Search found 109 matches

by projectileman
Sat May 21, 2011 10:42 pm
Forum: General Bullet Physics Support and Feedback
Topic: Generic6DofConstraint - strange behaviour of y-axis - jitter
Replies: 4
Views: 3259

Re: Generic6DofConstraint - strange behaviour of y-axis - ji

Due of its mathematical nature, Generic6DofConstraint has a "Gimbal Lock" issue at Y axis: http://en.wikipedia.org/wiki/Gimbal_lock So both Y axis as Z should be limited to a reduced angular range of freedom: [-PI/2 to PI/2]. Only the X axis could be configured with "free" rotation (as a Hinge). So ...
by projectileman
Sat Feb 26, 2011 3:16 pm
Forum: General Bullet Physics Support and Feedback
Topic: btGImpactCollisionAlgorithm : Performance issue
Replies: 4
Views: 2366

Re: btGImpactCollisionAlgorithm : Performance issue

Use convex decomposition instead. See the class btCompoundShape and take a look to the ConvexDecompositionDemo:

http://code.google.com/p/bullet/source/ ... onDemo.cpp
by projectileman
Sun Jan 02, 2011 4:53 am
Forum: General Bullet Physics Support and Feedback
Topic: Gimpact for a character...
Replies: 2
Views: 1889

Re: Gimpact for a character...

Just use the btCompoundShape class. Is the legal way to implement composed collision bodies. May btGImpactCompoundShape is deprecated.
by projectileman
Sat Aug 07, 2010 6:53 pm
Forum: Research and development discussion about Collision Detection and Physics Simulation
Topic: OpenCL vs. CUDA
Replies: 6
Views: 7401

Re: OpenCL vs. CUDA

Concerning to OpenCL, which hardware do you recommend for this? I already have an NVIDIA geforce 8600 which could execute OpenCL tests. However, I'm want to buy a more powerful GPU for OpenCL processing. And I'm tempted to purchase the new ATI HD 5870, which seems that offer a lot of GPU power for a...
by projectileman
Wed Jul 21, 2010 12:50 am
Forum: Research and development discussion about Collision Detection and Physics Simulation
Topic: OpenCL vs. CUDA
Replies: 6
Views: 7401

Re: OpenCL vs. CUDA

Hi. On NVIDIA cards, OpenCL has nearly the same performance as CUDA. But OpenCL has more advantages in portability and it offers better integration, extensionality and modularity in software projects, because it could load and compile kernel code at runtime. Also, OpenCL is supported in more hardwar...
by projectileman
Tue Dec 15, 2009 10:49 pm
Forum: General Bullet Physics Support and Feedback
Topic: Freeing a rotational axis with a 6DofConstraint
Replies: 2
Views: 6598

Re: Freeing a rotational axis with a 6DofConstraint

Take a look into the commented text in btGeneric6DofConstraint.h file at line 222, there specifies the valid angle ranges for rotational constraints. You could only freeing the X rotational axis, and you must keep Y and Z limited with values between (-PI/2 , PI/2) . That's because the complexity of ...
by projectileman
Fri Jun 12, 2009 5:06 pm
Forum: General Bullet Physics Support and Feedback
Topic: btBvhTriangleMeshShape bug?
Replies: 6
Views: 4963

Re: btBvhTriangleMeshShape bug?

What if it is caused by Quantization? Just for trying, use a btGImpactMeshShape which has the same functionality as btBvhTriangleMeshShape. But also you have to make some setting before using it: At this file btGImpactShape.h , at line 44, change the line which is typed as : typedef btGImpactQuantiz...
by projectileman
Fri Jun 12, 2009 4:55 pm
Forum: General Bullet Physics Support and Feedback
Topic: btGImpactMeshShape bouncing for first collision?
Replies: 7
Views: 4912

Re: btGImpactMeshShape bouncing for first collision?

Hi, I've found what is causing that problem.

For solving this, just uncomment the line 301 at file btGImpactCollisionAlgorithm.h. That enables the GJK algorithm for triangle collision. It was set for using SAT by default, which in some cases has bugs.
by projectileman
Tue Jun 09, 2009 8:04 pm
Forum: General Bullet Physics Support and Feedback
Topic: btGImpactShape penetrates each other?
Replies: 2
Views: 1997

Re: btGImpactShape penetrates each other?

Hi. The bad thing about GImpact collision meshes is that they are hollow in their interior. The triangles are just an thin surface like a paper, so this surface could be penetrated easily. It's necessary to increase the mesh thickness by setting the collision margin of the GimpactShape with the meth...
by projectileman
Mon Jun 08, 2009 10:46 pm
Forum: General Bullet Physics Support and Feedback
Topic: No wheel's mass parameter in Bullet??
Replies: 2
Views: 2482

Re: No wheel's mass parameter in Bullet??

Are you trying to make a Monster Truck? If so, then the Raycast car model is not for you. Bullet has a helper class for making vehicles at a low computational cost, implementing them with a sort of raycast chasis; that means that their wheels aren't bodies at all, but instead they are just rays that...
by projectileman
Thu May 14, 2009 2:52 am
Forum: General Bullet Physics Support and Feedback
Topic: Questions about the btGeneric6DofConstraint
Replies: 9
Views: 8625

Re: Questions about the btGeneric6DofConstraint

Hi. If you want to set the rotation axis as fixed in a 6DOF Joint, you must provide epsilons for convenience. I'm not sure, but maybe zero-length ranges could have problems in the constraint solver. So this code: _constraint->setAngularLowerLimit( btVector3( -PI/4, -PI/4, 0 ) ); _constraint->setAngu...
by projectileman
Sat May 09, 2009 11:10 pm
Forum: General Bullet Physics Support and Feedback
Topic: Bullet MSVC Release build has SSE2 enabled by default
Replies: 14
Views: 8095

Re: Bullet MSVC Release build has SSE2 enabled by default

Thanks Erwin for your correction. It's a good thing that bullet can select the best approach for solving constraints at runtime, either the SSE version or the standard. That reflects a good software design. However, I have to mention that selecting different code approaches for taking advantage of S...
by projectileman
Fri May 08, 2009 3:34 am
Forum: General Bullet Physics Support and Feedback
Topic: Bullet MSVC Release build has SSE2 enabled by default
Replies: 14
Views: 8095

Re: Bullet MSVC Release build has SSE2 enabled by default

Hi reptor. I think that you're taking this issue more complicated that it requires. Just develop an application launcher that detects which kind of hardware profile the user posseses, and it should launch the respective executable which has been compiled for that target. Managing SSE extensions is n...
by projectileman
Thu May 07, 2009 5:18 am
Forum: General Bullet Physics Support and Feedback
Topic: Bullet MSVC Release build has SSE2 enabled by default
Replies: 14
Views: 8095

Re: Bullet MSVC Release build has SSE2 enabled by default

Selecting SSE capabilities dinamically could slowdown the performance of bullet seriously, since most numeric routines in Bullet are inlined. Are you suggesting function pointers to do that? Conditional selection of strategies through if-else instructions? That gives looses in performance so you cou...