View Full Version : FXWars! NUKE!: JohnnyRandom, Freeform

03 March 2010, 12:53 AM
Well better almost late than never, sorry for the WIPless thread. Not to mention I completely lost track of the time.

Single shot 2.5 kiloton Test Nuke

I literally didn't have a lot of time to spend on this and it certainly isn't without its flaws, but what the heck, thought it would be fun:

Vimeo (

03 March 2010, 03:26 AM
Hmm, I really like the little debris streamers, must of had the sim steps to what like 15 to get those looking that good?
Whenever I try to make those things they look to eradic, more like little puffs, but I'm never patient enough to wait for 4-8 hours for a single sim.
Other then that, I like it, all the research I watched you never really see much of the start so you can never get a good feel of exactly how fast it is.
There could have been a little color variation. ie. the ground shockwave. would basically be dense mud and earth but I see that more as a limitation of FumeFX then anything else.
Cough: color by source, : cough cough.
really good job incorporating it into the backplate, your good at that.
I remember watching your blur contest shots and thinking the same thing.

Good luck.

03 March 2010, 04:21 AM
Great work...
BTW you got ONE week yo tweak it.

Good luck!

03 March 2010, 04:48 AM
Thanks :)

Hmm, I really like the little debris streamers, must of had the sim steps to what like 15 to get those looking that good?
Whenever I try to make those things they look to eradic, more like little puffs, but I'm never patient enough to wait for 4-8 hours for a single sim.

The secret to streams and fume is all about the amount of particles and their age. The longer the particle lives the more fuel/smoke/whatever it is going to generate especially if it's params are set to add ;) Sim steps were at 3,the grid itself was pretty big, dimensions were something like L1200xW800xH180 at 0.7 voxel size.

edit: Oh yeah the color, it is only two grids, which as you know really limits the amount of shader tweakage of course unless I re-render it in krakatoa, I might try it, I'll have to check the fume sims to see if I cached all of the right channel data, if not I'll have to resim it and rerender it, might be more work than I have time for.

It would be great to see a shader by source, that's a pretty damn good idea :)

03 March 2010, 01:31 PM
Whoa, that's a huge grid. See that's where my problems with fume come is with the scale. Firstly I would never be able to sim something that detailed. Secondly when I go about setting up a viewport background and aligning a plane to represent the ground (for matte/shadow material) and setting up a camera. My grid ends up having to be like 500 units tall to even reach the top of the title/action safe lines.

I know I'm doing something wrong, How do you get the explosion to look like it's half a mile away with still having the grid only be 180 units tall?
My problem lies in setting up the scene. I'm working on a new one as we speak and I'm running into the same problem, the explosion is closer then I want it to be to the camera but at this point the grid is already 500x500x730 which will only allow me to go down to around 2.0 spacing before max runs out of available memory.

03 March 2010, 02:07 PM
Excellent! That turned out really well. I also like the little streamers they add a lot of detail to the sim.

03 March 2010, 04:39 PM
Thanks Mustan9 :)

The grid size was large but the adaptive didn't use the whole thing so it was quite reasonable. Also the spacing was 1.7 not 0.7, I had to open the scene to double check. I can't remember the ram usage but I do know that it didn't go over 4gb's. I rendered this in HD too and it holds up quite well, I had to scale it down because one of my post plugins isn't licensed for HD :(

The plume grid is actually quite a bit taller but it is also a LOT smaller in regards to length and width. Its dimensions were W300xL300xH500 with a spacing of 2.0, believe it or not, and it simmed under 2.5gig peak usage.

Speed is mostly what helps determine the scale, the grid even uses a very small amount of velocity dampening to help keep speed in check much like a drag operator in pflow (use it sparingly as it is really sensitive). As well as an animated temperature parameter, killing the temp after the burn finishes helps slow things down.

03 March 2010, 05:35 PM
Nice idea with the animated temperature, I just assumed you animated either the buoyancy or timescale down over time to get it to slow down the higher it got. I always assumed temperature drops off depending on the dissipation parameter after fuel burns off anyways.

03 March 2010, 06:30 PM
It definitely can/does dissipate using the Temp dissipation, it was animated from 1 to 40 in the plume grid, and I dumped the temp from 100 to -30 (ignition was set to 40 btw) so it almost stalls near the apex and too much lower it might start to fall again.

I would rahter dump the temp instead of tweaking the timescale too much, as timescale values less than .8 +- .1 can have some odd effects, you have to take a lot of other parameters into account to compensate for the change.

This is a test ( I did a few years ago that demonstrates timescale and higher FPS rates. You'll notice how you start to gain some inacuracies in the fluid, either because of the time it takes the fluid to develop or the amount of turbulance affecting the grid. If you notice the 480fps behaves way differently that the 30 and 60 fps sims even though the sim steps were raised (not drastically only 3 to the previous sims 1 step)

CGTalk Moderation
03 March 2010, 06:30 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.