New version from NewtonPhysics released

[yoda64]

I have released today a new version from the NewtonPhysics plugin for A:M .

 

Some words before .

- the new version works only with V12.0p and higher

- Sorry for the mac guys, at this time a mac version is not avaible, because there is no V12 SDK avaible for the mac . I have sent the source to Hash , hopefully they can compile them ....

- there are two known problems, but I would not longer wait with releasing this version (and workarounds are available), read the known problem section attentively

 

What's new

- breakable joints between pieces of a model

- fixed and breakable joints between dynamic and static objects (and Nullobjects)

- A:M forces now considered during the simulation

- extended materialsystem

- all settings now as pluginproperties and (I hope) more clearly

 

 

To find at http://www.sgross.com/plugins/

[pleavens]

Your new example video is great!

 

Thanks,

 

Phil

[martin]

You're such a stud, Steffen. I'll have Steve contact you about your Mac problems. We want everything working well for the V13 introduction.

[NancyGormezano]

This is fabulous stuff - I just watched your domino movie - amazing! - I am so excited to try this out - I just got the 2006CD.

 

So many possibilities...

 

(all the example movies are terrific)

 

Thank you thank you.

[Rodney]

Outstanding. Thanks Steffen!

 

and this:

A:M forces now considered during the simulation

 

Whoaa.

[KenH]

This is great. Much appreciated Steffen!

[Kamikaze]

Super Great !!! Very amazing stuff here.......thank you very much ..

 

Michael

[johnl3d]

Thanks..!!

[PlJack]

For those of you that are unable to find the sample animations, they are here http://www.sgross.com/plugins/plugin6/samples.html

[Ganthofer]

Fantastic Steffen!

 

I haven't even tried all the features in 12 yet, but I'll have to push this to the top of the list.

 

Much Appreciated!!!

[serg2]

SuperCool!!!!! Great Thanks Steffen!!!!

[3DKiwi]

Awesome stuff. Thanks.

 

3DKiwi

[Paul Forwood]

Thank you, Steffen!

 

You helped to make 2005 an exceptional year for the A:M community and I am pleased to hear that Newton Physics is being fully implemented in A:M 13. Well done!

 

Wishing you another outstanding year in 2006.

All the best!

[yoda64]

for the next release

[Paul Forwood]

Hmmm. It's a bit floaty.

 

I jest! I jest!!! Good floaty!!

 

Very nice, Steffen!

[Stuart Rogers]

Yes, very floaty!

 

If that movement's entirely simulation, I'm very impressed. All we need know is a plug-in to automatically generate animated bump/displacement maps for water ripples!

[yoda64]

It's completly simulatet, fluids are the wizard word (can be used also for air movement,examples are in the renderqueue ) .

Displacement/bumps (but when I remember correct aaver have a plugin in this direction), no , but I'm thinking over fluidsurface animation , but this need more research and some (or better much more :-)) more time .

[KenH]

It would be great if you had the time Steffen! It's looking great as is....ripples would be the...um.....ripples on the water.

[John Bigboote]

CAN'T WAIT to get my hands on the upgrade! THANKS STEFAN!

 

Curious...does it do figurative gravitations? (As in...apply it to a rigged character and throw the character out the window to see how he falls...)

 

We had talked about that on the 1st version and I think Stefan said that the newer version of Newton was needed...

[yoda64]

@John

Yes , I remember and I have my hand on this also . But at this time I need a genius idea , how to find the assigned cp's from given bones , to build the collision hull .

[aaver]

[...]how to find the assigned cp's from given bones , to build the collision hull .

 

Steffen,

 

I'm not sure if I understand your problem exactly, but you are familar with the method HSegmentCache* HCP::GetSegmentCache(); aren't you?

 

My first idea would be to traverse the HModelCache the standard way and call the GetSegmentCache() for every CP. In that way you could generate your own Bone-CP Table, something like this pseudo code:

 

for ( HSpline *spline = hmc->GetHeadSpline(); spline; spline = spline->GetNextSpline() ) {

for ( HCP *cp = spline->GetHeadCP(); cp; cp = cp->GetNext() ) {

tableBoneCP.assign( cp, cp->GetSegmentCache() );

}

}

 

Will this solve your problem? Probably not, but if you'd like to share your problem in more detail, I'd like to try to help you

[aaver]

[...]when I remember correct aaver have a plugin in this direction[...]

Yes. The first version is almost ready, but I haven't got much time to spare lately and it has been almost ready for a couple of months...

 

Except for the lack of time, the new way A:M handles bitmaps has offered some additional problems, but I think I start to get the hang of that now so if nothing exceptional happens, I think I will have an alpha version ready for testing within this month.

[yoda64]

[...]how to find the assigned cp's from given bones , to build the collision hull .

Steffen,

Probably not, but if you'd like to share your problem in more detail, I'd like to try to help you

 

I'm working in this way , but I stick up , when the CP's are weightet .

I have asked for this problem in this thread in the sdk forum a week ago with a bit more information .

 

 

And back to the next version Fly,fly away

[markeh]

One thing I've found is that the newton dynamics doesn't affect objects that are 10000 cm away in either the x, y, or z but not cummulative. In other words you can have a sphere with translate y = 9000, x=9000, z=9000 and it works on that object but move any one of those values past 10000 cm and the sphere won't be acted on. This is not a problem in my eyes but just pointing it out just in case someone gets frustrated. But then I'm probably the only one that puts objects 10 trillion miles away.

