I found a computer that has just enough power for 3drad. And looks like it all works great, I put together a project just for testing and I did find some strange results with flips. Rolls and spins are flawless though.
Check it out. It seems like for every front flip, you get one spin.

The mysterious spin (Y) factor had me thinking really hard there for a moment, until I visualized the actual rotations, and it all made sense. What actually happens is, that unless there is a perfect circular flip along the X axis of the car, there will always be at least one spin per flip. In other words, there will always be a spin around the Y axis, that is not perceptible to the �UP� minded human brain; but which is mathematically correct.

I have tried to explain this phenomena in the image below.

If the car were to start off in the direction of the Red arrow on the image, and it were to rotate 360 degrees as shown by arrow A. If we were to take the blue line B, to represent the World's Z axis, so that we have a point of reference, then you will notice something very interesting happening.

Notice how the first part of arrow A, crosses on �this� side of the blue line, and how the last part of  arrow A crosses on the �other� side of the blue line. Besides describing a complete circle around the X axis, the front end of the car also describes a full circle around the Y (for UP/Down) axis, as depicted by arrow C. Because the car is measured as rotating around it's own center, this is why, you have a spin (Y) recorded when you get a flip (X).

The measurements are correct, but the visual interpretation is what is confusing. To compensate for this, I suggest subtracting 1 spin per flip.. in other words, adjusting the the Y factor to compensate for this imperceptible phenomena.

Quote

I suggest subtracting 1 spin per flip.. in other words, adjusting the the Y factor to compensate for this imperceptible phenomena.

Doesn't sound too hard. Only I probably need to round the numbers to single digits first. I don't know really, but I'll try not to mess anything up while doing it.

Something like this..(not tested)
if(totalX>1){         //if there are flips..
   totalY-=totalX;   //subtract the flips from the spins
   if(totalY < 0){    //No negative spins wanted..
      totalY = 0;      //so we zero it.
   }
}

Let me know if you have problems, then I'll make the change in the plug-in for you

Gald to hear..
Just another thing now that I think of it..
I just uploaded an update to my OrbitCam.. this uses adifferent z_vp00 mesh, to the one that this plugin uses.. Rather not use that cam for now, till I've updated the plugin to use the newer mesh.. I added a new bone to the mesh to allow for backwards camera movement, and this screwed up the bone numbers, hence the incompatibility.

K, found another problem. It's compiling right now... It'll be a few seconds...

2.30 am here. off to sleep.. Got to finish some animations for my polo game tomorrow, then get Allan's demo done. So I'll look at your new prob soon as I get the chance

Head...Throbing...too...much...math... (Head Exploded)

Sounds good. And thanks for helping out.

Anyway, I think this project calls for a separate floats "carFlipForward, carFlipBackward" because there are different results with each direction.

This is possible, since the measurement is done for a full 360 degrees. I'll see what I can do.