Re: [osg-users] particle sim on GPU
Hi Crezy, Have a look at glsl_confetti.osg in the OSG dataset http://www.openscenegraph.com/osgwiki/pmwiki.php/Downloads/SampleDataset Regards, Brede On 7/11/07, Crezy Elmo [EMAIL PROTECTED] wrote: Hi all, I'm trying to implement a particle system that runs entirely on the GPU by storing positions and velocities as texture maps, and then using shaders to do the physics calculations/updates. I'm having difficulty figuring out how to set up the use of a texture (position texture) as a source of vertex data, as well as how to write the updated positions/velocities back to the textures from within the pixel shader (if this is possible?). Anyone have any pointers? I'm basically trying to do something like this: http://www.2ld.de/gdc2004/ only in GLSL w/OpenSceneGraph. Thanks for any tips! PC Magazine's 2007 editors' choice for best web mail—award-winning Windows Live Hotmail. Check it out! ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] problem applying textures to .flt models
Hi, You need texture coordinates. Regards, Brede On 7/4/07, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote: Hi all, I can apply textures to geometry, shapes and models .3ds, but not to models .flt. The code I'm using is: // set up the texture state. osg::Texture2D* texture = new osg::Texture2D; texture-setDataVariance(osg::Object::DYNAMIC); // protect from being optimized away as static state. osg::Image* textureImage = osgDB:: readImageFile((LPCTSTR)texturePath); if(!textureImage){ AfxMessageBox(Impssibile caricare la texture); return; } // Assign the texture to the image we read from file: texture-setImage (textureImage); osg::StateSet* stateset = node- getOrCreateStateSet(); stateset-setTextureAttributeAndModes(0, texture,osg::StateAttribute::ON); stateset-setMode( GL_LIGHTING, osg:: StateAttribute::OFF ); node-setStateSet(stateset); How can I do? Thank you. ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] Getting error w/ OSG v 2.0
Hi Nelson, The OpenFlight loader should handle version 13 so I tried the apache model in debug mode. There's an assert when the color index is out of range in ColorPool::getColor(index). I will submit a patch to fix this. PS. You didn't have to post the entire database collection. Regards, Brede On 7/3/07, Cysneros, Nelson SPAWAR [EMAIL PROTECTED] wrote: Thanks Robert, Tried using osgconv and got the same results. Double checked my installation and it seems to work ok on version 1.0 but not 2.0. Attached is the model file and all related files that came with it. Best Regards Nelson ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] osg-2.0 and multigen openflight 16.0 shader example world
Hi, If you refer to the Creator gallery world model I believe it only works with cg shaders. The osgviewer doesn't support cg shaders out of the box. The glsl version of the bumbmap shaders in the gallery doesn't use the tangent and binormal information so it only works for geometry in the xy plane. Regards, Brede On 7/3/07, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote: Which options must be choosen in osgviewer to see the expected result of a bump mapped world (openflight 16.0)? Jetzt Handykosten senken mit klarmobil - 14 Ct./Min.! Hier klicken http://produkte.shopping.freenet.de/handy_voip_isdn/klarmobil/index.html?pid=730025 ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] Regarding osgDem in version (OpenSceneGraph-1.9.9)
Hi, You find osgdem in VirtualPlanetBuilder. More information here: http://www.openscenegraph.com/projects/VirtualPlanetBuilder Btw. 2.0 is released. Brede On 6/22/07, Anurag M [EMAIL PROTECTED] wrote: Dear Mr. Robert, I have downloaded OpenSceneGraph-1.9.9 version, using cMake I generated MSVC7.0 solution; All 138 projects are compiled. the problem is osgdem is missing. I checked it throughly. Can you please clerify... did I missed something? reply back Please, Thanks, Anurag ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] OpenSceneGraph-2.0 released!!!
