View Full Version : New Challenge: Making Comet + Dust

04 April 2010, 11:21 PM
Anybody here know of a good approach to creating a dusty / fiery / comet trail that dissipates nicely and swirly? I was going to try and use fluids here at home with the unlimited version but at work I can't... so I'm trying to stick with particles.

I have made a comet, and to get that comet tail dusty feeling I used geometry with procedural texture work. I added flying pieces of rock to make it feel more real and I am almost satisfied with it, but I still think it would look way nicer if I can get it to work with particles instead.

here is the example

youtube vid (

04 April 2010, 12:22 AM
Try cloud type particles with a transparency map

04 April 2010, 12:26 AM
Do you have any reference and a render of what you have so far?

04 April 2010, 07:59 AM
Remember that comets fly around in space = vacuum (or solar wind to be more exact) so the particles on its tail won't swirl. The swirl only happens when they are in a medium (like air or water). Maybe the solar wind swirls on the "shadow" of the comet, but I doubt it.

What I actually mean to say is that you can achieve a realistic effect with particles only, no need for fluids. But "pretty" and "realistic" aren't always the same so you might go for fluids anyways.

04 April 2010, 05:11 PM
thanks DJ n tpalamar

and Aikiman I will post some reference tomorrow!

04 April 2010, 04:10 PM
Do you have any reference and a render of what you have so far?

here is a render

and what i mean by swirlyness / gassy dust can be seen here on the first 2 mins of the vid

armaggedon (

04 April 2010, 09:31 PM
As mentioned you dont need fluids here, but particles are definitely the way to go. Id even go as far as just using HW particles with motion blur (maybe), color accum and losts of transparency. I couldnt really see much dust debris on the vid but from what I did see those patterns could be replicated using a curl noise function or maybe some noisy turbulence from a volume field. Turn the conserve to zero apply the volume field with a low turb speed but good frequency and push play. Essentially you want the turbulence to break the particles apart just enough so that you can see the noise of the field in the particles. I would try at least 100,000 point particles and if you need more then try multi point. But the best results would be to use over 500,000 - million particles.

This gives you the general idea of the method I was explaining.

600,000 point particles

CGTalk Moderation
04 April 2010, 09:31 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.