fracture solvers comparison

General Discussions about realflow, fluid simulations,...
dynamic_boy
Posts: 62
Joined: Mon Feb 19, 2007 4:56 pm

fracture solvers comparison

Postby dynamic_boy » Wed Jul 28, 2010 10:48 pm

This is performance test for broken dominoes made with three different fracture solvers, Houdini, Fracture FX and Pulldownit,

http://vimeo.com/12278051

http://vimeo.com/13563856

http://vimeo.com/13660078

there are the same number of objects in all of them and the same distance among pieces, important to say each test have been made independently for each developer.

According to the frame rate showed, Pulldownit is 56 times faster than the first tool and 5,8 times faster than the second one computing fracture.

what shot do you think have more convincing dynamics?

cheers
Carlos Pegar

Thinkinetic-Product Manager
http://www.pulldownit.com


User avatar
tmdag
Posts: 1023
Joined: Thu Jun 28, 2007 2:22 pm
Location: New Zealand
Contact:

fracture solvers comparison

Postby tmdag » Wed Jul 28, 2010 11:06 pm

I must say that speed is not important. Or maybe I will say it in other way - more important is stability, detail you can achieve and level of control rather then speed.

Houdini was allows more into control than speed.
Within H10 and H11 (because it was available in H10 too as well as previous versions) You get a preset - digital asset with voronoi algorithm that u can use in many ways - like dynamically fracture geometry

Second thing is that it seems that you have less voronoi pieces then in the other samples ;)


Of course sometimes it is only about speed that is also why I am using Realflow and FumeFX.
And in such situations probably I would try with pull it down.
"Do not feed the trolls"
Albert 'tmdag' Szostkiewicz
FX Technical Director
Weta Digital

dynamic_boy
Posts: 62
Joined: Mon Feb 19, 2007 4:56 pm

fracture solvers comparison

Postby dynamic_boy » Wed Jul 28, 2010 11:40 pm

Im agree with you tmdag, control is a key issue, I talk about pulldownit becouse It is my tool :), with pdi you can undo the shatter operation change shatter parameters and do it again, also repeat simulation tweaking parameters as many as needed, and it counts with an incredible resume mode which allow you to pause simulation, makes changes and continue from this point.

If you watch at the number of objects/fragments label in the videos, all of them have more or less the same, even pulldownit is the one with more final fragments in simulation, what happens is the pdi fracture solver keeps the fragments together while the pieces are moving, and they break only upon impact, thats why you think there are less fragments, I gess.
Carlos Pegar

Thinkinetic-Product Manager
http://www.pulldownit.com

User avatar
tmdag
Posts: 1023
Joined: Thu Jun 28, 2007 2:22 pm
Location: New Zealand
Contact:

fracture solvers comparison

Postby tmdag » Wed Jul 28, 2010 11:54 pm

yes I know that this is Your tool and I like it but i think that You should show off hard possibilities rather than speed that is a second thing.



At this moment most problematic is shattering for many maya users that use rayfire... but they don't want to :) and stability issue when u have really hard sim. This is key issue u should show Your possibilities in real examples.
Like movie styles - breaking dome with millions of pieces, or buildings like in 2012 (but I mean like enormous number of pieces:) )
"Do not feed the trolls"
Albert 'tmdag' Szostkiewicz
FX Technical Director
Weta Digital

dynamic_boy
Posts: 62
Joined: Mon Feb 19, 2007 4:56 pm

fracture solvers comparison

Postby dynamic_boy » Thu Jul 29, 2010 12:26 am

Yeah, large scale scenes are another big issue, you count with several shots of thousands of objects/fragments in our site and vimeo, and you will see very soon pulldownit used in the big screen aswell, I cannot tell you becouse of an NDA but it is huge!
Carlos Pegar

Thinkinetic-Product Manager
http://www.pulldownit.com

User avatar
tmdag
Posts: 1023
Joined: Thu Jun 28, 2007 2:22 pm
Location: New Zealand
Contact:

fracture solvers comparison

Postby tmdag » Thu Jul 29, 2010 1:00 am

sounds fantastic! I hope to see it soon :)
"Do not feed the trolls"
Albert 'tmdag' Szostkiewicz
FX Technical Director
Weta Digital

User avatar
BrianLooney
Posts: 585
Joined: Mon Feb 11, 2008 2:19 am
Location: Sand Springs OK

fracture solvers comparison

Postby BrianLooney » Thu Jul 29, 2010 1:09 am

Personally, I see the Houdini simulation as the most realistic.
Brian Looney
Tulsa, Oklahoma

User avatar
Vitor Teixeira
Posts: 312
Joined: Thu Jul 23, 2009 1:47 pm
Location: Porto, Portugal
Contact:

fracture solvers comparison

Postby Vitor Teixeira » Thu Jul 29, 2010 1:24 am

Fracture for Maya is quite good too.
But with these new features in what order you would put Houdini in comparison with Max, Maya, RF, FumeFx.
Just a question, I never used it and I think I'm losing a lot.

dexter
Posts: 33
Joined: Thu Sep 10, 2009 1:42 am

fracture solvers comparison

Postby dexter » Fri Aug 06, 2010 10:06 pm

Very interesting test, I think the best of Houdini is its node-based workflow, aside when someone from sidefx comes to your studio and changes everything to make it work for your project ;) , it is clear by this test houdini counts with a very slow rigid bodies/fracture solver, and I still find realflow fluid solver far better than Houdini even rf4, but the node-based issue I must say, is very powerfull.

msalek
Posts: 7
Joined: Sat Aug 07, 2010 5:18 am
Contact:

fracture solvers comparison

Postby msalek » Sat Aug 07, 2010 5:27 am

Other than houdini (the god/father of destruction)

I would recommend to look into cebas tp4 (Mother of All Particle F/X & destruction ...)
http://www.cebasstation.com/index.php?pid=product&prd_id=109&feature=1052

or Cebas Volumebreaker.

http://www.cebasstation.com/index.php?pid=product&prd_id=77&feature=912

-----


rayfire anyone?

http://mirvadim.com/


Return to “General Discussion”

Who is online

Users browsing this forum: No registered users and 1 guest