Congrats and thanks to everyone involved! Brede ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] Position prerender camera for reflection
Hi Robert, I solved my problem by flipping the scene upside-down for the prerendering. It has some limitations so I keep your suggestion in mind. Thanks, Brede On 6/12/07, Robert Osfield [EMAIL PROTECTED] wrote: Hi Brede, You effects like this I tend to have a custom Node that manages a set of internal Camera that it updates then traverses. The osgShadow::ShadowMap and osgSim::OverlayNode are examples of this. Using the approach allows you to manage multiple cameras and multiple threading easily. Robert. On 6/11/07, Brede Johansen [EMAIL PROTECTED] wrote: Hi, I'm looking for a way to update the view matrix of a scene camera based on the position of the main (matrix manipulator controlled) camera. This is needed to render the scene from under the water surface to update a reflection texture for the water. I tried a camera cull callback but the model-view is pushed before the callback is used in CullVisitor::apply(Camera). Regards, Brede ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
[osg-users] Position prerender camera for reflection
Hi, I'm looking for a way to update the view matrix of a scene camera based on the position of the main (matrix manipulator controlled) camera. This is needed to render the scene from under the water surface to update a reflection texture for the water. I tried a camera cull callback but the model-view is pushed before the callback is used in CullVisitor::apply(Camera). Regards, Brede ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] OpenFlight plug-in problems with light point nodes
Hi Chuck, On 5/21/07, Cole, Charles E. (LARC-B702)[GENEX SYSTEMS] [EMAIL PROTECTED] wrote: So, my question is ... is the OpenFlight plug-in (as far as anyone knows) fully capable and compliant with the OpenFlight specification? Or are there known issues with it and it's just not complete yet? Is there an obvious problem with how I've generated the OpenFlight file with Creator? I don't know if any application is fully compliant with the (latest) OpenFlight specification. Theres also the possibility for applications to follow the specification and produce different results. The OpenFlight format keeps evolving and the OpenFlight plugin in tries to catch up. If you don't have the time to dive into the code and fix it yourself theres a better chance someone will pick up the challenge if you provide a small test model to demonstrate the problem. Another option is to save the file in an older version like 15.7 or try to disable the palettized light points. Regards, Brede ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
[osg-users] CoordinateSystemNode and map projections
Hi all, I'm experimenting with the osgsimulation example where a plane is flying round the world. The plane model works in latitude, longitude, height and is converted to x,y,z by the EllipsoidModel in the parent CoordinateSystemNode. Now I want to replace the round earth model with a flat Mercator map but the CoordinateSystemNode doesn't support anything but the ellipsoid model. Before I make a submission I would like your comments. Regards, Brede ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] CoordinateSystemNode and map projections
Hi Paul, On 3/13/07, Paul Martz [EMAIL PROTECTED] wrote: Hi Brede -- So it sounds like EllipsoidModel projects from lat/long/alt to a 3D XYZ coordinate in space, but what you really want is more of a standard map projection: from lat/long/alt to a 2D map XY. Yes and to do that I need to replace the EllipsoidModel in the CoordinateSystemNode with UTM or some other transformation. The purpose of this is to be able to show the simulated entities on a map or on the ellipsoid model of the world. Btw. I have made a submission. Regards, Brede ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] ossimPlanet meets tiled display
Hi Norman, That's a nice setup of displays but a bit over my budget. OT but I noticed the widgets on the right side of the picture. Is that a widget library for OSG? A long time ago Marco worked on a node kit called osgWidgets but I think he got swamped with other tasks so it never materialized. Regards, Brede On 3/1/07, Norman Vine [EMAIL PROTECTED] wrote: Hi all I am visiting CalIT2 at UCSD in San Diego and have been working nn getting ossimPlanet running on their tile walls and in a cave This snap is of LandSat data from JPL laid on top of SRTM3 elevation on the Varrier http://ossim.telascience.org/ossimdata/nhv/DSCF0129.JPG This is libOssimPlanet and OpenSceneGraph running inside of Covise http://www.hlrs.de/organization/vis/covise/ Cheers Norman ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] Alpha priority
Hi Zach, I'm still trying to figure out how to automate this process since having our modelers find each instance of these cases in all of our databases would be very costly. Roberts first trick: Use a AlphaFunc setting that rejects all alpha values below 1.0. Try to add this line to your code. The node variable is the root node of your model. I think this should work because the (new) OpenFlight loader doesn't set AlphaFunc. node-getOrCreateStateSet()-setAttribute( new osg::AlphaFunc(osg::AlphaFunc::GEQUAL,0.9f) ); Regards, Brede On 3/1/07, Zach Deedler [EMAIL PROTECTED] wrote: Thanks a lot Brede. I've attached the flt file if anyone is interested (I haven't fixed it yet). I'm still trying to figure out how to automate this process since having our modelers find each instance of these cases in all of our databases would be very costly. It looks like if I turn blending off it is better also, but then I get a dark outline to the tree. Zach From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Brede Johansen Sent: Wednesday, February 28, 2007 14:20 To: osg users Subject: Re: [osg-users] Alpha priority Hi Zach, It's much easier to experiment with .osg files so I converted the model. I couldn't extract the texture from the ive so I had to grab a screenshot from osgviewer and add an alpha channel manually. Your model used linear filter, I changed this to near in the attached model. Creator saves filter modes in the attr files so make sure it's present and set correctly. Regards, Brede On 2/28/07, Zach Deedler [EMAIL PROTECTED] wrote: Hi Robert, Just modifying the texture to have 0 or 1 alpha values didn't work. Also, I looked at this same tree using VTree (a graphics API), and it doesn't have the same problem. This is weird because Multigen Creator does show the problem. So, somehow VTree is doing what we want to do, without being given any extra information. I know it doesn't just reject alpha values between 0 and 1 because we have other objects that have alpha values between 0 and 1 show up just fine (vehicles). VTree: I can try changing AlphaFunc, but how can I detect that I want to do this for only my trees, as VTree does? Anybody know? Thanks. Zach ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] the differences between the openflight loaders
Hi Melis, The new loader supports OpenFlight from version 9 or 10 and up to the latest. Most features are implemented including multitexture and shaders (GLSL). I suggest you give it a try and report if you find something missing. Also take a look at www.openscenegraph.org/osgwiki/pmwiki.php/KnowledgeBase/OpenFlight for a list of reader options. Regards, Brede On 2/18/07, melis melis [EMAIL PROTECTED] wrote: Hi, Does any of you know the differences between the old and the new versions of these openFlight loaders? do they support different openflight version standards or using different design or something else; what I could found is that old thread; http://openscenegraph.net/pipermail/osg-submissions/2006-March/001490.html, is the situation still the same? Any ideas? Thanks a lot :) Melis ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] Does thee OpenThreads license conflict with the OSG license?
