setting particle emit separation distance

[Kris Rivel]

Is there a way to specify in ICE somewhere the minimum distance a particle can be emitted from another particle?  I'm using low particle counts for instancing and I'm getting a lot of doubling up or duplicates very close together that I'd like to space out a bit more uniformly.

Kris

[Kris Rivel]

Ok so I assume I could test the distance to neighbors and use an "if" statement but what would I use to reset its surface position?  Its like I need to keep it moving around until it finds itself not within the minimum range specified.  Any ideas?

Kris

[Matt Lind]

Compute the location where you intend to emit the particle, then test to see if it meets your spatial criteria.  If so, emit the particle, otherwise do nothing. 

 

You shouldn't have to worry about moving the particle because as time progresses future particles will find the gaps and emit there.  If your system comes out a bit sparse, then up the emit count to find the gaps quicker.

 

just one way to do it.

 

Matt

 

 

 

---

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Kris Rivel
Sent: Tuesday, January 13, 2009 1:58 PM
To: Softimage List
Subject: Re: setting particle emit separation distance

[Kris Rivel]

Ah, didn't think about doing that...doh.  So what would I look up though without having an intial point or something to use as a starting point?  Would I look up a vertex or something?

Kris

[Matt Lind]

After thinking about it a bit more and realizing I answered the wrong question.... ;-)

 

If you need to move the particle to create some space, for mathematical purposes you can assume each instance is a sphere whose radius is equal to the min distance you're willing to accept and compute artificial surface normals to use as a direction vector to offset the emitted particle.  That is, if a particle gets too close to an object, find the object's artificial surface normal and use that vector to push the particle away from it.

 

Another route that I remember Olivier suggesting a few years ago to address a different problem: start with a uniform grid or other shape, then emit particles from the vertices varying their positions to the desired degree of randomness.  This will give you some control over placement while retaining the random nature.

 

Matt

 

---

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Kris Rivel
Sent: Tuesday, January 13, 2009 1:58 PM
To: Softimage List
Subject: Re: setting particle emit separation distance

[Matt Lind]

How are you emitting particles?

 

 

Matt

---

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Kris Rivel

Sent: Tuesday, January 13, 2009 2:19 PM
To: soft...@listproc.autodesk.com

Subject: Re: setting particle emit separation distance

[Ciaran Moloney]

Hi,
If you don't feel like filtering emission locations against each other: a quick and hacky approach would be to test on emission for neighboring particles with a given radius and delete if true. There are several handy compounds on the community site that do just this.You'll probably need to increase your emission rate to compensate.
 

On Tue, Jan 13, 2009 at 3:49 PM, Kris Rivel <kris...@gmail.com> wrote:

[Matt Lind]

hack method just to get you started:  emit the particle, then do the test. if it doesn't fit spatial criteria, delete it.

 

A bit wasteful in resources, but it'll allow you to test the technique.  If it works, you can spend more time figuring out how to compute the birth location and do it more cleanly later.

 

Matt

 

 

---

From: softimag...@listproc.autodesk.com [mailto:softimag...@listproc.autodesk.com] On Behalf Of Kris Rivel

Sent: Tuesday, January 13, 2009 2:19 PM
To: soft...@listproc.autodesk.com

Subject: Re: setting particle emit separation distance

[Kris Rivel]

Ah ok, thanks guys....I'll try these suggestions.

Kris