Yeah, absolutely. I see the inherent complexity in something like that. I'll be able to really test your solution on an actual work project this coming week. So I can report back, depending on what I wind up doing.

Have a great weekend!

Wow, cool. Thanks, Dr. Sassi. I'm studying the key aspects of what you built, so I can understand more of what the operations are. Initially, it does look like the operation is generating individual polygon edges > splines and that's what you used to drive the eventual sweep.
So it's not necessarily one single spline, but many. So I might not be able (for instance) navigate an object / light around those pieces with something like the align to spline function. But overall, this still seems like a very solid option for other effects.

Cheers!

Good morning!

I'm wondering whether there is any method, to generate a parametric spline from a boolean of solid objects?
I've attached the simplest of scenes, for example. In it the figure is being cut away by a cube. I would like to be able to generate a spline outline of that red region. I would then use that spline for other effects; generate particles, sweep with other geometry, etc.

Simple enough as a still...but not if the cube is animated.

Is there a function or series of functions, which can accomplish something like this?

Thanks

Thank you, very much, Dr. Sassi.
I'm studying your invention, and I'll see what I can gain from your generous investment of time.

Being that I still need to find a solution for a current job. Maybe I can pivot this a little. In my example scene, I need to change the shape of the splines & spline mask. One thing I need to do, is soften / chamfer the harder edges; but I need to maintain the parametric nature of the object.
As it stands, the only "chamfer" solution I know for splines, is destructive. I've tried throwing the "smooth" deformer at the splines, as well as the delta mush deformer; but neither seem to have any effect on splines - only polygon objects.

Any thoughts, Dr. Sassi?

Cheers!

Thanks, Dr. Sassi. I'll watch that video, in its entirety. Once I get through the job that required figuring something like this out...

Hey,

I'm dipping my toe (barely) into object node creation. I'm wondering what the steps might be, to recreate the parametric shape in my attached scene, as nodes. Right off the bat, I didn't see any obvious node for MoGraph; so not sure whether that is simply not available in that method of object creation? Is there another node / combination of nodes, which afford similar features as MoGraph?

Thanks, for any initial guidance.

reference shape.c4d

Dr. Sassi, thank you so much, as always, for your efforts.
I worked around the issue, to get through my deadline. But I haven't solved the actual problem; which I believe may be somewhat related to the actual geometry object, which is acting as the surface emitter.
I'm attempting to build a version of that project, with a proxy object serving the same purpose, but I'm finding the results are different enough, to likely not warrant the effort.
Complicating matters (as you certainly know) is that the final banding results don't entirely manifest until the whole particle system is cached, and even then, only at final render.
I'll continue to see whether I can create a scenario which illustrates the issue, then I'll post the example project.

As a side-note: I've come to find an serious issue between the "sphere instances" and "optimized spheres" as they relate to rendered particles.
When I view the optimized sphere particles through the RS viewport, they appear as expected; just like the sphere instances.
However, upon final render, the optimized spheres tend to have an offset from the emission source, by a noticeable amount of pixels.
I'm aware that the Maxon documentation warns that there may be certain limitations for the optimized spheres, but the fact that the issue only manifests in final render, makes it extremely time-consuming to find.
But this is likely more a Redshift issue, being that it directly relates to their render tag, right?

Thanks, Dr. Sassi.
I'm moving ahead with a workaround; on deadline right now. I'll circle back, and build it a representative scene, after I finish this.

Greetings!

I'm encountering some obvious banding, in particles which are emanating from the surface of an object (mesh emitter).
I'm utilizing a Follow Spline modifier, with rather strong settings so that the emitted particles rapidly move away from emitting surface. This, is where I'm getting banding. I've tried the obvious, by radically increasing the simulations / particles / substeps.
But to no avail.
Are there any other critical settings that which I'm unaware, that would affect particle banding?

I'm sorry, this is happening in a client project, and I don't have the time to reduce it down to something I can share; but maybe in a little bit I can.

Thank you!

Good idea. While, it doesn't necessarily present itself as a bug, perhaps it's an efficiency aspect which could be improved, in the future.

