2025-06-05

pressure

emits continuous scalar values ∈ [0,1] representing how much a voice or behavior wants to act - based on rhythmic, structural, or contextual forces. pressure is non-triggering, non-allocating, and deterministic: it modulates downstream activation decisions via activation, constraint, or direct summation.

introduction

pressure replaces fixed rules with field-like intention. each form defines a behaviorally meaningful gradient of urgency: pressure builds with silence, aligns with rhythm, tracks prior activity, or expresses arc structure. pressures are not events - they are conditions. they are computed per voice (or per group) across time, and later combined (e.g. summed, multiplied, gated) to decide if an onset should occur. this domain is used together with:

  • role: per-voice personality traits
  • activation: threshold-based onset emission
  • constraint: global or group limitations

overview

each form uses a and b ∈ [0,1] to shape its behavior:

  • pulse attraction

    • description: aligns pressure with a metrical pulse
    • analogy: beat gravity, periodic energy
    • a: grid resolution (0 = long cycle, 1 = high subdivision)
    • b: sharpness of pulse (smooth → impulse)
  • silence tension

    • description: pressure builds the longer a voice has been silent
    • analogy: restrained energy, buildup urge
    • a: ramp rate (how fast tension grows)
    • b: clamp softness (hard cap → soft curve)
  • inverse density

    • description: pressure increases when local activity is low
    • analogy: compensation, dynamic fill-in
    • a: lookback window (short → long)
    • b: relief strength (mild → strong inverse)
  • inertial memory

    • description: pressure carries forward recent activity
    • analogy: continuation tendency, behavioral momentum
    • a: decay time (fast forget → slow drift)
    • b: stickiness (inertia or activation friction)
  • segment emphasis

    • description: pressure peaks inside a defined segment region
    • analogy: swell to climax, arc of attention
    • a: peak location (early → late in segment)
    • b: width of pressure curve (sharp → wide)

parameter behavior summary

  • pulse attraction

    • a: grid resolution (beat unit scale)
    • b: curve sharpness (sine → impulse)
  • silence tension

    • a: buildup rate (0 = slow, 1 = fast)
    • b: ceiling curve (0 = hard cap, 1 = soft taper)
  • inverse density

    • a: history span (short memory → long context)
    • b: inverse pressure gain
  • inertial memory

    • a: how long pressure persists after activity
    • b: resistance to pressure decay
  • segment emphasis

    • a: pressure peak timing (relative to segment)
    • b: spread of the emphasis (sharp → broad)

why these were chosen

each form contributes a structurally distinct and musically intuitive activation tendency:

  • pulse: metrical anchoring
  • silence: time since last event
  • density: response to local activity
  • memory: behavioral continuity
  • segment: macrostructure shaping used alone or in combination, these allow voices to emerge from pressure landscapes, not rigid rules - enabling micro- and macro-level phrasing, independence, and expression.

what is not included

  • direct onset emission (handled by activation)
  • voice assignment or feedback (system is unidirectional)
  • post-hoc adaptation (all pressure is precomputed)
  • energy shaping of audio signals (belongs in dynamics, envelope)

conclusion

pressure turns deterministic sequencing into emergent phrasing. it shifts decision-making from absolute rules to weighted conditions, allowing voices to build, persist, or fade organically - all within the system's additive, parametric structure. it is one of the core drivers of activation behavior, working hand-in-hand with role, constraint, and structural domains like segmentation.