Hi Robert, While you're at it, could you add the OSGPL license to the new OpenFlight loader too? I gave the green light some time ago and assumed the license would appear in the source, like magic ;-). Regards, Brede On 2/6/07, Robert Osfield [EMAIL PROTECTED] wrote: Hi Guys, On 2/5/07, Robert Osfield [EMAIL PROTECTED] wrote: I have just written to Sean to see he'd be happy changing to OpenThreads. Sean remembered the decision to go with the OSGPL, so was a bit surprised that OpenThreads was still under LGPL too. I guess me and Don must have updated everything in own projects while Sean must have assumed it be automatically be done for OpenThreads too. Sean has given me the go ahead to update the license entries to OSGPL/WxWindows license. I'll make these changes in the CVS version of OpenThreads tonight. Ack, another late night :-| Technically one can't go back and change the license of already released version of OpenThreads so these will still be under the LGPL. However, if you comply with the OSGPL then you are safe as none of the Copyright owners related to OpenThreads will come after you. If a troll does kick a stink, then point them to me or Sean, as unless you are a Copyright owner you have no authority or a legal leg to stand on. Robert. Robert. ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] OpenFlight plugin
Hi Zach, The new loader doesn't try to be clever. E.g If your flight file specifies a lit face without normals thats what you get. By default the OpenFlight loader performs some optimizations to give better performance out of the box (osgviewer) but this can be disabled with the reader option preserveFace. Rendering options in the comment field is not part of the OpenFlight specification. Without a model I can only guess but you could try to perform a calculate shading in Creator to regenerate normals. Regards, Brede On 2/4/07, Zach Deedler [EMAIL PROTECTED] wrote: Hmmm, it used to have problems finding external references and textures, but it didn't seem to have that problem just now. However, the scenes do look different. If I turn off lighting, the new plugin looks similar. However, I use the comment field to add additional rendering parameters to flt files. This does not seem to be getting into the new OpenFlight plugin. I'm assuming the new OpenFlight plugin is ignoring the comment fields. When, I get some time I can take a look at adding this. Anybody know why the new OpenFlight plugin is rendering the scene darker? Is something different in normals? Thanks. Zach -Original Message- From: [EMAIL PROTECTED] [ mailto:[EMAIL PROTECTED][EMAIL PROTECTED]] On Behalf Of Joakim Simonsson Sent: Sunday, February 04, 2007 11:37 To: osg users Subject: Re: [osg-users] OpenFlight plugin On Sun, 04 Feb 2007 17:29:11 +0100, Zach Deedler [EMAIL PROTECTED] wrote: I was wondering if people could chime in on their use of the flt plugins? Do you use? 1) old flt plugin or 2) OpenFlight plugin I've used the old plugin until some weeks ago. When caching of textures was implemented, i migrated to the new plugin. And as far as I have experienced, it works great. My flt models do not convert correctly using the OpenFlight plugin. What is not converting correctly? -- Joakim Simonsson ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] OpenFlight plugin
Hi Zach, Comments are added as node descriptions like before. The new loader doesn't parse for carriage return and line feed but adds the whole comment as one description string. Be aware that optimizations might merge or remove records loosing the comment in the process. To avoid this I do as follows. 1) Read file with reader options preserveFace and preserveObject. 2) Traverse model with comment parser vistor. 3) Tag nodes I need to preserve during optimization with setPermissibleOptimizationsForObject() 4) Optimize I'm familiar with the use of comments to add additional information to the OpenFlight format. OpenGVS had the possibility to add lights with sectors and blink sequences using comments before it was available in the OpenFlight specification. Regards, Brede On 2/4/07, Zach Deedler [EMAIL PROTECTED] wrote: Hi Brede, I know rendering options in the comment field is not part of the OpenFlight specification. But, the old loader would read the comment fields into description field of groups/objects. This allows one to write a node visitor that does use the comment field to do rendering things not in the OpenFlight specification. Using the comment fields to extend OpenFlight files has been a common method used by many. Since comment fields are in the OpenFlight specification, it would be nice if they were preserved in the scene graph, or at least an option to do so. You know what I mean? I can do this at a later date. I know you put a lot of work into the new plugin, so I want to help as much as I can. Zach ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] problem: flt with OpenFlight
Hi Yong, Opcode 101 is the light source record. I believe it's missing in the new OpenFlight plugin. I'll have a look at it this weekend. The warning message about NaN (Not a Number) values in a vertex record has nothing todo with the opcode 101 message. I plan to add better warning messages to help locate the problem in the database. Regards Brede On 2/2/07, Yong LIU [EMAIL PROTECTED] wrote: hi, I use osgcluster to open a flt file. 0.97 OSG works well. But int the latest version, flt uses OpenFlight plugin. I suffer the warning information: Unknown record, opcode=101 size=64 Warning: data error detected in VertexCNT::readRecord uv=nan nan OpenFlight doesn't define that opcode, but flt dose define it. How to figure it out? Yong ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] Problem using a Vertex Shader with TexGen
