AutoMovie API
    Preparing search index...

    Interface IAutoMovieProductionSoundEvent

    One semantic event lowered from the compiled film edit into audible space.

    interface IAutoMovieProductionSoundEvent {
        acousticResponse?: IAutoMovieProductionAcousticResponse;
        attenuation: number;
        densityGain: number;
        distanceMeters: number;
        emitter: IAutoMovieVector3;
        event: string;
        frame: number;
        id: string;
        kind: "transition" | "contact" | "break" | "arrival" | "reveal";
        listener: IAutoMovieVector3;
        memberCount: number;
        pan: number;
        propagation?: {
            arrivalFrame: number;
            arrivalTimeSeconds: number;
            boundary: "inside-segment" | "carried-across-cut" | "trimmed-at-segment";
            distanceGain: number;
            highFrequencyGain: number | null;
            profile: string;
        };
        seed: number;
        shot: string;
        spreadRadiusMeters: number;
        timeSeconds: number;
    }
    Index

    Consumer-supplied room-path result. The planner does not derive it, and the mixer uses the dry event when it is absent; acousticProfile alone does not establish a response or enforce a complete response population.

    attenuation: number

    Distance-derived dry gain from zero through one, taken at the root-mean-square source/listener distance hypot(distanceMeters, spreadRadiusMeters) rather than at the centroid alone.

    The declared dry gain law is evaluated at the modeled RMS distance. Population variance comes from uniform-box approximations of compiled position bounds, so this calculation does not measure individual member distances or their combined physical attenuation.

    densityGain: number

    Level factor for memberCount mutually uncorrelated sources, sqrt(memberCount).

    The model assumes equal-level uncorrelated sources, whose powers add: p_total^2 = N * p_single^2, giving amplitude gain sqrt(N) and a 10*log10(N) dB increase. It does not measure actual source correlation; coherent sources would require a different law.

    distanceMeters: number

    Euclidean distance in meters from the listener to emitter.

    Sampled world-space source point.

    event: string

    Authoritative event-contract id.

    frame: number

    Exact film-global event frame.

    id: string

    Stable occurrence id, including the owning film segment.

    kind: "transition" | "contact" | "break" | "arrival" | "reveal"

    Procedural sound family selected by the event contract.

    Sampled world-space camera point.

    memberCount: number

    How many individual sources the event's subjects contain: one for a scene node, the member count for a formation or an instance set, summed over every subject the event names.

    This is what makes a mass sound like a mass. A group used to contribute exactly one emitter no matter how many stood in it, so three people and a hundred thousand were acoustically identical.

    pan: number

    Camera-relative stereo position from -1 (left) through 1 (right), narrowed by the source's own angular width: local.x / hypot(distanceMeters, spreadRadiusMeters).

    Modeled spread reduces the centroid's pan magnitude. As spread becomes small relative to distance, this formula approaches the point-source result; it does not measure each member's stereo contribution.

    propagation?: {
        arrivalFrame: number;
        arrivalTimeSeconds: number;
        boundary: "inside-segment" | "carried-across-cut" | "trimmed-at-segment";
        distanceGain: number;
        highFrequencyGain: number | null;
        profile: string;
    }

    Direct-path propagation result when the production selected a profile. Omitted only on the legacy dry path.

    Type Declaration

    • arrivalFrame: number

      Exact film-global listener-arrival frame.

    • arrivalTimeSeconds: number

      Exact frame-derived listener-arrival time.

    • boundary: "inside-segment" | "carried-across-cut" | "trimmed-at-segment"

      Applied shot-boundary decision.

    • distanceGain: number

      Distance gain derived by the declared law.

    • highFrequencyGain: number | null

      High-frequency linear gain, or null when no spectral model was claimed.

    • profile: string

      Selected production propagation profile id.

    seed: number

    Stable unsigned 32-bit procedural seed.

    shot: string

    Compiled shot that owns the event.

    spreadRadiusMeters: number

    Square root of the modeled source variance in metres about emitter. Formations and instance sets use a uniform-box approximation from compiled position bounds, rather than measuring every member.

    Zero for a single node. A crowd has a radius, and a source with a radius is not a point: it cannot be all in one place in the stereo field, and the listener is not the same distance from all of it.

    timeSeconds: number

    Exact frame-derived film time.