Hi, Dr. Sassi.
Thanks for the suggestions / questions.
Unfortunately, I really can't get around the direct use of the "live" vertex maps, on the active objects. There is simply far too much direct control over the RS materials, to do anything else.
Over the weekend, I tested different ways to actually construct those vertex maps; some using freeze objects, others, directly linked to the polygon tags, etc.
Ultimately, no real difference in performance. Interestingly, the performance hit comes when the actual objects (with the vertex tags) are animated. When those are static, and only the camera moves, it's as fast as can be.
So, clearly, there is something "under the hood" with the way Cinema processes vertex maps, when motion is involved.

I already just posted one question, regarding the creation and modification of vertex maps.
However, I'm finding that regardless of that, I'm still encountering extreme viewport performance degradation, directly related to my objects' vertex maps.

So...is there any way to optimize such maps, or more realistically, a way to only have them active at render-time?
I'm only using them for material creations; not for deformations, nor other FX. So, they only need to be active when RS needs to render a frame.

Does this question even make any sense?

Thanks.

Morning!

Note: I posted this identical question, over on the Maxon RedShift General Forum.

This isn't entirely a RS question, however, my RS material specifically relies on using many Vertex Attribute nodes, so there is an important corollary.
I've noticed that if I drag a polygon selection tag, into an existing vertex map field, the resultant "variable tag" absolutely chokes the viewport performance. There must be some horribly inefficient process C4D is using, to generate the vertex map from the polygon selection, in realtime. However, if I simply highlight those same polygons, then generate a new vertex map, there is no more slowdown in performance.

Does this make sense to anyone else?

So, is there a way to add polygon selections to an existing vertex map, without having to just drag the polygon selection tag into the vertex fields?

Thanks!

Great. Thanks, as always, Dr. Sassi. I'll run through your notes, and your new scene, to see whether I can get a handle on the situation. Ultimately, I may just have to forge ahead, and just deal with masking the problem areas in AE, given the deadlines on the work. Then, I can try to learn more about this, after.

Thanks, Dr. Sassi. I see what you did, and that might otherwise work. However, in the actual scene in my project, the camera is trained at the sphere, so I can't quite get away with just blowing those unwanted particles to the left.
It seems like they're getting "captured" inside the collider object. So, I'm still trying to figure out what's causing that to happen.

Good morning!

I've attached a simplified scene, with a basic (C4D) particles simulation.
I'm having a devil of a time, preventing the particles from "collecting" on my collider object (sphere) at the end of the animation, when the sphere accelerates. The attached screen caps highlight those errant particles.
As it stands, I have the "collision iterations" in the collide modifier set to 10. I also have the scene "substeps" set to 10.
But those settings don't seem to be preventing the issue. Further, when I grossly exaggerate those settings, for instance, 20 units for each parameter, it still doesn't seem to help. But it sure does choke down there performance of the sim, to an unacceptable level.

So, are there other settings that which I'm unaware, that might help solve this? Or some other modifier trick, that will simply work around that?

Thanks!

particles issue.c4d

Thanks, as always, Dr. Sassi.
I rather figured there wasn't a (at least as of yet) direct way to do this sort of effect.
Noseman's example you reference, certainly is interesting. However, by the time I would rig everything up in my actual project, it's like the same amount of labor as what I'm currently doing: Emitting from the vertex map, caching, then reverse time mapping the alembic file.
I did experiment with some splines, to direct the motion, but ultimately those only deal with the distant particles. But I need great fidelity at the point that which those particles "land" on the surface of the object, as those have to perfectly match the vertex map - which is also controlling the RS shader material and displacement.

Hi,

I'm wondering whether there is any combination of (new) C4D Particles tools / modifiers, which might allow particles emitted from a separate source, to be attracted to the mesh of an object, based on a vertex map?
I know the technique for generating particles, with a mesh emitter, and using the vertex map to control the location of the emission. But there doesn't appear to be any (obvious) method to do the reverse.

My attached scene is set up with what would seem like the basics, and the minimum number of controls; but the closest I get is to just attract / stick the emitted particles to the object mesh, but with no regard for where precisely. The Surface Attract and Stick external modifiers have no provision to input a vertex map or polygon selection. So it's clearly something which would have to involve more complex math or conditional modifiers...even it is even possible at all.

Thanks!

Attract to Vertex Map.c4d

Great, thanks. I'll test those options, further. I initally did try to employ the Simulation Scene, because it seemed logical. But for some reason, my results still seemed off. But I'm sure it's because I was just rushing to test things, and didn't pay close enough attention.