Cookie on Copy SOP geo?

[Benjamin Fiske]

I'm trying to make a very basic polygonal cloud. I'm starting with an isooffset'd sphere, scattering points on that, and then copying a bunch of randomly sized polygonal spheres to that point cloud via the copy SOP.

Having never used the cookie SOP before, is there a simple way to take all of this resulting geo and cookie each into one uniform surface?

[rafaelfs]

I'm trying to make a very basic polygonal cloud. I'm starting with an isooffset'd sphere, scattering points on that, and then copying a bunch of randomly sized polygonal spheres to that point cloud via the copy SOP.

Having never used the cookie SOP before, is there a simple way to take all of this resulting geo and cookie each into one uniform surface?

You could optionally use an isooffset to convert all the spheres to an isosurface.

If you must use a cookie sop you might have to set it up inside a foreach sop cause the cookie only takes two objects at a time. it could be a bit confusing a setup, plus nothing guarantees that the cookie won't corrupt your geo at some point...

Take a look at Peter Quint's videos on the foreach sop if you decide to go down that route.

Cheers

Edited May 2, 2012 by rafaelfs

[Benjamin Fiske]

You could optionally use an isooffset to convert all the spheres to an isosurface.

If you must use a cookie sop you might have to set it up inside a foreach sop cause the cookie only takes two objects at a time. it could be a bit confusing a setup, plus nothing guarantees that the cookie won't corrupt your geo at some point...

Take a look at Peter Quint's videos on the foreach sop if you decide to go down that route.

Cheers

I thought of isooffset last night, as well, but the result was a swiss cheese mesh. I'm guessing it has a tough time navigating which part is inside and which part is outside in a big mess of geo like that.

Tried foreach last night as well, but it didn't seem (offhand) that one could grab the "next each" to use - nor would that necessarily work. Ideally this would be solved in a system that progresses from the first and ends at the last, but that's pretty un-Houdini .. tricky problem!

[Benjamin Fiske]

So, didn't think of metaballs. I think that'll work.

That said, still very curious how one would solve this problem if anyone can think of how to do it.

[3dbeing]

So, didn't think of metaballs. I think that'll work.

That said, still very curious how one would solve this problem if anyone can think of how to do it.

Here ya go benjamin. It's a bit convoluted, I feel there is a more elegant solution somewhere, but it works none the less...

volume2surface.hip

[eetu]

It's a bit convoluted, I feel there is a more elegant solution somewhere, but it works none the less...

Yes, that feels a bit convoluted

You can just copy volumes and get the isosurface.

volume2surface_ee.hip

[zarti]

without volumes .

oc , the Merge Results option on the ForEach node is turned OFF

since the cookie operation mode ( Union ) adds them all by itself .

well , the cookie ' accuracy ' .. is another song . =)

--

--

scene is a H11 file . while Clip SOP is just to see whats going on inside ..

--

cloudz.hipnc

--

.cheers

[3dbeing]

Yes, that feels a bit convoluted

Let's call it friday night derpa derps.

But I thought I tried your way and got wierd results.. must have done something wrong anyway... =P

[tjeeds]

I'll just throw this in here, if you set a metaball to Threshold Radius with a weight of 10 it almost exactly matches a sphere of the same radius when you isoOffset it in metaball mode. On my last show I had a simple asset that contained a linked metaball and sphere so I could pop it down and model with primitive spheres then select all of them and toggle to metaballs for the isoOffset.

[woodenduck]

How about this -

Sphere sop (set to polygon),

mountain sop,

isooffset(fogVolume),

scatter points,

copy metaballs,

convertmeta.

metaConvert.hipnc