Link Search Menu Expand Document

Quaternions

Quaternions represent rotation through an extension of complex numbers. A full explanation of quaternions and their intrinsics is out of this document’s scope, in this section however enough information is provided to use quaternion’s effectively as a tool.

In practice quaternions are rarely constructed directly as it is fairly difficult to get an intuition for its argument values.

val q = Quaternion(0.4, 0.3, 0.1, 0.1)

Instead quaternions are created from Euler-rotation angles and concatenated in quaternion space. Working in quaternion domains warrants consistent rotations and avoids gimbal locks.

val q0 = fromAngles(pitch0, yaw0, roll0)
val q1 = fromAngles(pitch1, yaw1, roll1)
val q = q0 * q1

Slerp

Spherical linear interpolation, or colloquially “slerping” solves the problem of interpolating or blending between rotations.

val q0 = fromAngles(pitch0, yaw0, roll0)
val q1 = fromAngles(pitch1, yaw1, roll1)
val q = slerp(q0, q1, 0.5)

Quaternion to matrix

Naturally quaternions can be converted to matrices. Quaternions have a matrix property that holds a Matrix44 representation of the orientation represented by the Quaternion.

val q0 = fromAngles(pitch, yaw, roll)
drawer.model *= q0.matrix.matrix44