AutoMovie API
    Preparing search index...

    Namespace AutoMovieHumanSkeleton

    Lightweight posable skeleton of the generated Human, the stickman mode, and its forward kinematics under params.pose.

    Joint centres come from AutoMovieHumanJointCentres, scaled with the body exactly as AutoMovieHuman.generate scales the skin (resolved stature over calibration stature, millimetres to metres), so the rest skeleton overlays the neutral skin. Bones use the engine's humanoid names where the vocabulary has one; the scapula, atlas, subtalar, midfoot and per-toe bones are the Human's own, because the shared humanoid vocabulary lacks them.

    Three postures stay apart. The bind is the achieved posture the skin was generated in, and bones hold it. Neutral is the standing body clinical angles are measured from, and AutoMovieHumanNeutral reaches it from the bind. A pose is a further turn of neutral: every field at zero, or omitted, gives the neutral body, so {} is neutral and not the bind. Each placement's rotation runs from the bind, because that is what a skin reads.

    Every bone declares three orthonormal clinical axes in the registered frame (forward +X, up +Y, left +Z) from its neutral direction d: flexion turns d toward the joint's flexion direction, abduction toward the side away from the midline (the figure's right for midline bones), twist runs along d. The foot bones evert about their long axis; the eyes and the scapula use fixed body axes. Each pose field maps onto one bone axis through the canonical channel table, clamped to its range of motion.

    The skin's binding and the engine's pose come from these placements through AutoMovieHumanRig.

    generate
    prepare