I have a catmull-rom spline represented by its end vertices plus control points. I would like to be able to scale it so that it can be resized/stretched/squeezed in a way that looks linear. However, by simply scaling the coordinates and control points linearly only works if the scaling is isotropic ( same factor in x and y ). Scaling the control points anisotropically causes the curve to interpolate differently since the equation is non-linear. So i wonder if there is a formula or an iterative approach to find a new pair of control points that would produce the desired "linearly-scaled" curve?
in theory, if i have a number of correct samples, shouldn't i be able to reconstruct thw control points by solving some sort of equation?
Turns out linearly scaling the control points works just as it should.. -.-