VCC Blaupunkt XSI 7 ICE demo

[Leonard]

Hi guys,

Stefan Rueffer from VCC Perfect Pictures produced this great demo showing his work with ICE.

http://webrel2.softimage.com/open/video/xsi7/vcc_blaupunkt_xsi7.flv

Enjoy,

Leo

[Milosh]

brillant :)

[Photon]

I am drooling Ahhhhhhhhhhhhhhhhhhhhhh

[3rD_of_March]

Thank you Stefan for explaining the technique with that ICE setup!

[farhaad_yousefi]

exelent..!!

[xzof]

That's nice !

[Zafar Iqbal]

Wow - that was really really cool. Thanks for sharing.

[septopus]

madness!!
he makes it sound so simple. hehe

[helge]

This is one truly inspiring presentation. Thanks Stefan.

We should have a beer soon. I am drying out.

[Chris Bahry]

Amazing! Any clues on how to set the particle scale based on the area of the polygons in a mesh?

[Ronald]

Amazing! Any clues on how to set the particle scale based on the area of the polygons in a mesh?

Hi Chris,

This is pretty straightforward. Polygons have an attribute in ICE called PolygonArea you can access it and drive the size or scale attribute on the particles. You can access it by getting the polygon area at the emit location.

To do so use the Get Particle Emit Location and plug the location into the source port of a Get Data node. Then explore the Get Data to get the PolygonArea attribute. You can then plug the area in the Size port of the Emit node. I would do the square root of the area before plugging it so that the area scales linearly with the size of the particle. You can also add a Multiply by Scalar node to adjust the size.

You don't have to use the polygon area where the particle is emitted. You can, for instance, take the area of the closest polygon by using the Get Closest Location on Geometry node instead of the Get Particle Emit Location.

Hope this helps and have fun!

- Ronald

[phoenix]

Man this really nice look at behind the scene ... thanks


regards

[hamairon]

amazing video and statistics.

[Apo]

WOW! it looks great! Impressive thing that ICE can do for a production.

[liquidlight]

I love this demo. How do you get particles to respect the deformation of the character? All our attempts, the particles stay right where they were emitted, they dont move with the character.

Ronald? Any help?

Adrian

[3DMastermind]

Be sure to use 'emit from surface'.
Yes... weird since we're used to 'surface=nurbs'
I tested with a poly envelope and it works fine.

[liquidlight]

Be sure to use 'emit from surface'.
Yes... weird since we're used to 'surface=nurbs'
I tested with a poly envelope and it works fine.

I am using emit from Surface.. isnt that the standard node created when you ask for an ICE simulated particle system?

[3DMastermind]

I just received your scene in support...
Use "number of particles per second" not "total ..." in Select rate type.
Ciao

-manny

[Ronald]

I love this demo. How do you get particles to respect the deformation of the character? All our attempts, the particles stay right where they were emitted, they dont move with the character.

Ronald? Any help?

Adrian


Hi Adrian,

If you want your particles constrained to the deforming mesh you just need to set the Point Position of each particle Get Particle Emit Location -> Set Particle Position. Do this after the simulate node since it's a constraint.

Another option, if you want the particles to track the emitting surface with dynamics, you can use goals. The Move Towards Goal node with the goal type as "Use Emit Location" should do the trick for such an effect.

- Ronald

[liquidlight]

Thanks Ron,

That did the trick - another question for you - I cant seem to get the add point node to work. I want to add a point in the center of all the poly's of an object - but it seems to stay red no matter what I do. The Error says something about the creation port - but the documentation is a bit thin on what it is, and why the add point node requires it.

[Ronald]

Hi Adrian,

What does your graph look like? If I were going to emit particles at the center of polygons I would connect the polygon positions into the AddPoint position port. In the On Creation port I would connect a SetData("EmitLocation") and connect in the output location of Get Closest Location on Geometry (to the emitter). This is just a quick trick to get the Emit Location attribute set so that you can use it later on. For example, if you want to inherit velocity of the emitter and so on.

Let me know if this helps or post your graph.

- Ronald

[ebbmeister]

Hi Ronald

I've been using the AddPoint node to generate the correct amount of particles for the mesh and then place them at the polygons' exact locations.

This works fine if the ICETree is not under simulation construction mode as the points are created only once. However, when under the simulation mode, the particles are generated every frame.

How can I prevent this from happening? (as there is no rate or rate type parameters with the AddPoint node)

Cheers
Phil

[Ronald]

You could have two ICE trees. One ICE tree in the modeling region to generate the points and then put the simulation in an ICE tree that's in the simulation region. All the data you write in the modeling ICE tree is available to the simulated ICE tree. I think that's the best way to do this.

An alternative way, which I like less, is to plug your Add Point into an IF node that checks that the current time is equal to zero. Use the current frame or current time node with the local time option.

.. Ronald

[ebbmeister]

Thanks Ronald.

I had already tried the conditional approach to test against currrent frame but unfortunately it doesn't work. Once the condition is equal to false all the particles dissappear. That seems quite odd in my opinion (the false condition is not set to do anything).

I think I'll create two ICETrees as you suggest and see if that overcomes it. It would have been nice to avoid that though so that the tree could be turned into a self-contained compound at the end...

Cheers
Phil

[Ronald]

Could it be that you did the conditional method in the modeling region? That's exactly what would happen in that case.

.. Ronald

[ebbmeister]

Hmm, I think I had temporarily shifted it into Post-Sim stack to avoid loads of points being generated and, as you say, it caused it to fail. Doh - thanks for that.

I found another solution though and it works well. I'm using the EmitFromPosition node at a rate of 1 (total particles) to generate a particle for each polygon of the mesh. No problems in Sim mode now.

Thanks for all the help!
Phil