On 1/24/07, Wojciech Lewandowski [EMAIL PROTECTED] wrote: I could recommend chapter 9 Emulating OpenGL Fixed Functionality of Addison Wesley OpenGL Shading Language book. ShaderGen from 3Dlabs is another alternative. http://developer.3dlabs.com/downloads/shadergen/ Regards, Brede On 1/24/07, Wojciech Lewandowski [EMAIL PROTECTED] wrote: Vertex Shaders overrides TexGen functionality (like other transform and lighting functionality). You need to add Tex Coord generation code to shader. I briefly looked at your code and suppose that following change in shader would do it (sorry had no time to check it). I could recommend chapter 9 Emulating OpenGL Fixed Functionality of Addison Wesley OpenGL Shading Language book. uniform int transformModeX; void main() { #if 1 // Emulate TexGen Object mode at texture stage 0 gl_TexCoord[0].s = dot(gl_Vertex, gl_ObjectPlaneS[0]); gl_TexCoord[0].t = dot(gl_Vertex, gl_ObjectPlaneT[0]); gl_TexCoord[0].p = dot(gl_Vertex, gl_ObjectPlaneR[0]); gl_TexCoord[0].q = dot(gl_Vertex, gl_ObjectPlaneQ[0]); #else gl_TexCoord[0] = gl_TextureMatrix[0] * gl_MultiTexCoord0; #endif gl_FrontColor = gl_Color; gl_BackColor = gl_Color; vec4 v = vec4(gl_Vertex); // Change the test to transformModeX == 0 to enable. // This is here to help verify that this shader is actually being used. // Change the C code's gluOrtho2D call so the camera gets // a better picture. (Switch the one that is commented out.) if(transformModeX == 1) { // The multiply factor makes this log10() v.x = log2(v.x) * 0.301029995664; } gl_ClipVertex = gl_ModelViewMatrix * v; gl_Position = gl_ModelViewProjectionMatrix * v; } Cheers, Wojtek Lewandowski - Original Message - From: E. Wing [EMAIL PROTECTED] To: osg-users@openscenegraph.net Sent: Wednesday, January 24, 2007 6:32 AM Subject: [osg-users] Problem using a Vertex Shader with TexGen I am trying to produce a simple cheesy effect. I am trying to make a moving dotted line (think similar to a barber shop pole). - I start with a GL_LINE_STRIP. - Then I add a 1D texture to it so I can control how it looks and how it moves. (Currently it is an alternating off-on texture so it looks like a dotted line.) - I use osg::TexGen to generate texture coordinates for me. (In my general case, I may add or subtract points to my line so I didn't want to have to manually set TexCoords myself.) - I then use osg::TexMat to generate movement This all seems to work well enough. But my final step is where things seem to break. I then add a Vertex Shader to the mix. When I do this, my line does not render correctly. Often I seem to get a line that completely flashes off or on, rather than seeing the animated dots. To start with, I'm just trying to write a simple vertex shader that is effectively a pass-through program so it should look identical to the fixed pipeline program. I have been able to reproduce the problem on both Mac and Windows (with different video cards). So I'm guess I'm missing something important. If I remove the glTexGen stuff and set the glTexCoord manually, things seem to work, but I would like to know how to get it to work with glTexGen as well. (It makes my general case easier.) I assume I'm doing something wrong with either my shader or the way I set up TexGen. Attached is a tar-ball containing the source to two programs plus my vertex shader. One uses TexGen, the other does not. In the program that uses TexGen, there is a single line of code that will make or break the program. Search for KEY LINE. If the line is enabled, the shader is used and I get the flashing. If disabled, the program works as expected. (Also, I think I might have found a bug in OSG. As noted in the comment, if I set the stateset's setAttributeAndModes to osg::StateAttribute::OFF, the shader still gets loaded and run.) Can somebody tell me how to fix my program so it works with TexGen? This is the build line I use for Mac: g++ -D__USE_OSX_AGL_IMPLEMENTATION__ shader_notexgen.cpp -framework osg -framework osgUtil -framework osgDB -framework osgGA -framework osgProducer -framework Producer -framework OpenThreads -framework osgText -framework OpenGL (Adapt as necessary for your platform.) Thanks, Eric ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] Re: GraphicsWindowWin32 is here! :-)
Hi, I'm also having similar errors but I'm not able to trace into the code with the debugger. I suspect osgviewer application and osgviewer dll creates debug and temporary files with the same filenames. Regards Brede On 1/20/07, Adrian Egli [EMAIL PROTECTED] wrote: ATI X1600 (mobile) doesn't work, i will try to fix this bug Windows System Error #0 : GraphicsWindowWin32::setPixelFormat() - No matching pixel format found based on traits specified. Reason GraphicsWindow has not been created successfully. iewer::realize() - failed to set up any windows aegli ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] OpenFlight loader
Hi / Hei, When I use the new OpenFlight plugin, it seems that it loads every texture in the texture pallete (even those that aren't in use) for each external reference. Correct, read on demand and internal caching of textures are not implemented. I'll put it on my todo list but if you need it now I suggest you make a submission yourself. One sollotion could be that every loaded texture is cached. Have a look at the CacheHintOptions in osgDB::ReaderWriter. Regards, Brede On 1/12/07, Joakim Simonsson [EMAIL PROTECTED] wrote: Hi, I still use osg 0.9.7/0.9.8-2 to convert flt files to osg format. Since the newer ones(including Brede's) eats up all memory and makes the application crash(because there are no more addressable memory available). My flt database has hundreds of external references. Each external reference has a identical texture pallettes that contains hundreds of textures. But each external reference only use around 1-10 of them. When I use the new OpenFlight plugin, it seems that it loads every texture in the texture pallete (even those that aren't in use) for each external reference. This takes infinitely much time and memory. So before I dig deeper into this, I will ask all (especially Brede), if someone knows a quick fix to this problem. One sollotion could be that every loaded texture is cached. Comments? -- Joakim Simonsson ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] Flight DOFTransform bug?
