Well, I fixed the texture "drifting" (explained in response to Rune in 
previous post).  I've been coliding balls all sorts of ways and it seems to 
work very well now.  It's fun and I don't know why.  I think the ability to 
see physics in slow mo is cool.  I thought I'd post a similar animation to 
the previous one, but this time I've set the bottom ball spining like a top 
and added a couple of balls.  I've slowed it down (a little less than half 
speed) at the expense of simulation time.  Maybe it's not long enough, it's 
nice to see them come to a rest.  It's intesting how a bunch of balls 
together will absorb a lot of the kinetic energy in a faster balls when it 
collides with them.

I think I'll try filling the box with balls tonight to see what happens.

Post a reply to this message

Attachments:
Download 'BBStack4.mpg' (413 KB)

Very nice!

I wonder what these would be like in a hollowed out large sphere or a half
pipe...?

Scateboarding spheres... hehe! ;)

Post a reply to this message

"Neonux" <neo### [at] softhomenet> wrote in
news:3f6bd6dd@news.povray.org: 

> Very nice!

Thanks.

> I wonder what these would be like in a hollowed out large sphere or a
> half pipe...?
> [...]

Implementing trace() is my next step which should allow the balls to 
collide with any surface/objects.

Post a reply to this message

"Neonux" <neo### [at] softhomenet> wrote in
news:3f6bd6dd@news.povray.org: 

> Very nice!
> 
> I wonder what these would be like in a hollowed out large sphere or a
> half pipe...?

I've finished implementing trace() in my code to allow the balls to 
collide with anything.  It works very well, and the time penalty is 
minimal, I was really surprised.  I've tried different types of objects, 
simple and complex, including isosurface objects.  Now I just need an 
interesting obstacle course for the balls to go through.

Thanks Neonux for the idea of the hollowed out sphere and half-pipe.  It 
was one of the first objects I tried when I was implementing trace() 
because one of the primary objectives for me was to test the rotational 
(rolling) aspect of my logic with trace() and your idea was perfect to 
test this.  Although the attached animation isn't that exciting it does 
show my collision to trace() logic at work.

Post a reply to this message

Attachments:
Download 'BBTrace.MPG' (175 KB)

Magic!!!  :)  I love this anim... :)

Some of these animation techniques really need to be combined into a single
animation tool.

Open Source project would be wonderful with a nice interface... Hmmm..

It would enable people to make animations absolute masterpeices.


> Thanks Neonux for the idea of the hollowed out sphere and half-pipe.  It
> was one of the first objects I tried when I was implementing trace()
> because one of the primary objectives for me was to test the rotational
> (rolling) aspect of my logic with trace() and your idea was perfect to
> test this.  Although the attached animation isn't that exciting it does
> show my collision to trace() logic at work.

Post a reply to this message

"Neonux" <neo### [at] softhomenet> wrote in
news:3f6e43e0$1@news.povray.org: 

> [...]
> Some of these animation techniques really need to be combined into a
> single animation tool.
> 
> Open Source project would be wonderful with a nice interface... Hmmm..
> 
> It would enable people to make animations absolute masterpeices.
> [...]

Yeah, I also wish there was an easy to use free 3D physics simulation 
program.  I love "newtonian" physics.  Unfortunatly my little Povray SDL 
program is only specialized at handling balls, not so versatile.

MechSim in MegaPov is actually quite capable.  I wish more people would 
post MechSim animations here.  Unfortunatly, Povray alone is not an easy 
tool to use for the uninitiated, and then MechSim is a large step beyond 
that.  Add to that the fact that Povray is not real-time interactive 
(both design and render), and the audience becomes that much smaller.  
Many commercial 3D packages now support "physics" simulations, even 
Macromedia's latest Shockwave, which was typically not considered a 3D 
software (it uses Havok's real-time physics engine, which is quite cool 
(havok.com)).

If you use Linux, track down an old program called "Aero", it's a fun to 
use rigid body dynamics program and supports (old) Povray output.  
Problem is that it's no longer in development and a little buggy.  There 
are also some open source "physics" engine API projects going on which 
show promise, although I haven't tried any of them yet.

Post a reply to this message

"Neonux" <neo### [at] softhomenet> wrote in
news:3f6e43e0$1@news.povray.org: 

> Magic!!!  :)  I love this anim... :)
> [...]

Oh, and thanks, it's encouraging.

Post a reply to this message