Hi,

I post a new animation of fluid (mpeg1 encoded). This time, I tried to
render a more complex scene then in my previous posts. It is the filling of
a small room with a red liquid.

The simulation of the liquid surface took a little bit more than 48 hours.

The rendering done with POVRay took around 30 min per frame (the animation
has 82 frames). It is so long because I wanted to use the radiosity.
Moreover, both the liquid surface and the room walls are defined by an
isosurface.

I use a P4 3.2 Ghz with 1 Go of RAM for the simulation and the rendering.

Regards,

Fidos

Post a reply to this message

Attachments:
Download 'flow.mpg' (621 KB)

fidos wrote:
> Hi,
> 
> I post a new animation of fluid (mpeg1 encoded). This time, I tried to
> render a more complex scene then in my previous posts. It is the filling of
> a small room with a red liquid.
> 
> The simulation of the liquid surface took a little bit more than 48 hours.
> 
> The rendering done with POVRay took around 30 min per frame (the animation
> has 82 frames). It is so long because I wanted to use the radiosity.
> Moreover, both the liquid surface and the room walls are defined by an
> isosurface.
> 
> I use a P4 3.2 Ghz with 1 Go of RAM for the simulation and the rendering.
> 
> Regards,
> 
> Fidos

Nice!

Post a reply to this message

in news:web.42430adcab87a4af6da95bb50@news.povray.org fidos wrote:

> I post a new animation of fluid

Somehow it looks like liquid clay, I like that!

Ingo

Post a reply to this message

> The rendering done with POVRay took around 30 min per frame (the animation
> has 82 frames). It is so long because I wanted to use the radiosity.
> Moreover, both the liquid surface and the room walls are defined by an
> isosurface.


The radiosity worked pretty well for an animation.

I looked back at your reply to my post in your previous thread and read one
of the papers on simulating fluids with the particle level set method. Your
simulations look very good and I'm interested in playing around with this
myself. Are you going by any specific paper on that page you linked to? If
so, which one? How long did it take you to write the code? Could this be
easily extended to simulate fog/smoke?

 - Slime
 [ http://www.slimeland.com/ ]

Post a reply to this message

I love your simulations. Very good job!

One thing I'd love to see would be a rendering with a more realistic 
texture. I understand that this would increase render time a lot, but 
IMHO it's worth it ;)


Florian
-- 
camera{look_at-y*10location<8,-3,-8>*10}#local a=0;#while(a<999)sphere{
#local _=.01*a-4.99;#local p=a*.01-5;#local c=.01*a-4.995;<sin(p*pi)*5p
*10pow(p,5)*.01>sin(c*c*c*.1)+1pigment{rgb 3}}#local a=a+1;#end
/******** http://www.torfbold.com ******** http://www.imp.org ********/

Post a reply to this message

"Slime" <fak### [at] emailaddress> wrote:
> The radiosity worked pretty well for an animation.

It was long to find a good setting. The best results for quality and speed I
succeed to have was in 1 pass with a low error_bound (at .8) and a count of
200.

> I looked back at your reply to my post in your previous thread and read one
> of the papers on simulating fluids with the particle level set method. Your
> simulations look very good and I'm interested in playing around with this
> myself. Are you going by any specific paper on that page you linked to? If
> so, which one? How long did it take you to write the code? Could this be
> easily extended to simulate fog/smoke?

At the start, I read a paper of Jos Stam named "Stable Fluids" which
describe how to code in 2D a smoke animation program (if you make a web
search for "Stam Stable Fluids", you should find articles and source code).
I play with it and extend it to 3D. It was quite easy. I attach one of my
first results in DIVX format (don't look to much to the quality, it was
rendered with my own raytracer program).
The step to go to liquid motion was much more difficult. At the start, I
used mainly the paper of Nick Foster (http://www.cis.upenn.edu/~fostern/).
On the home page of Ron Fedkiw, the interesting papers are those of Frank
Losasso (for the octree idea) and Doug Enright (for the use of particules
to improve precision).

Regards,
Fidos

Post a reply to this message

Attachments:
Download 'fluid_2.avi.dat' (557 KB)

hey, that looks like something I could use for my explosion! :)

How fast is it?

cu!
-- 
ZK AKA Gaeriel
http://www.povplace.be.tf
"Surely you've heard of silicon heaven?"

Post a reply to this message

"Zeger Knaepen" <zeg### [at] studentkuleuvenacbe> wrote:
> hey, that looks like something I could use for my explosion! :)
>
> How fast is it?
>

I don't remember how long it tooks to do the simulation for the smoke
animation, In any case, it is much more faster to simulate smoke than to
simulate liquid.

Regards,
Fidos

Post a reply to this message

Source, Source!  <G>

It looks very impressive.  I think the smoke anim in particular would make a 
great mushroom cloud for an explosion.

Are you exporting each frame or is this a POVRay modification?

        == John ==

Post a reply to this message

> Are you exporting each frame or is this a POVRay modification?

I cut and past an explanation I wrote in a previous post :

The rendering is done in 2 steps.

First, I used a program that I wrote in C for the simulation of the liquid
surface. I use the level set method and the solving of the NS equations.
This method can give good results but is quite long (see for exemple
http://graphics.stanford.edu/~fedkiw/ for impressive animations). For each
frame, the program generate an isosurface file in the dfd format. For the
simulation, the environement is defined by a isosurface file (in dfd
format) which is loaded at the start of the program.

When I have all my frames simulated, I do the rendering in POVRay. I use
MagaPov 1.0 with the patch for density file :
http://staff.aist.go.jp/r-suzuki/e/povray/iso/df_body.htm

Regards,

Fidos

Post a reply to this message