Damping?

[juanpi]

Hi,
I found this line surfing through cpSpaceState

Code: Select all

cpFloat damping = pow(1.0f/space->damping, -dt);

line 503 cpSpace.c

Is this doing (1/space->damping)^(-dt)? This is just (space->damping)^dt
The first equation produces error for really small damping (like free space), while the seconds does not. I guess there is a reason for doing this, right?

Thanks

[ker]

correct me if I'm wrong, but the default value for cpSpace::damping is 1.0.
Therefore free space with no damping would have a damping value of 1.0 and 0.0 would be invalid.
This confused me at the beginning, too, since "no damping" would be 0.0 in my opinion and "full damping" 1.0.

[slembcke]

Well, the thought was that you would be able to double the damping value to double the amount of damping. It is a bit confusing though.

[ker]

but e.g. 0.3 would be really strong damping, and doubling it would be 0.6 which would be a lot less damping in the current setup, or at least that is how it feels to me

[juanpi]

Sorry guy,
Yes, you are write that damping 1.0 is no damping. What was the range for damping in the original design? Is 1.0 < damping < \infty, where damping= 1.0 means no damping or was it 0 < damping < 1, where 1 is fully damped?

Anyway, what I was trying to point out is that

pow(1.0f/damping, -dt)

is the same as

pow(damping,dt)

but you save 2 operations.

Thanks!

Do not miss the new videos in youtube! Magnets!

[slembcke]

/me smacks forehead

Right. I see where you were going with that now.