stand-still and the magic daemon

Discussions about the use of Realflow 3
komma
Posts: 2
Joined: Sat Mar 31, 2007 11:28 am

stand-still and the magic daemon

Postby komma » Sat Mar 31, 2007 11:33 am

Hell guys,
I'm doing the infamous splashing/drop/logo thing in realflow, and I'm running in to some problems.

The thing is my particles just won't come to a stand-still once collected.
I figure it might be a scale problem as I read on this board, and my simulation
is at scale 1.00, my logo mold is about 4 grid units high and I'm using a resoultion of
12 with maximum 50000 particles.

What is to be considered in my case, what should I try?
I dont want them to be perfectly still just not that unrealistic plasma-motion constantly.


shaun_michael
Posts: 4766
Joined: Sun Sep 10, 2006 8:04 am

stand-still and the magic daemon

Postby shaun_michael » Sat Mar 31, 2007 4:47 pm

Hi there,
Welcome aboard.

OK, first off, you'll never completely get rid of the jitter that you're seeing. You can start to minimise it by increasing the size of your objects, which will in turn increase the number of particles you need to fill it, and in turn increase simulation time. This should help reduce the noise you're getting in the particle motion.

Secondly, think smart about what your end result needs to be. I dont know exactly what your project is, but if its filling a logo with fluid then letting it fall for instance, it's likely that you'll need a certain number of frames at the start to be static before the real action starts.

Well, realflow will produce a series of sequencially numbered files for you. So how about running the simulation with the "hold" at the beginning as if realflow was doing it all perfectly, but when it comes to render time, take your mesh files and delete all but the one directly before the action starts, copy the one remaining and renumber them to make up the sequence again. This way, as you're only using one frame of the simulation, it'll be perfectly still until your action happens and as long as it's the frame directly before things start, there shouldn't be any noticeable jump.

So it really depends on what you're after at the end of the day. I wouldn't get in the mindset that realflow can not do everything for you, so considering your needs and pipeline would be a smart way to incorporate RF into your effects arsenal.

From a liquid point of view, your particle count sems awfully low. Are you restricted by using the demo version? If not, bring your particle count up somewhat, watch your pressure settings, viscosity setting and surface tension. If the water is to collide with something (a plane for instance) pay some attention to the bounce/friction/sticky and roughness settings to ensure you get the sort of splash you're after.

Good luck, and let us know how it goes.

Shaun

ENSLAVER
Posts: 58
Joined: Mon Jan 29, 2007 7:31 am

stand-still and the magic daemon

Postby ENSLAVER » Sun Apr 01, 2007 4:17 am

Also could try a K_Speed deamon (limit and keep) and key framing when you want it to come on..

You could animate down the max speed from something ridiculously high to 1 (or lower?) when you want your particles to stop moving around so much, I haven't tested this but it seems like an idea that might help

shaun_michael
Posts: 4766
Joined: Sun Sep 10, 2006 8:04 am

stand-still and the magic daemon

Postby shaun_michael » Sun Apr 01, 2007 9:56 am

Indeed the k_speed daemon may work, although it never stops the particles completely (at least not in my expereince). I suspect, again, the route to go depends on what the end result needs to be, and in this particular case, how close to the logo are we in the final render. Are we far enough away for the jittering that is still left from a k_speed daemon to not be noticeable, or not.

Seems like a few solutions, which is always good :)

Shaun

komma
Posts: 2
Joined: Sat Mar 31, 2007 11:28 am

stand-still and the magic daemon

Postby komma » Sun Apr 01, 2007 11:53 pm

Hello guys!
Thanks for the answeres, awsome dedication!
shaun_michael: good tips, like you're suggesting, I sort of went with a non-fluid simulation solution to get my particles to stop. I just imported the sequence in my NLE, time remapped it to a freeze frame and Voila! So my particle number is low ey? I thought it was big, no there are no limitation regarding software.
ENSLAVER: I have indeed used the k_speed and it was good. But my was less realflow and more everything less.

Integrating realflow in ones worklflow is very sufficent as you mention shaun_michael, I think its the key if one wishes to get functional results. Unfortunately my 3d bundle suffered from a severe crash and has not come alive since. Thanks for nothing Autodesk.
I really like this forum, and I have a feeling that as far as membership, the amount will
skyrocket as soon as people realize the power of the app. And besides, there is indeed
very few other resources on the net for Realflow.

Thanks again guys!


Return to “Realflow 3”

Who is online

Users browsing this forum: Google [Bot] and 1 guest