovered pickup channels after their wide-band piezo preamps and before the octave/PA witness. MIC > LOOP distorts the two moving-microphone returns before they are monitored and re-enter the governed feedback input. Because each position receives different source dynamics and sits at a different place in the regenerative path, the same pedal setting develops a different character.

The actuator output travels through a fractional propagation delay, excites eight material modes, and is observed by two related pickup positions. Each virtual contact pickup combines modal displacement and local velocity with a high-impedance piezo preamp whose 5.5 Hz offset pole preserves low vocal and octave-down energy. A dual-grain shifter can lower the result by one octave before output, and a controlled portion can enter the governed return.

The return drives its own PA system: asymmetric preamp, sagging power stage, four-mode loudspeaker, cone breakup, and fractional air path. THROW moves the virtual microphone through non-periodic inertial distance and angle trajectories, simultaneously changing delay, level, air loss, microphone blend, and polar phase. Stressed high-THROW feedback can fracture into short windowed captures before the Stack/LF Synth-style governors regain it.

PICKUP is not a dry/wet or brightness shortcut. It changes virtual travel distance and the signs and strengths with which the contact point receives individual modes. Sweeping it can cross nodes, reverse partial polarity, and reorganize the vocal spectrum.

In STEREO mode, the left and right inputs retain their identity by driving two independent complete material, pedal, chamber, PA, microphone, and feedback systems. A restrained cross-image at the final pickup field keeps the result spatially related without collapsing it. CHANNEL 1, CHANNEL 2, and SUM MONO instead feed one virtual body, which still generates stereo output through its dual pickups, chamber, and moving-PA image. On the alternate mono input port, all listen choices receive the one available channel.

Materials

Material	Character
Broad materials	METAL, GLASS, PLASTIC, WOOD, WATER, and SKIN cover plate, pane, polymer, timber, pressure/hydrophone, and membrane tendencies.
DIRECT / vessels	DIRECT models an attached actuator and pickup for more truthful voice transmission. METAL VESSEL, GLASS VESSEL, and PLASTIC VESSEL add decorrelated internal chamber paths.
Bronze	DEEP BRONZE, TIERED BRONZE, BROAD BRONZE, and BRIGHT BRONZE range from low gong clusters to close tier pairs and sparse bright branches.
Plates / laminates	CARBON LAM., GLASS PLATE, ALUM. PLATE, and SPRUCE PLATE use distinct paired plate relationships, loss, and propagation.
Shells / ceramics	STEEL SHELL, PORCELAIN, and EARTHENWARE combine curved or fired-body modes with internal air response; porous clay damps fastest.
Membranes	TENSIONED SKIN, LOADED MEM., COUPLED MEM., CAVITY MEM., and LOOSE MEM. provide taut, weighted, paired, cavity-coupled, and slack low-frequency bodies.

The sixteen specific bronze, plate, shell, ceramic, and membrane bodies share the public material vocabulary and paired modal relationships developed for Ambi Encoder Modal, translated here into Conduit's transducer, piezo, chamber, and feedback system. Material names describe designed tendencies, not sampled objects or calibrated measurements of particular specimens.

Controls

Control	Meaning
MATERIAL	Selects the virtual body's modal ratios, loss, propagation speed, and brightness.
INPUT LISTEN	Selects CHANNEL 1, CHANNEL 2, SUM MONO, or STEREO. STEREO passes both inputs through independent complete Conduit bodies and is the default.
INPUT	Sets microphone gain before the actuator.
DRIVER	Raises actuator saturation and mechanical asymmetry while narrowing its bandwidth.
SIZE	Makes the body larger and lower while lengthening travel between actuator and pickup.
TENSION	Moves the modal foundation and changes upper-mode dispersion.
DAMPING	Shortens the body's ring and darkens returned energy.
PICKUP	Moves the virtual contact point through modal nodes and changes propagation distance.
CONTACT	Adds pickup velocity, bite, asymmetric overload, and wavefolding.
FEEDBACK	Returns the recovered pickup signal through a bounded, filtered path to the actuator.
PEDAL	Selects SHRED, WOOL, RAT, ZONE A/B, FUZZ I/II, or DIODE.
POSITION	Moves that pedal between PRE DRIVER, PIEZO > PA, and MIC > LOOP circuit points.
PEDAL MIX	Crossfades between the unprocessed signal at the selected insertion point and the pedal output.
DRIVE / TONE	Sets the selected pedal's distortion intensity and voicing.
OCT DOWN	Blends the recovered material with a true -12 semitone dual-grain shift.
DRAG	Lengthens and darkens the octave grains for a broader, less articulate slowed-vocal character.
PA DRIVE	Pushes the feedback-only PA preamp, power stage, speaker modes, sag, and cone breakup.
THROW	Sets the depth and speed of erratic virtual microphone movement and high-stress short feedback fractures.
CHAMBER	Sets the internal return of vessel, shell, ceramic, and cavity materials.
WIDTH	Opens the two-pickup, chamber, and moving-PA stereo image.
MIX / OUT	Sets the recovered-material blend and final output level.
PRESET / RANDOM	Loads one of ten topology-aware starting points or generates a constrained material/pedal/feedback state. Both preserve INPUT LISTEN, INPUT, main MIX, and OUT exactly.
PANIC	Clears propagation, modes, chambers, PA/air paths, breakup captures, pickups, pedals, and feedback memory without changing controls.

Vocal and Harsh-Noise Performance

PEDAL, DRIVER, and CONTACT are different distortion gestures. PRE DRIVER damages speech before transduction; PIEZO > PA damages the recovered object; MIC > LOOP makes the PA return itself nonlinear. DRIVER damages the actuator's energy, while CONTACT changes the virtual piezo response before any post-piezo pedal. Low DAMPING and moderate FEEDBACK let the body answer between phrases. PICKUP is especially effective on held vowels because a steady excitation makes its moving nodes audible. Slow SIZE movement produces structural bends; faster TENSION changes pull and shear the partial field. OCT DOWN at high DRAG trades consonant clarity for the heavy temporal impression of slowed tape without requiring offline time stretching.

Containment and Limits

The feedback system governs the complete PA, air, microphone, material, and pickup loop. A live-source envelope prevents permanent no-input oscillation; a fast loop envelope catches immediate excess; a second slow-recovery governor controls dense returned tone; and level/roughness detection gradually lowers and spectrally masks an overloaded return. Requested feedback follows a nonlinear performance curve but remains below unity. Modal, chamber, speaker, and microphone excursions are bounded, non-finite input is rejected, and both outputs are soft-limited.

The CONTAINMENT HISTORY display makes that behavior visible. Its upper ENG trace is an eight-second logarithmic history of recovered material energy; the lower CUT trace shows the amount of active governor reduction. Live readouts report energy and cut, while QUIET, OPEN, TRIMMING, and MASKING identify the current containment state. These protections constrain the digital system but cannot make excessive monitor level safe. Use conservative gain and appropriate hearing protection for live microphone work.

Processor Conduit is not convolution and does not claim component-exact simulation of a loudspeaker, piezo disc, hydrophone, material specimen, or room. Material names describe playable tendencies within one virtual system rather than measurements of particular physical objects.