PDA

View Full Version : Intermediate user - help with making movement of a space ship appear more realistic!



xandercom
04-05-2013, 02:23 PM
Good evening all,

I consider myself an intermediate user of Lightwave, currently using 10, but am spending far too much time on something which may have a much easier solution?

I have a spaceship in my scene which the camera follows. I want to make the ship's movement as it dodge debris in it's way appear more fluid and natural.
To date, all I have been doing is manually keyframing the ship at particular points along it's motion path, but it's hit and miss to get it to feel like it has realistic physics as it twists and turns along the way.

In my head I would imagine a physical solution would be something akin to a null following the motion path, with the spaceship attached to the null with some kind of invisible elastic cable, so it dampens the sharp curves, pulling the spaceship along the path but not as rigid as simply the ship with it's own, manually keyed path. Imagine a toy airplane hanging from a rail above it, and how the toy would bank and turn as it hangs below the rail on corners and straights.


I'd love to hear from anyone who might be able to a) understand what I am trying to describe, and b) know of a possible solution!
Any and all help would be greatly appreciated :)

shrox
04-05-2013, 02:48 PM
What I do is move the ship, then the camera. Create a null, PARENT it to the ship, then TARGET the camera at the null. As the ship moves so will the null, but you can move the null to move where the camera is pointing. It makes for less wild movement of the camera when trying to make interesting cinematography.

xandercom
04-05-2013, 03:12 PM
Hi
The camera's direction isn't really the issue. I've used nulls in that manner before for camera direction.

In this case it's the ship. It's very easy to modify the ship to bank if, for example, the path is an "S" shape, however in this case it's a complicated path in which the ship doesn't just bank left and right, but also goes under objects, above them etc and toward the end it's speed increases as it zooms off out of the scene.
If I try to keyframe it it always looks unnatural. I can tinker with keyframes to attempt manual adjustments, but the end result is just a manual attempt at replicating realistic physics, and flaws are obvious. I end up having to start again from scratch after the motion path keyframing becomes so convoluted with inaccurate attempts at realistic physics modeling.

xandercom
04-05-2013, 03:16 PM
here's an example for you:

http://www.youtube.com/watch?v=V0JrZbUhsbI

For this I had to manually adjust every keyframe to give the blue box the feel of bouncing around with any kind of realistic physics at work, rocking back and forth, bouncing off of things. It took forever, and there were still places where the keyframing was wrong.
There has to be some kind of plugin to allow a motion with realistic physics, gravity etc.

jrandom
04-05-2013, 03:33 PM
When I was working on getting a marble rolling, I got an interesting effect when I tied each rotation axis amount to be directly derived from the distance traveled on the perpendicular axis.

For example, Y_Rotation = (X_Position_Current - X_Position_Last_Frame) * Some_Constant

Gave me fairly interesting banking behavior. Definitely not what I needed for a rolling marble, but possibly useful for other endeavors. Not sure how well that translates to movement on all three axises... axisis? Axis'?

As far as actual movement is concerned, if you're aiming for Newtonian Physics you'll want your motions curves to be very linear until the object accelerates to change direction -- you should see a lot of nice, straight diagonal lines that sharply ease into/out-of keyframes.

Cryonic
04-06-2013, 01:20 AM
Axis -> Axes

If you want newtonian, then think Asteroids. You don't lose velocity without accelerating against it. So, any acceleration either adds to the total velocity (and changes your course) or counters your current velocity. So, the ship would actually rotate in advance of a change of direction of motion. And then it all depends on the max acceleration the engines of a given ship can generate. But those curves will be smooth. No sharp changes in direction.