PDA

View Full Version : turbulance tips



vonpietro
10-17-2015, 12:21 AM
I just got turbulance fd -

any tips or reference sites you guys swear by?


I'm currently exploring making desert dust from a robot -

so far - to get the turbidity i'm using paddles to physically move the smoke around -
surprisingly it works quite well.
of course you can't move it to fast, and you have to play with interesting ways to swing the paddles -
too fast , not fast enough - there is a fine line.

turn shadows off and you get some really interesting perturbulation(is that a word)

Farhad_azer
10-18-2015, 12:46 AM
Sorry that I cant help bec I don't have TFD, but I would like to see your results at very stage. I have read somewhere it is easy to use so I would like to see how does your first results look.

vonpietro
10-18-2015, 02:10 AM
i'm still trying to understand the settings

looks like you can control how much quality - or refinement of the smoke, but the data sets get quite large

smaller swirly details seem to require larger refinement otherwise it turns into little balls



this image shows the dust trail.
its too high, and too large right now, so i have to decrease how much comes out of it.
i'm also trying to disturb the top of it using paddles- it works, its just about dialing how the paddles move. do i move them, rotate them, how fast - too fast and the smoke reacts like real smoke and dissipates too fast. too slow and there is no perturbation.
130398

gordonrobb
10-18-2015, 02:40 AM
I've had it for years, and never seem to get the time to play with it. Must get that sorted :)

Farhad_azer
10-18-2015, 09:17 PM
For god sake do more works. nothing is obvious here, maybe a darker BG could have helped to bring more details into appearance. Please don't be insulted , I am just trying to motivate you to see sth that LW can not do it. I am sure your next upload will be a lot better.

vonpietro
10-18-2015, 10:34 PM
its just the png format - it's alpha is making it white for some reason

note the ball forms at the base.

the emission point is the bottom right area where the array of ball puff clouds start

Farhad_azer
10-19-2015, 07:06 AM
Ok then upload a jpg version please. I don't think jpg has the problem of alpha channel showing. I have had similar issues with PNG. have u surfed utube for tutorials?

m.d.
10-19-2015, 02:21 PM
jpeg definitely wont have the problem of the alpha channel showing :)

m.d.
10-19-2015, 02:47 PM
i'm still trying to understand the settings

looks like you can control how much quality - or refinement of the smoke, but the data sets get quite large

smaller swirly details seem to require larger refinement otherwise it turns into little balls



this image shows the dust trail.
its too high, and too large right now, so i have to decrease how much comes out of it.
i'm also trying to disturb the top of it using paddles- it works, its just about dialing how the paddles move. do i move them, rotate them, how fast - too fast and the smoke reacts like real smoke and dissipates too fast. too slow and there is no perturbation.
130398

you can uprez it as well....simulate lower voxel size then uprez
also remember the sub grid detail on the rendering tab....that is really the key.... to displace velocity you will need to cache velocity on the simulation settings.

Since this is dust, I would emit density from the emitter....and remember to enable density it in the simulation. Use an object....maybe enable some pressure from the emmiter, and then play with turbulence and vorticity in the simulation tab.

Sub grid detail will give it that fine look without the high rez sim.....displaces things after simulation.

this is more smokey then dusty....but literally did it in 30 seconds. 15mm grid 1x1x2 box, emit density of 1....turbulence 1, vorticity 1, and sub grid noise intensity of 500mm
130430

m.d.
10-19-2015, 03:03 PM
no sub grid detail....
130431

with sub grid detail
130432

same simulation settings for the 2, uprezzing would also do similar...but sub grid detail is computationally/filesize free

vonpietro
10-19-2015, 09:49 PM
Sub grid is my new best friend

thanks for showing the difference -

i was trying to get that kind of detail using rotating paddles -
which is yielding some very very interesting results

m.d.
10-20-2015, 01:28 AM
i was trying to get that kind of detail using rotating paddles -
which is yielding some very very interesting results

That is one cool thing with turbulence....you can treat it like a traditional 'real world' effect like you are with your paddles and such and get really cool results....

Farhad_azer
10-21-2015, 07:09 AM
Waiting....

Oedo 808
10-21-2015, 12:34 PM
Mark Hennessy-Barrett did a couple of videos for the LightWave L.A. User Group (https://www.youtube.com/channel/UC6NL3EUX90auYzSB_VTUkYA) which are quite good.

Also Kat has some training up on Liberty3D (http://www.liberty3d.com), and I believe there's a Halloween sale in effect to the tune of 25% off with the coupon code of SPOOKY at the moment.

m.d.
10-21-2015, 01:11 PM
yup those are both good....

Kat's is pretty encompassing....but long....but good :)

Also these sims are pretty well the same algorithms used by Houdini and others, so you can even learn something from watching those as well....they haven't re-invented the wheel here....the basics are the same as houdini and Maya voxel fluids....so tutorials on those will help somewhat

Farhad_azer
10-23-2015, 10:17 AM
what about this one:

https://www.youtube.com/watch?v=AiHMjuCd-Mc

I envy this friend as he is going to make those nice,hot and mesmerizing fire/explosion effects.

prometheus
10-23-2015, 10:34 AM
yes..the nuke effect is very nice in shading and with itīs vacuum suck effect, he would just need a higher resolution and get away with some render artifacts, then I would say it qualifies as a high end visual nuke effect.

Michael