Hi Robert, It could be the DOF local coordinate system. I recommend going to the MPI website and download the latest version of the specification for details on the DOF record. Regards, Brede On 12/6/06, Robert Nitti [EMAIL PROTECTED] wrote: Hi – I loaded an OpenFlight model with several DOFTransforms. I noticed, upon looking in the debugger, that the InversePut matrix of the DOFT(s) seem to have all its values in absolute coordinates despite several DOFTransforms being children of other DOFTs; I know they are children because I parse the scene graph and I see that one DOFT is a child of another, etc. The Transform::ReferenceFrame clearly states that the values are in RELATIVE_RF, but it's obvious by just looking at the model (and in the debugger) that the values could not possibly be relative; if they were the subordinate parts of the DOFT would be way off the model; but they are not. They are spot-on and when I animate the DOFTs, everything seems to work perfectly. All the (sub)vertices also seem to be in absolute. I'm not very familiar with the OpenFlght standard, so I don't know if it's the model that's build that way, or if it's OSG's interpretation when it imports the model. If the O.F is build that way, then why does OSG say RELATIVE_RF? If its OSG, is there not some redundant calculation being done since the values appear to all be in absolute; but they must as some point be converted to relative, no? I've provided part of the callstack of the 2 DOFTs (that are of concern) below. The cannon should be relative to the turret. And just looking at the model, there is no way the cannon is ~2 meters away (in both y, and z axis) from the turret if they are relative to each other. It just doesn't make sense; however, it looks fine in the visual – so values are somehow interpreted properly. In Z, they should be about 15cm away (ie: 2.0553188323974609 - 1.8663945198059082) from each other but its showing 2.0553188323974609m. Any help would be appreciated. Thanks - osg::Transform {_referenceFrame=RELATIVE_RF } const osg::Transform + osg::Object {_name={TURRET} _dataVariance=DYNAMIC _userData={_ptr=0x {_refMutex=??? _refCount=??? _observers=??? } } } osg::Object - _inversePut {_mat=0x0aa3fa90 } osg::Matrixd - _mat0x0aa3fa90 double [4][4] + [0] 0x0aa3fa90 double [4] + [1] 0x0aa3fab0 double [4] + [2] 0x0aa3fad0 double [4] - [3] 0x0aa3faf0 double [4] [0] -0.0068642422556877136 double [1] 0.082249075174331665double [2] 1.8663945198059082 double [3] 1. double - osg::Transform {_referenceFrame=RELATIVE_RF } const osg::Transform + osg::Object {_name={CANNON} _dataVariance=DYNAMIC _userData={_ptr=0x {_refMutex=??? _refCount=??? _observers=??? } } } osg::Object - _inversePut {_mat=0x0ab78d80 } osg::Matrixd - _mat0x0ab78d80 double [4][4] + [0] 0x0ab78d80 double [4] + [1] 0x0ab78da0 double [4] + [2] 0x0ab78dc0 double [4] - [3] 0x0ab78de0 double [4] [0] -0.0068642422556877136 double [1] 1.5622954368591309 double [2] 2.0553188323974609 double [3] 1. double ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] External Reference Problem
Hi Stuart, Could you try to replace ReadExternalsVisitor in ReaderWriterFLT.cpp with the code below and give it a try. The magic word (reader option) is cloneExternalReferences. class ReadExternalsVisitor : public osg::NodeVisitor { bool _cloneExternalReferences; osg::ref_ptrReaderWriter::Options _options; public: ReadExternalsVisitor(ReaderWriter::Options* options) : osg::NodeVisitor(osg::NodeVisitor::TRAVERSE_ALL_CHILDREN), _options(options), _cloneExternalReferences(false) { if (options) _cloneExternalReferences = (options-getOptionString().find(cloneExternalReferences)!=std::string::npos); } virtual ~ReadExternalsVisitor() {} virtual void apply(ProxyNode node) { // Transfer ownership of pools. _options-setUserData( node.getUserData() ); node.setUserData(NULL); for (unsigned int pos=0; posnode.getNumFileNames(); pos++) { std::string filename = node.getFileName(pos); // read external osg::Node* external = osgDB::readNodeFile(filename,_options.get()); if (external) { if (_cloneExternalReferences) external = dynamic_castosg::Node*(external-clone(osg::CopyOp(osg::CopyOp::DEEP_COPY_NODES))); node.addChild(external); } } } }; Brede On 11/29/06, Paul Martz [EMAIL PROTECTED] wrote: Does OSG have a similar function? node-clone(osg::CopyOp(...)) Great! Then I suggest that that is how users will expect flt external references to be implemented. Shared instancing using multiple parents was used for internal instances and replications, but not external instances. If you implement this in the OpenFlight loader, could you do so as an option? The current behavior of multiply parenting a single instance would be preferable in situations where the ext ref is not to be dynamically manipulated. -Paul ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] Is it time to put the old OpenFlight plugin to pasture?
