- System Objective
I am building an interactive containment effect. The goal is to emit particles from a dynamically moving origin (driven by Face Track CHOP data) and completely trap them inside a wireframe Box. The particles should bounce off the internal walls and accumulate via turbulence. - Network Architecture
• Source: Face X/Y data is referenced to the Center of a micro Sphere SOP (Radius 0.1).
• Collision Boundary: A Box SOP (5x5x5) followed by a Primitive SOP (Vertex set to Reverse to flip normals inward).
• Routing: The Sphere connects to Input 1 (Source) of the Particle SOP. The inverted Box connects to Input 2 (Collision).
• Dynamics: Particle SOP Hit Behavior is ‘Bounce on Contact’, with extended Life Expectancy and high Turbulence. The Box is also rendered via a separate Geo COMP to visualize the physical boundaries. - The Topological Anomaly
Despite this setup, the particles completely ignore the physical boundary. They either spawn outside the box or pass right through it, exploding outward. Visually, the particle origin and the wireframe box appear spatially decoupled. - Debugging Attempts
I suspect a World Space vs. Local Space mismatch between the SOP physics and COMP rendering. I have forced the
Xform parameters (Translate/Rotate/Scale) on all external Geo COMPs to default (0,0,0 / 1,1,1), but the visual separation and collision failure persist.
How can I strictly unify the SOP collision matrix with the visual render matrix to prevent this escape?
I would be easier to study your project if you publish the .toe file.
Thank you so much for your help,here is the .toe file link.
Here is the new link.
Hi,
I just wanted to suggest my SPH tool. It should make this quite easy:
https://www.patreon.com/posts/146471881?utm_campaign=postshare_creator
https://www.patreon.com/posts/150931235?utm_campaign=postshare_creator