[ODE] integration with cal3d?

[Adam Moravanszky [Novodex]]

> but what you're really talking about with a realistic
> physics simulation running the inputs to the rag doll
> instead of key frame animation is reproducing the
> human brain.

I don't think that was the question!

>> How easy/smart would it be ... have an animated model collapse like a rag
doll

There is quite a bit more work involved because you have to get all the
transforms between cal3d and ode right, but it is certainly doable and the
results will be worth your time.  I think the main problem is actually not
the programming part but rather the fact that no free authoring environment
exists where one could easily rig skinned characters with the extra stuff
needed for physics, like masses, joint limits, and bounding boxes.


--  Adam Moravanszky
/*================*\
|Chief Software Architect
|NovodeX AG
|physics middleware
|www.novodex.com
\*================*/

----- Original Message ----- 
From: <jnilson_99 at yahoo.com>
To: "Jason _____" <jgmath2000 at hotmail.com>; <ODE at q12.org>
Sent: Saturday, January 10, 2004 4:12 AM
Subject: Re: [ODE] integration with cal3d?


>
> i've use cal3d a little bit. it looks really cool but
> in reality it's just using slirp/quaternions to
> interpolate between the key frames. don't be fooled by
> the pretty models. that's all easily done in 3dstudio
> max/character studio.
>
> cal3d has added some extra spice in that you can slirp
> from one set on key frames to another.
>
> for example their demo where a female character is
> walking and you can make her run through a simple
> "volume" like control.
>
> all that's happening is you're controling which sets
> of keyframes to use. that's great if you want to use
> what is essentialy key frames on steroids.
>
> but what you're really talking about with a realistic
> physics simulation running the inputs to the rag doll
> instead of key frame animation is reproducing the
> human brain.
>
> in fact this was the very problem i believe Russel
> Smith (hello and thank you for ode if you're reading!)
> was trying to solve in his phd thesis and why he
> created ode. (he has a movie somewhere on his web site
> of his character moving using a neural net to run the
> inputs to ode).
>
> beware of the folly of building the "a tower of babel
> to heaven". in trying to get ode to run cal3d with the
> same precision as keyframes you would effectively be
> modeling the entire human motor cortex. yes you would
> win the nobel prize but you would also be making us
> humans obsolete and some people might not like that
> :-).
>
>
>
> john
>
>
> --- Jason _____ <jgmath2000 at hotmail.com> wrote:
>
> ---------------------------------
>
> How easy/smart would it be to try to somehow
> seamlessly merge ODE and cal3d so that you could have
> an animated model collapse like a rag doll similar to
> what you see in modern ragdoll physics in 3d games?
> Would this is a smart approach or would it be way too
> difficult to pull off?