AutoMovie API
    Preparing search index...

    Variable sampleFluidSprayConst

    sampleFluidSpray: (
        props: IAutoMovieSampleFluidSpray.IProps,
    ) => IAutoMovieFluidSpraySample = ...

    Sample the decorative spray of a fluid domain at one solved state.

    Decorative particles do not change depths or volume ledgers. The supplied state and domain must have matching ids; that check alone does not establish state provenance, array layout or finite fields.

    Particle i has nominal birth i / rate and age time - i / rate. Its position is p0 + direction * speed * age - gravity * age² / 2 * y, where p0 is the emitter cell's free surface in the supplied current state, shared by all its particles. Historical spawn heights are not reconstructed. Direction normalizes the axis, adds seeded per-component jitter scaled by spread, then normalizes again.

    LOD keeps spawn indices divisible by 1 + floor(max(cameraDistance, 0) / lodDistance); the maxParticles cap then keeps the newest survivors. This assumes representable live-window and stride arithmetic.

    Particles are returned per emitter in declaration order and, for representable indices, in increasing spawn order. LOD and caps are applied in the CPU sample; this does not prove GPU display agreement.

    Throws for a mismatched state domain id or non-finite camera distance. Negative distance uses the nearest LOD, and omitted distance defaults to zero. Caller-run domain admission remains required; finite authored fields do not prove finite/safe spawn indices, finite stride or ballistic arithmetic, or a finite nonzero jitter norm. Those derived values are not checked here.