Hi, I vote with both hands to send it to oblivion :-) Regards, Brede On 11/28/06, Robert Osfield [EMAIL PROTECTED] wrote: Hi All, We've had the new OpenFlight plugin part of the OSG since March this year, with 1.1 and 1.2 released with it as the default OpenFlight loader. After the initial flutter of activity fixing bugs, added old features things have quietened down quite a bit, if the quiet is anything to go by then perhaps we might assume that new plugin is now pretty stable. With the stability of the new OpenFlight plugin keeping the old plugin around is a burden on build and disitribution that we should safely be able to remove. As the old plugin is open source removing it for the core OSG isn't killing it, it can quite happily live out side of the core OSG, in case users need to use it as a fallback. Thoughts? Robert. ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] OpenFlight Loader Discarding Material
Hi Brian, I just upgraded from OSG 0.9.7 to 1.1.0 ... 1.2 is the latest. ...the new OpenFlight loader is discarding materials defined for geometry that has lighting turned off. This is a bug. The code : // Material if (isLit()) should be changed to something like: // Material if (materialIndex=0) in the Face class. Then you can enable lighting and create missing normals in your application and your models should be OK. 1) create a custom version of the new loader to preserve the materials 2) use the old loader 3) edit all my legacy flt files to enable lighting on the specific geometry so that the material is preserved 4) Fix the code and make a submission. Can anyone tell me if removing the materials is by design or an oversight? The idea with the new loader is to be able to recreate the information in the OpenFlight file as close as possible. By default the reader performs some optimization but a few reader options should do the trick. Regards, Brede 2006/10/5, Brian R Hill [EMAIL PROTECTED]: Folks, I just upgraded from OSG 0.9.7 to 1.1.0 and found that the new OpenFlight loader is discarding materials defined for geometry that has lighting turned off. Our application uses the materials and toggles the lighting on/off through runtime controls and we have a large number of legacy flt files modeled this way. I don't see a loader option that preservers the materials. I'm trying to decide if I want to: 1) create a custom version of the new loader to preserve the materials 2) use the old loader 3) edit all my legacy flt files to enable lighting on the specific geometry so that the material is preserved If removing the materials is by design, then I'll probably use the old loader for the short term and edit the legacy files over time. If removing the materials is an oversight, then I'll look into modifying the new loader. Can anyone tell me if removing the materials is by design or an oversight? Brian This is a PRIVATE message. If you are not the intended recipient, please delete without copying and kindly advise us by e-mail of the mistake in delivery. NOTE: Regardless of content, this e-mail shall not operate to bind CSC to any order or other contract unless pursuant to explicit written agreement or government initiative expressly permitting the use of e-mail for such purpose. ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
[osg-users] VS2005 warnings
Hi Robert, I downloaded latest OSG from CVS and got plenty of warnings like this: Warning 1 warning C4250: 'osg::ClusterCullingCallback' : inherits 'osg::NodeCallback::osg::NodeCallback::releaseGLObjects' via dominance c:\Projects\OSG_CMAKE\OpenSceneGraph\include\osg\ClusterCullingCallback 70 Classes reporting this warning are: ClusterCullingCallback GUIEventHandler CompositeGUIEventHandler All Manipulator classes Here's a link on warning C4250 http://msdn2.microsoft.com/en-us/library/6b3sy7ae.aspx Regards, Brede ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] VS2005 warnings
Looking good on my side. No warnings so far. Brede 2006/10/3, Robert Osfield [EMAIL PROTECTED]: On 10/3/06, Robert Osfield [EMAIL PROTECTED] wrote: So it should be safe to strip it from the other callbacks. I'll have a tinker. I have just done this and everything is compiling ok on my linux box, and got past any of the likely flash points so I've checked these changes in. Fingers crossed VS2005 will compile with these warnings. Robert. ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] OpenFlight flipbook animation node options
