Been looking at my code for a couple days now and it SHOULD work, if my math is correct, which after many sit downs, I think it is.

The goal:
Get my model to tilt properly to the terrain under it (without using the physics engine)

The technique:
I started off with just the tilt and would move on to the roll after the tilt was working.

I started with using c_trace to find the distance to the ground from the center of my model. Then use another c_trace to find the distance to the ground from the front of my model. Once I had those two distances, I could get the difference between them, and use atan to find the angle that the model should be angled to.

Here's a picture to show what I mean:


The formula based on the picture would be:
theta = atan((A-B)/L)

I think this is correct... but for some reason I'm not getting the angle I think I should be.
I know the slope my model is standing on is exactly 15 degrees, but it constantly returns 7.125


Can anyone see anything wrong with my method here, or if there is an easier way to do this (without using the physics system).


Thanks in advance for any help.

if you use the trace anoyhow, you can read out the normal vector of
the surface that was hit.

Do several traces to the front in different angles,
get the normal vector, and make an weighted average of them.

if you turn your model 90 degrees to the normal vector
you have the same effect.


Your current method assumes a very steady / flat terrain,
if some obstace comes, the angle can jump quite quickly.


---

if the math in your calculation for the angle is correct, you can
check on paper.
if the values differ from what you have in the engine,
you probably uses some math function with the
wrong parameters.

Based upon the debug values I have for the c_trace distance returns, the paper math is giving me the same answer as the engine.

That is what led me to believe it was a problem with the c_trace function. I don't think it's giving me accurate distance values.

Here's the code I have for c_trace and setting up the vectors to use in it. Let me know if I've done something wrong here, please.

vec_set(t_vec.x,vector(0,0,-500));
vec_for_ent(t_vec.x,me);
floor_dist_center = c_trace(me.x,t_vec.x,IGNORE_ME | IGNORE_PASSABLE | IGNORE_PASSENTS | IGNORE_MODELS | IGNORE_SPRITES);
vec_set(t_vec.x,vector(60,0,-500));
vec_for_ent(t_vec.x,me);
vec_set(t2_vec.x,vector(60,0,0));
vec_for_ent(t2_vec.x,me);
floor_dist_front = c_trace(t2_vec.x,t_vec.x,IGNORE_ME | IGNORE_PASSABLE | IGNORE_PASSENTS | IGNORE_MODELS | IGNORE_SPRITES);

if (floor_dist_center != 0 && floor_dist_front != 0)
{
targtilt = atan((floor_dist_center-floor_dist_front)/60);
}