catmull rom implementation

Unity Engine
1

Hey guys, so I’m trying to implement something simple where an object will jump around between a N number of control points (i use 8) and at the same time trying to learn exactly how this interpolation works. Below is code for what I have generating the vector that i’m constantly translating to in update

	Vector3 ComputePointOnCatmull(double u, int segNumber)
	{
		Vector3 travelpoint = new Vector3();
		Vector3 temp1 = new Vector3();
		Vector3 temp2 = new Vector3();
		Vector3 temp3 = new Vector3();
		Vector3 temp4 = new Vector3();
		//Debug.Log(u);
		//Debug.Log (segNumber);
		
		int p0=0;
		int p1=0;
		int p2=0;
		int p3=0;
		
		if (segNumber == 0) 
		{
			p0=7;
			p1=0;
			p2=1;
			p3=2;
		}
		else if (segNumber ==7)
		{
			p0=6;
			p1=7;
			p2=0;
			p3=1;
		}
		else if (segNumber ==6)
		{
			p0=5;
			p1=6;
			p2=7;
			p3=0;
		}
		else
		{
			p0=segNumber-1;
			p1=segNumber;
			p2=segNumber+1;
			p3=segNumber+2;
		}
		temp1 = 2f*controlPoints[p1];
		temp2 = (controlPoints[p2]-controlPoints[p0])*(float)u;
		temp3 = ((2f*controlPoints[p0])-(5f*controlPoints[1])+(4f*controlPoints[p2])-controlPoints[p3])*(float)(u*u);
		temp1 = ((3f*controlPoints[p1])-(controlPoints[p0])-(3f*controlPoints[p2])+(controlPoints[p3]))*(float)(u*u*u);
		
		travelpoint = 0.5f * (temp1+temp2+temp3+temp4);
		
		return point;
	}

u is a constantly being updated by a counter of .01 and goes from 0.0 to 1.0, segNumber goes from 0.0 to 7.9, controlpoints are the vector locations of where i’m traveling to and interpolating between.

the temps just break up the math behind the equation for a catmull rom (i believe, trying to get it from this site Introduction to Catmull-Rom Splines)

basicaly when this is run, every segment starts at 0 and flys randomly off on some curve and when it reaches the next segment it starts back at 0,0,0 which is not what i want. it doesn’t appear to hit any points, and it shouldn’t be starting at 0,0,0 everytime but i think thats because when u = 0 it resets the travelpoint vector to 0, but when i tried to force when u = 0 change u to 0.01 it still didn’t work. I feel like theres some fundamental piece missing as i’ve checked the equation and i think it’s right.

Let me know if anyone has any ideas.

2

Your equation for temp3 seems to be missing a “p”

temp3 = ((2f*controlPoints[p0])-(5f*controlPoints[1])+(4f*controlPoints[p2])-controlPoints[p3])*(float)(u*u);
                                                  ^
                                             should be p1?
3

wow haha thanks, that at least made it hit the first point but its still starting over at the 0,0,0 instead of starting at the next control point. Thanks alot though i hate missing something so

4

well with that i figured it out, so in case anyone is interested i just had to combine all the temps into one equation because point was defaulting to zero everytime instead of carrying over from the last point.