Hi Paul, In my last submission for the billboardCenter option not yet in CVS I have added two virtual functions for PrimaryRecords, pushLevel() and popLevel(). By the time popLevel() is called for the Group record the children are added so you can calculate the duration. Regards, Brede On 9/29/06, Paul Martz [EMAIL PROTECTED] wrote: Yes, you're right about the loop duration. This might be a more complex change than I originally thought, as the code can't specify the per-frame duration in the osg::Sequence until it knows how many children the Group record has. Let me spend some more time on this. -Paul -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Thom DeCarlo Sent: Friday, September 29, 2006 1:38 PM To: 'osg users' Subject: RE: [osg-users] OpenFlight flipbook animation node options Oh, rats. I figured that would be the case. I'll look forward to seeing your submission. However, I don't think that I interpret the loop duration the same way you do. I would expect the duration to be the total time to complete the whole loop, regardless of the number of subnodes in the animation, not the time to display each of the subnodes. Thanks, Thom -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Paul Martz Sent: Friday, September 29, 2006 3:19 PM To: 'osg users' Subject: RE: [osg-users] OpenFlight flipbook animation node options Hi Thom -- The current OpenFlight loader ignores the duration fields. I've modified it to pay attention to the loopDuration field, but there is no obvious way (that I can see) to make the final frame have a unique duration. So, all frames will be 2.5 seconds with my change. I'll post it to osg-submissions shortly. -Paul -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Thom DeCarlo Sent: Friday, September 29, 2006 12:52 PM To: OSG-user Subject: [osg-users] OpenFlight flipbook animation node options I've set the loop parameters on my openflight file. I have the loop duration=2.5 sec, and last frame duration=1.0 sec. However, the animation flips at a constant frame rate. I guess my question is, does the OSG understand these settings and is there some way to turn them on? Thanks, Thom ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] osgconv to ive, keep textures external?
Hi Ben, If anyone knows how to describe any of the other secret ReaderWriter options, please come forth and add them to the documentation! No need to repeat the information on the OpenFlight page so I added a link. Brede On 9/1/06, Ben Discoe [EMAIL PROTECTED] wrote: Thanks Zach. I have updated the Wiki at http://www.openscenegraph.org/osgwiki/pmwiki.php/UserGuides/ReaderWriter If anyone knows how to describe any of the other secret ReaderWriter options, please come forth and add them to the documentation! Thanks, Ben ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] New OpenFlight plugin
Hi Zach, I tried the test models on this site with success. http://www.joakims.com/osg/testdb_flt/ Regards, Brede On 9/1/06, Zach Deedler [EMAIL PROTECTED] wrote: Hello, The new OpenFlight plugin doesn't load anything with external references in other directories correctly. Anybody else see this? Zach Deedler FAAC Incorporated 1229 Oak Valley Dr. Ann Arbor, Michigan 48108 Phone: (734)761-5836 Fax: (734)761-5368 email: [EMAIL PROTECTED] http://www.faac.com ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] need help in exporting options for models
Hi Abhinandan, To verify if osg sees the floor object you can use osgconv to convert the Flight file to .osg and look at it in your favorite editor. You probably need to turn off optimization to preserve the structure of the Flight file. To disable the optimization in osgconv you can set the environment variable OSG_OPTIMIZER to OFF and to disable the optimization in the new OpenFlight reader you can set the option string preserveFace. Regards, Brede On 8/17/06, Abhinandan jain [EMAIL PROTECTED] wrote: Hi, I am trying to set seperate material property for a perticular node in a model and I could do it for turret of tank in example 8 of tutorials successfully but when i tried to it with my own model(Its a FLT model exported from maya 8) I got an debug error (I tried to change the material property of floor of a room model). can u tell me is there any care has to be taken while exporting(I had named floor objects as floor in maya itself) thanks in advance -- Regards, Abhinandan ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] strange thing in osgvertexprogram
Hi Paul, The cull callback have a *reference* to the texture matrix. Cubemap texture generation uses the normals to calculate the texture coordinates in eye space. That means the +Z face of the cubmap will always face the camera. To use cubemaps to simulate reflection you have to rotate the cubmap from eye-space to world-space and this is what the TexMat is used for. In the osgvertexprogram the cubemap is used for the skydome as well as the reflected/refracted sphere. Both objects have a TexMat state attribute that is updated in separate cull callbacks. Regards, Brede On 8/14/06, Robert Osfield [EMAIL PROTECTED] wrote: On 8/14/06, Paul Melis [EMAIL PROTECTED] wrote: Can someone (Robert :)) perhaps enlighten me of the function of the TexMatCallback() class in the osgvertexprogram example? Although it is set as a cull callback on two nodes, the result of the calculation in operator(), which is stored in _texMat, doesn't seem to be used anywhere... I'll pass the buck to the author of osgvertexprogram... Brede ;-) ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] 2 param or 3 param sched_setaffinity...