[yoda64]

One thing I've found is that the newton dynamics doesn't affect objects that are 10000 cm away in either the x, y, or z but not cummulative.

 

In this case set "Use calculatet worldsize" to ON (pluginproperties in the chor tree). That solves the problem ...

[markeh]

In this case set "Use calculatet worldsize" to ON

Ah good. Thankyou.

I had a little question and maybe I'm trying to break the laws of newton physics but.....

I've been playing with the flotation properties. In the ball.mov example, it works great although I should have put some damping on.

 

What I'm trying to do in the logs.mov example is have a flood (model) come down a river and wash away a dam. The object moving horizontally is the flood water and I have it set as "static fluid surface" and I changed no settings. The logs that end up moving vertically up are set to be "Dynamic object" and I changed the density to .7. The logs are suppose to become somewhat immersed in the moving water model and then float within it but they just float straight up. Am I breaking the law or is there some other setting I should be considering.

Thankyou for all the plugins you've created by the way.

ball.mov

logs.mov

[yoda64]

Without the projectfile it's hard to find the problem ...

But a idea .

A "Static Fluidsurface" is a static object , means, his position and size is only calculatet at the simulationstart , I don't update this values during the simulation time . I think , this can be changed ,but I must first check for sideeffects. Especially if two fluidsurfaces have influence to one dynamic object .

Your projectfile where helpfull for this .

The second thing , what comes in my mind is , that the Normals from the flood model looking in the false direction .

[markeh]

It could be that one of my problems is that I probably need to add newton groups to the logs perhaps? The normals are suppose to be upward, right? If I simulate further they do seem to come to a equilibrium float but well above the water model.

I know I'm trying to use it in a fasion that it is not meant so if it's not possible, it's still a great plugin. Here is the simplified project file.

Testing.zip

[yoda64]

I think you have found a bug :-), but I must look deeper into this problem tomorrow .

[yoda64]

Yes, markeh you are right , it is not possible with this pluginversion .

Why ,

first

at this time I'm adding a new feature (generating waves for dynamic objects) that would not work if there is a moving buoyance object

second

a buoyance object can be explained as a flat plane , where the surface is at the (depends on the averaged normals) side from the model boundingbox (not turned around some, degrees only in y+,y- and so one) and in your project the dynamic bodys are starting full immersed , so they get a massive amount of lift strength , thatswhy they are springing out of the water and when they are coming to rest, they are in top of the surface from the flood wall model (and not piecewise immersed as expectet)

 

The only thing that you can try is, to use a extra grid for the watersurface at the surface values from the flood wall , but this will not solve the problem with springing :-(

[markeh]

Dynamic wave generation will be awesome! Thankyou for the very timely information (even if it can't be done in the crazy way I was trying). I think what I will do for a work around is blow up these old and splintering logs whose bark and wood will splinter as a flood comes crashing onto them.

[ssappington]

I have this compiling for the mac now! HASH will include it (both PC an Mac) in the next v13 Alpha release-

[MMZ_TimeLord]

Yoda64,

 

I'd also like to thank you for your dedication and hard work on all of these wonderful plug-ins you've crated!

 

I haven't yet played with the Newton Physics plug-in, but I've been using splitpatch for several versions now and it is a very useful tool for me, along with 'smooth'.

 

Thanks again, can't wait to see V13.0 released with your wonderful plug-ins included!

 

Thanks to Steve, Martin and all the guys at Hash, Inc. for their tireless dedication to a product that we can all afford!

 

Thanks, thanks ... and THANKS!

[MMZ_TimeLord]

Yoda64,

 

One thing to note: The new Newton Physics Plug-in breaks in V13 Alpha 4.

 

But you probably already knew that.

[yoda64]

Yes , I know this problem and I have also a version for V13 compiled already , but I will not give them out, because a problem with pluginproperties in the avaible V13 alpha , that make the plugin not usable .

[frosteternal]

I have released today a new version from the NewtonPhysics plugin for A:M .

- there are two known problems, but I would not longer wait with releasing this version (and workarounds are available), read the known problem section attentively

 

Hmmm...very nice, but for some reason it hangs at "Build NewtonWorld 50%" when using any of the explode object types (explodable model created w/ the e_rebuild plugin)

 

Anyone else have this problem?

[yoda64]

Can you send the projectfile to me , please ?

In general , the creation time is depending to the number of pieces from the explodeobject.

[martin]

Yes , I know this problem and I have also a version for V13 compiled already , but I will not give them out, because a problem with pluginproperties in the avaible V13 alpha , that make the plugin not usable .

Steffen, what are you talking about? Is this problem in the "Newton Physics" plug-in we included in the V13 release yesterday? (Are you getting Steven Sappington's emails asking you about this stuff?)

[yoda64]

Steffen, what are you talking about? Is this problem in the "Newton Physics" plug-in we included in the V13 release yesterday?

Properties attached to HChor are not showing in V13 , and yes the problem still exist also in Alpha5 (affects also the simcloth plugin, so I assume it's not plugin specific)