Yeah, no, I am not looking for accuracy. Just a good look. Right now, I am creating an axis of rotation based on the cross of the plane normal and the reflected motion vector and I am creating an angle of rotation scaled by the dot of the rotation vector to the plane normal (smaller equals less rotation because it is a more dead on hit). This works ok. My problem is that I think I also want to maintain some of the existing rotation, but I am really flubbing this part
I think it is one of those things I am better off putting aside for a little while and coming back to.
Funny you should mention a tutorial. I was hoping to do a quick tutorial using quarternions to accumulate accurate rotations around arbitrary axis in a sphere, both to provide some code for a sphere which people are always asking for and to introduce quarternions. I will get to it at some point.