Re: [osg-users] Manipulating a loaded 3D model
Hi Praveena, Create a static cylindrical geometry and decorate this subgrpah with a transform node that you update dynamically to align the cylinder with the points your want. Robert. On 24 June 2012 07:45, Praveena Sarathchandra pravee...@gmail.com wrote: Hi, I want to have 2 points in 3D space and use their coordinates to align/scale a 3D cylinder model. Thus the cylinder should have a dynamic length that will be updated each time the points are moved. Illustration below: [image: Inline image 1] How to achieve this in OSG? Thanks ___ osg-users mailing list osg-users@lists.openscenegraph.org http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org image.png___ osg-users mailing list osg-users@lists.openscenegraph.org http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
[osg-users] Strange error with multiple textures in a state set
Hello,
I've been playing with multi-texturing and have encountered a situation where I
get a vector subscript out of range error. Typically, it's easy to spot the
cause of this, but this one has me a bit stumped.
I've boiled down the problem to what appears to be an issue with converting
geometry tex coord arrays to vertex attributes using the osgBuiltIn code. That
is, if I specify more than five tex coord arrays for a geometry object, the osg
built-in vertex attribute conversion code seems to have issues.
For reference, here's the function to enable osg built-in's:
Code:
void enableOsgBuiltIns( osgViewer::Viewer viewer )
{
// switch on the FnUniforms that track the modelview and projection
matrices
osgViewer::Viewer::Windows windows;
viewer.getWindows(windows);
for(osgViewer::Viewer::Windows::iterator itr = windows.begin(); itr !=
windows.end(); ++itr)
{
(*itr)-getState()-setUseModelViewAndProjectionUniforms(true);
(*itr)-getState()-setUseVertexAttributeAliasing(true);
}
};
Next is my code which adds textures to an existing geometry object:
Code:
void RenderStage::addInputTexture ( osg::Texture2D* tex, const std::string
texName = )
{
osg::StateSet* ss = this-getOrCreateStateSet();
int textureCount = ss-getNumTextureAttributeLists();
ss-setTextureAttributeAndModes( textureCount, tex );
if( usingRttQuad_ )
{
osg::Geometry* geom =
geometry_-asGeode()-getDrawable(0)-asGeometry();
geom-setTexCoordArray( textureCount, geom-getTexCoordArray(0)
);
}
};
The quad code is:
Code:
osg::Geode* RenderStage::createRttQuad()
{
osg::ref_ptrosg::Vec3Array quad_coords = new osg::Vec3Array; // vertex
coords
// counter-clockwise
quad_coords-push_back(osg::Vec3d(0, 0, -1));
quad_coords-push_back(osg::Vec3d(1, 0, -1));
quad_coords-push_back(osg::Vec3d(1, 1, -1));
quad_coords-push_back(osg::Vec3d(0, 1, -1));
osg::ref_ptrosg::Vec2Array quad_tcoords = new osg::Vec2Array; // texture
coords
quad_tcoords-push_back(osg::Vec2(0, 0));
quad_tcoords-push_back(osg::Vec2(1, 0));
quad_tcoords-push_back(osg::Vec2(1, 1));
quad_tcoords-push_back(osg::Vec2(0, 1));
osg::Geometry* quad_geom = new osg::Geometry;
quad_geom-setVertexArray(quad_coords.get());
osg::ref_ptrosg::DrawArrays quad_da = new
osg::DrawArrays(osg::PrimitiveSet::QUADS,0,4);
quad_geom-setTexCoordArray(0, quad_tcoords.get());
quad_geom-addPrimitiveSet(quad_da.get());
osg::Geode* node = new osg::Geode;
node-addDrawable( quad_geom) ;
return node;
};
So here's what I've observed:
1. If I do not enable OSG builtin's, I have no error.
2. With built-in's enabled, I can eliminate the error if I comment out the
following line in my addInputTexture() fuunction:
Code:
geom-setTexCoordArray( textureCount, geom-getTexCoordArray(0) );
The reason this line is there is because I'm using an RTT camera and,
naturally, wish to use the same quad geometry for multiple textures. Of
course, for each new texture, I also assign texture coordinates to the
corresponding unit, which is just a copy of the original texture coordinate
array pointer. Perhaps that approach is wrong, but it escapes me as to why it
appears to work fine up to tex unit four.
Granted, I could just reference the unit 0 tex coord array for all textures,
but now that this inexplicable error has appeared, I am worried there may be
more going on that I need to understand. Given that the gl_MultiTexCoord
built-in has indices ranging from 0 to 8, I am confused as to why the
osgBuiltIn's code appears to choke on texture coords assigned to units over 4...
--
Read this topic online here:
http://forum.openscenegraph.org/viewtopic.php?p=48461#48461
___
osg-users mailing list
osg-users@lists.openscenegraph.org
http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
[osg-users] Antialiasing in Volume rendering
Hi, I got aliasing problem on volume rendering. I tried to set parameter samples to 8 in class osg::GraphicsContext::Traits, but I still can see aliasing. Does anyone have an solution to stop aliasing. Thanks. Regards, Clement ___ osg-users mailing list osg-users@lists.openscenegraph.org http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
Re: [osg-users] ECEF to openGl
will look into this and probably post back with new problems -- Read this topic online here: http://forum.openscenegraph.org/viewtopic.php?p=48458#48458 ___ osg-users mailing list osg-users@lists.openscenegraph.org http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
[osg-users] Parallel Rendering Problems using NVidia GPU cards in Windows
Hi, all I want to use two GTS 250 cards to do the parallel rendering job based on OpenSceneGraph。 In my experiment, a scene full of complex models is split to 2 parts according to viewport。 The 2 parts are rendered in seperated windows (actually two slaves). I hope the LEFT part is rendered in the first screen and by the first GPU, and the RIGHT part in the second screen and by the second GPU。 The results are a little wierd and disappointing. Experiment condition: GigaByte Motherboard EX58-UD3R, 2 GTS 250 Cards, Windows XP, driver 285.58 gpuz are used to monitor the GPUs. There are 3 SLI options: Maximize 3D performance, Activate all displays, Disable SLI CASE 1: 1 screen and 1 GPU, 1 window Frame Rate: 8.4 GPU Load: 100% CASE 2: 2 screens and 2 GPU, 2 windows, the scene are split Frame Rate: 3.5 GPU1 Load: 60% GPU2 Load: 60% CASE 3: 2 screens and 2 GPU, 1 window in 1st screen Frame Rate: 8.4, 14 if only half of the scene is rendered GPU1 Load: 100% GPU2 Load: 100% CASE 1,2,3 are the same for Activate all displays and Disable SLI mode。 When I use Maximize 3D performance mode, and use the SLI connector to connect the 2 GPU cards, there is only 1 screen left。 But according to the gpuz, only 1 GPU is working。 CASE 4: 1 screen and 2 GPUs, 1 window Frame Rate: 8.4 GPU1 Load: 100% GPU2 Load: 0% CASE 5: 1 screen and 2 GPUs, 2 window Frame Rate: 3.5 GPU1 Load: 60% GPU2 Load: 0% My assumption that 2 GPUs can work independently seems not right. And the SLI is also not working。 It looks like there is always only 1 GPU working, even the other GPU is full of load。 So my problem is: 1. In windows, how can I let the 2 GPUs work independently and bind the screen with the GPU. 2. Why the SLI mode seems failed? Are there some special configurations? Any advice will be appreciated! -- Bob ___ osg-users mailing list osg-users@lists.openscenegraph.org http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
