particles rendered as Geo -> control over geo objects Tweet

[crasse]

Hi !
I'm working frequently with xparticles and use a lot the ability to render particles as geometry objects, works great.
However I think there is no control over the object scattering whatsoever ? (actually as in the scatter object) like adding a layer of control would be great (size variation, rotation offset and so on) currently it only rely on particles information if i'm not wrong ?
for example what if I want to randomize geometry sizes (without relying on random scale/radius information from particles)
like it's possible with the built-in Cycle 4D instance coming with xParticles (shown in the capture)

or maybe there is a way to use the Octane scatter object for particles ?

[frankmci]

IF I understand you correctly, you can use the particles as the source object for a Cloner, then vary the geometry through the Cloner settings, as well as any effectors you want to influence the Cloner.

[crasse]

I don't think I can use particles as source object for cloner in a typical xparticles setup, i'm using an octane tag on the XParticle emitter, and put geometries in the rendering tab (as shown in the capture),
so I don't know what I could do of a cloner or where it can be inserted somewhere in this workflow?
In my setup each particles from the emitter is rendered as the object(s) referenced in the particle rendering tab (from the octane tag put of the emitter).
I'm more thinking of additional parameters there (as In the cyinstance tag from cycle 4D I shown in the first post), and in fact those parameters exists in octane but only on the scatter object, and the scatter object isn't compatible with Xparticle emitter I think
I know that I can for instance offset rotation by rotating the source geo an putting into a Null (and referening the null containing the object instead of the object itself)
and also of course I can (and I do) produce those randomized values within the particles already but a second layer of parameters like this would be great, and was wondering if it was actually already possible (but hidden somewhere, or not integrated clearly in the interface)

[frankmci]

I don't think I can use particles as source object for cloner in a typical xparticles setup, i'm using an octane tag on the XParticle emitter, and put geometries in the rendering tab (as shown in the capture),
so I don't know what I could do of a cloner or where it can be inserted somewhere in this workflow?

Instead of using Octane Object Tag particles, you might try using a Cloner to place geometry onto each particle location. You just need to drag the xpEmitter into the Ojbect slot of the Cloner. Now the cloned geometry is using the X-Particles data for initial location, but can be further modified with any of the Effectors. In this case, I've added a Push Apart to prevent collision/interpenetration, as well as a Target Effector to make all the cone particles point toward the red sphere.

[crasse]

Hi frankmci !

yes that actually a method that I came across in the past days, seems to work great ! however is there any difference in calculation method to calculate/display those geos ? anyway i'll test that method and compare with tag one, on a millions+ particles setup

thanks for the help (this method also help when I want a different amount of différent geo objects as the tag method does not allow to put multiple occurence of the same object, I need to duplicate it in the scene to duplicate it in the list of the rendering tag tab in order to get various proportions of object, which is kinda stupid )

[crasse]

to stay on the xparticles topic,

does anybody experience very bad stability with octan live viewer and (heavy) xparticles scene ? i'm still on an average of 15 xdl64 crashes a day, which is very bad... Arnold and redshift are much more stable on this usage, does xparticles handling is something that need to be worked on ?
Actually, only the liveViewer crash, when I do render export it never crashes. but god does it crash. ( crashed occurs when I hit the green wheel to refresh my preview after i changed things in the scene, and only when there is a fair amount of particles rendered as geo

[frankmci]

Hi frankmci !

yes that actually a method that I came across in the past days, seems to work great ! however is there any difference in calculation method to calculate/display those geos ? anyway i'll test that method and compare with tag one, on a millions+ particles setup

I imagine that for simple particles, the Octane tag version will be faster, but for more complex geometries, I suspect it doesn't make much difference if you use Render Instance or Multi-Instance modes in the Cloner. As I understand it, Octane treats C4D Render Instances the same way it treats its own Scatter instances. But it's definitely something you should test for your situation.

[crasse]

Oh ok, thanks for those informations ! got to switch on another scene but i'll evaluate that asap !