Interpolation is the calculation of an attribute value that lies in between two key frames.
For example, if an object should move in a circle, then we can define an animation that mutates its X and Y coordinate attributes.
The animation definition can represent this using 5 key frames:
|Key Frame #||X coordinate||Y coordinate|
For a perfect circle, the values in between the key frames can be calculated by the
function for the X coordinate, and the
cos(..) function for the Y coordinate. So, if we
were trying to calculate what the coordinates should be when
t = 0.5, we could go
sin( 0.5 * π ).
However, what if we do not have such perfect coordinate control, and we only have the values at the specified key frames?
To move in a circle, the X coordinate first increases with a larger step, and the step size decreases as it approaches the circle boundary on the X axis, where it then flips, and increases in the negative direction. For the Y coordinate, the magnitude of the step size increases downwards, then decreases once it has gotten past the halfway point.
The changing step size means, given the first two key frames, 0 and 1, the values do not change in constant step increments — linearly (LERP) —, but spherical linearly (SLERP).
The spherical linear function is a way of saying, given these two key frame values, and some proportion of time between the two key frames, what should the actual value be given that the step increments change as they would on a sphere?
In computer graphics, there are a number of methods commonly used to calculate the interpolated
values. The following functions are available in Amethyst, implemented by the
minterpolate library, namely:
Amethyst also allows you to specify your own custom interpolation function.