Nuke Explosions..for Ka-boom freaks only!

#241 
 I was reading your post and I said to myself: isn't your demo reel supposed to be the most important thing in terms of getting a job in this field? I am kind of surprised that you not having an university degree plays a more important factor as to whether you get hired or not.  Somebody correct me if I am wrong.
#242

In this day and age everything counts. Most of the larger companies have HR departments that filter applicants before their demo reels even get viewed. If your credentials don’t match up (degree required) your done.

But that’s a different thread, lets blow stuff up!LOL

#243

Thanks for the reply Todd :slight_smile:

#244

Hi everyone
This thread is really epic! Lots of useful information! However, like said by others, it seems that getting the perfect nuke is somewhat time consuming and cannot be done by reading. However, I would like to spend that time because I’m a really big fan of nukes.

I have many questions, don’t know where to start g
-So my nukes always blows up when i’m tweaking too much. Sometimes, I get a mushroom that looks like a fork or it grows much more on the left side than on the right side, which looks very bad.
After reading duncans post, i may know what i’m doing wrong: i use high details except velocity and use 1-2.5 swirl and high temp turbulence/buyonc and no fuel.
Duncan recommends to use “high details on all grids”, no swirl and to add fuel to the explosion so that it releases heat on the right position to get that “swirl effect”. My english is not good so i want to ask, did i understand that correctly? Do you guys make it this way?

If yes, how do you add the fuel? Do you make a seperate emitter that is moving with the mushroom and adds fuel to the center?

-My nukes are always too tall. How do i get a bigger mushroom? I usually stop density buoyancy for the first 2-3 frames, animate density/sec from ~60 or 80 to 2 in the first 10-20 frames.

Maybe you have some other tipps for my, maybe how i can get better knowledge of what which attribute is doing. Actually, it looks for me like every attribute can affect every aspect of the simulation a little bit :frowning:

#245

hmm, I guess it all depends who, where and how you go about it.
im quite sure in London that KD could get a job doing FX if he is that good using fluids (and other fx such as cloth, particles, hard bodies) just from some of his examples…

but yeah, an insider to get him past HR could help :wink:

#246

I totaly agree with you mon (I love trols from war3 :stuck_out_tongue: xD).So this is my 1st try with some of the new 2011 adds.
So aaaaam basicly from maya 2011 I used that new lighting features and Velocity Noise(0.500).From emitter features I keyed the Density Method (I dont know it looks like that with replace its going up faster and it can give some extra burst xD could be wrong ofcourse :> )1st Add then Replace then Add again and Directional Speed .
Its not looking exactly as I wanted (it needs more detail) but… :>.Becouse its only 1 container it looks like it has that sucking real boombs have which is kinde a interesting.
link

#247

http://vimeo.com/11325679

you can see the left side the particles get cut off by the resizer. I’m don’t have resize on emitter so maybe that’s my fault. hah, makes sense that it would be. Anwyay here’s a cool test sim.

#248

Nice, that is looking very cool so far.

#249

Yea I totaly agree the resize feature is kinde a fail T_T.Nice sim and awesome res man O_O.

#250

@destruct ;
yeah I agree possibally a different setup or a different fluid container for the shooting particles will help much to complete the explosion look,
in this way it wont have to depend on the main sim & u can extend the particles directions further. just a thought !
btw very nice test dave,
Apparently I m not having enough luck to use maya 2011 much…strict deadlines at work is killing my time.
Are u using forward advection? does motion streak in the emitter helps much to smooth/smear the particle emission? I also notice if I run a longer explosion sim say atleast more than 200 frames then the overall sim gets too noisy with high deatils solve + swirl ON . So I prefer to animate it, is tht how u also do it often?
thanks
Vik

#251

not using forward advection yet. I’ve played with it but haven’t figured it out yet. I will, its on the short list of features to understand.

I turned it on but honestly couldn’t see if it was or wasn’t working.

