[ODE] Physics Engines comparisons and benchmarks

erwin@erwincoumans.com erwin at erwincoumans.com
Wed Dec 20 07:49:17 MST 2006


Agreed, it would be useful to have some standard benchmark/test for 
different areas indeed. I think COLLADA Physics files can help for that. 

My main point was that in my opinion it is open source versus proprietary 
physics engines. ODE and Bullet are both frameworks with many contributions, 
some of them are shared already: both include GIMPACT and quickstep. Any 
feature that lacks in one of them can be added to the other (Zlib and BSD 
are very comparible/compatible). So if stiff motors are your focus: ODE's 
Dantzig might be one option (but it has stability drawbacks), or you might 
want to push me for adding a Featherstone-with-motors contribution to 
Bullet. A good thing about ODE as a whole is its API and community. That API 
has been ported to many languages/frameworks, including Java, Python etc. 

In proprietary/closed source I think Havok is superior to anything out 
there, but they focus on commercial game companies, not individuals. 

There are some old physics comparison tests by Chris Hecker and Jeff Lander. 
It might be a good starting point for setting up tests.

Dimitri Ognibene writes: 

> Thank you all,
> Grazie..
> Still every body tell me a different opinion.. to define a standard set of
> test to compare the various engine would be good.. ex: a test for coriolis
> force.. convex collision, many and few objects.. 
> 2006/12/20, Antonio_Martini at scee.net <Antonio_Martini at scee.net>:
>> if you want to simulate a realistic robot with related controllers i 
>> think
>> you would need a stiff solver for constraints/motors and i have the
>> impression that only ODE is providing one at the moment. Do you need to
>> model the velocity dependent forces(coriolis etc..)? if so i get the 
>> feeling
>> that Bullet and other engines are excluding those terms in favour of
>> stability. Ultimately it depends on your specific application and on the
>> level of accuracy required. 
>> you may find the following link(where ODE has been used) interesting: 
>> http://www.droidlogic.com/ 
>> cheers,
>> Antonio 
>> Antonio Martini
>> Sony Computer Entertainment Europe
>> http://www.scee.com 
>>  *"Dimitri Ognibene" <dimitri.ognibene at gmail.com>*
>> Sent by: ode-bounces at q12.org 
>> 19/12/2006 22:35
>>   To
>> vector at acm.org  cc
>> ode at q12.org  Subject
>> Re: [ODE] Physics Engines comparisons and benchmarks 
>> thank you..
>> I know that paper.. I would prefer a test suite 
>> 2006/12/19, Brian Allen <*chokma at gmail.com* <chokma at gmail.com>>:
>> You might find this useful:
>> *
>> **http://www.cs.umu.se/education/examina/Rapporter/SeuglingRolin.pdf*<htt 
>> p://www.cs.umu.se/education/examina/Rapporter/SeuglingRolin.pdf> 
>> On 12/19/06, Dimitri Ognibene < 
>> *dimitri.ognibene at gmail.com*<dimitri.ognibene at gmail.com>>
>> wrote:
>> > Where can i find some docs on physical engines and some benchmark
>> >  suite?
>> >  I was trying some by my self.. I must write a simulator for an
>> >  anthropomorphic learning robot...
>> >  but other then the API I don't know how to compare speed and accuracy
>> >  is there any benchmark suite or a standard set of test?
>> >  Or is someone interested in building one?
>> >  let me know.
>> >  I'll post this message in every forum I know related to physics
>> >  engines
>> >  C u
>> >  Dimitri Ognibene
>> > _______________________________________________
>> > ODE mailing list
>> > *ODE at q12.org* <ODE at q12.org>
>> > 
>> *http://q12.org/mailman/listinfo/ode*<http://q12.org/mailman/listinfo/ode 
>> >
>> >
>> >
>> >
>> _______________________________________________
>> ODE mailing list*
>> **ODE at q12.org* <ODE at q12.org>*
>> **http://q12.org/mailman/listinfo/ode*<http://q12.org/mailman/listinfo/od 
>> e>
>> _______________________________________________
>> ODE mailing list
>> ODE at q12.org
>> http://q12.org/mailman/listinfo/ode 
>> **********************************************************************
>> This email and any files transmitted with it are confidential and 
>> intended
>> solely for the use of the individual or entity to whom they are 
>> addressed.
>> If you have received this email in error please notify 
>> postmaster at scee.netThis footnote also confirms that this email message 
>> has been checked for all
>> known viruses.
>> ********************************************************************** 
>> Sony
>> Computer Entertainment Europe

More information about the ODE mailing list