Yep, I started to do that on this series of tests. I notice that directional noisiness as soon as 60 frames or less, but yeah it’s no good. My friend was telling my the the best fume guy he knows keys a lot of attrs and basically thats the trick know what to key when and how much. haha. But its a change in philosophy than I used to have where I never keyed anything and always let the sim work it out.

#252

That is one amazing sim destruct, regardless of the bugs in the auto-resize. Would love to see the final sim if you get it working. Thanks for sharing.

#253

jealous! :wink:

#254

http://www.youtube.com/watch?v=gybY4S75U6s

I heard (source) this has been done in maya? WTF!

#255

lol well, it was, but with Dnegs proprietary Fluids simulation/Voxel renderer…

#256

In this scene, Yellowstone eruption, made by Double Negative the software used was Maya, Houdini, Renderman & in-house tools.
This info is from the number 124 of the 3D World.
Cheers

#257

Quote:
Originally Posted by viki164
I also notice if I run a longer explosion sim say atleast more than 200 frames then the overall sim gets too noisy with high deatils solve + swirl ON . So I prefer to animate it, is tht how u also do it often?

Yep I started to do that on this series of tests. I notice that directional noisiness as soon as 60 frames or less, but yeah it’s no good. My friend was telling my the the best fume guy he knows keys a lot of attrs and basically thats the trick know what to key when and how much. haha. But its a change in philosophy than I used to have where I never keyed anything and always let the sim work it out.

  • David

Just litlle more experimenting & to address the above Issue I ran few test with my old fx_war nuke sims in maya 2010.
I really do like using high detail solve ON (all grids) as it gives great amount of detail without texturing. I have seen lot of great explosions with the high detail solve On method. but mostly most of the sims are either to short in frames or retimed to compensate the length of the overall explosion.
Anyway with all due respect I am trying to post my experience here : plz see it in full screen or hd

http://www.youtube.com/watch?v=ZCrASEYnRCk

I created this test in maya 2010 to see how long the fluid can hold its shape wihtout loosing much detail. the first three are using high deatil solve ON & the last one is high detail except velocity.

I notice in high deatial solve on + swirl gives a gr8 detail in the first few 100 frames but after running it long the sim starts getting a directional noise which makes it look little noisy especially in the end frames. so the workaround would be to key the attributes like swirl & temp turbulence by lowering the values.

I haven’t used any texture here, My idea here was to get the scale of big explosion & have it run for a longer period so I never used much higher density buoyancy like (e.g value more than 2) & higher simulation rate scale (e.g value more than 2) as it kills the scale.
I just keyed my sim scale from 1.3 to 0.6 for first 6-7 frames & then allowed the solver to do rest without keying swirl & temp turbulence.

But no doubt in most fast explosion situations using higher density buoyancy & keying simualtion rate scale often gives pretty good results where u want ur explosion to be very aggressive & fastly moves out from the camera/screen. Fluid tends to swirl a lot more at higher velocities.

I`ll be posting my tests soon on keying swirl & temp turb maybe with maya 2011 & would post my findings here on the nuke study. Most of my explosion are temperature & fuel driven & I like to keep the fuel life longer than temperature/density emission.
the max resolution here in Y is 240 & in X & z are around 190 & 160.

best,
Vik

#258

Now this is awesome man.I have an idea so lets say that you want not so much details or alot of details so aaaaaam why dont you try to key the solver quality.So lets say you start with low solver quality 20-30 and after that you increase it to smth like 150-200 or even more so it can save its shape or even now in maya 2011 you can key sub-steps as well.

Just a link to what I’m trying now :>.

#259

whats the difference in those 2? why does the 2nd one look alot better? :wink:

#260

I knew that someone would ask that :>.I used too much nParticles at the 1st one :>.Its the same fluid container with different nParticles set up :>.And 2nd one is with buoyancy 1 1st is with .1 xD.