Robert, FYI Being a complete noob on CVS I usually just accepts the 'default (keep sticky)' revision when I want to get an update. To get the GetNumberOfProcessors() update I had to select 'Main trunk's head (reset sticky)'. What revision do you recommend I use for the future? Regards, Brede On 8/11/06, Robert Osfield [EMAIL PROTECTED] wrote: Thanks for the code. I have just checked in the following to OpenThreads CVS: int OpenThreads::GetNumberOfProcessors() { SYSTEM_INFO sysInfo; GetSystemInfo(sysInfo); return sysInfo.dwNumberOfProcessors; } ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] Re: Working towards OpenSceneGraph-1.1.1 release
Hi, VC7 and 8 have the possibility to disable warnings in the project settings. So when you convert the OSG project from VC6 to say VC8, disable the warning for all projects. This can be performed in on go by selecting all projects in the Solution Explorer, open the Properties dialog, select All Configurations, enter the warnings you want to disable under C/C++ - Advanced, Press OK and Save. Brede On 8/10/06, Robert Osfield [EMAIL PROTECTED] wrote: Hi Gary, On 8/10/06, Quinn, Gary [EMAIL PROTECTED] wrote: You could turn the warnings off... #pragma warning( disable : 4996 ) but give me until the end of next week to see if I can clear them without the need to do this, because I don't really like to hide warnings. I suspect the changes are going to be quite intrusive, lots if #ifdef all over the place, and this will certainly increase the changes of introducing bugs and making the whole code less maintainable. So in this case hiding the warning will keep the code more robust. Robert. ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] Geometry Normals
Hi Rolf, Try the SmoothingVisitor. Regards, Brede On 7/27/06, Rolf Anders [EMAIL PROTECTED] wrote: Hei, is there any methodes for auto generating normals for the geometry primitives ? Or do I need to calculate this my self ? -- Rolf Anders Master student at NTNU (The Norwegian University of Science and Technology) [EMAIL PROTECTED] ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] OpenSceneGraph-1.1 Release Candidate 2 upload
Hi Robert, I tried the runexamples.bat and found a couple of issues. The osgscribe example didn't look right, see attached screenshot. I traced this to the Radeon workaround in PolygonOffset, remove the workaround and the artifacts disappeared. Btw. glGetString(GL_RENDERER) return RADEON X800 XT Platinum Edition x86/SSE2 on my system. The second issue I came across was the glsl_mandelbrot.osg shader. This produced a white screen and hung the viewer. Later I found the glsl_julia.osg shader to do the same thing. The shaders are from 2005 so I may have to update my data distribution. Regards, Brede On 7/14/06, Robert Osfield [EMAIL PROTECTED] wrote: Hi All, I have just tagged OpenThreads, Producer and OpenSceneGraph for the release candidate 2, the tag is osg_v1_1_rc2_sync_release. You can download the triple set source code from : http://openscenegraph.org/osgwiki/pmwiki.php/Downloads/Downloads My plan is to go for the full release next Wednesday, only build and bug fixes will be accepted between now and Wednesday. Since we are so close to the release can you please test this new version throughly and report and build or runtime issues. Many thanks in advance, Robert. ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ scribe.gif Description: GIF image ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] About to make 1.1-rc2 last call for feedback
Hi, osgconv.cpp reports an error in VC 2005. 'setenv' :identifier not found Brede On 7/14/06, Robert Osfield [EMAIL PROTECTED] wrote: Hi All, I have now cleared my inbox box and my todo list for 1.1 release candidate 2, so I'm ready, but is the code... :-) I'll wait another a couple of hours before making the release. Please could you do a cvs update and test the build and executation and report back any success failure. Thanks in advance, Robert. ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] About to make 1.1-rc2 last call for feedback
Hi Robert, Ack. Is putenv available? _putenv This is my solution #if (_MSC_VER = 1400) // Visual Studio 2005 #include sstream int setenv(const char *name, const char *value, int /*rewrite*/) { std::stringstream sstr; sstrname'='value; return _putenv(sstr.str().c_str()); } #endif Brede On 7/14/06, Robert Osfield [EMAIL PROTECTED] wrote: Hi Brede, On 7/14/06, Brede Johansen [EMAIL PROTECTED] wrote: osgconv.cpp reports an error in VC 2005. 'setenv' :identifier not found Ack. Is putenv available? Robert. ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] SIGGRAPH 2B0ST0N6?
Hi, I'm coming too. Going to Siggraph have been on my wish-list for years. Brede On 6/7/06, Mike Weiblen [EMAIL PROTECTED] wrote: Hi all, I went ahead and made a request for BOF space. I'll post as to the details when they come in. Cheers -- mew ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] OpenSceneGraph-1.1 Release Candidate this week
Hi Martin, Did you uninstall the old driver before installing the new? It could make a difference. Regards, Brede On 7/5/06, Martin Naylor [EMAIL PROTECTED] wrote: Hi Paul, Yes it's certainly odd? Tried changing the refresh rate from 60hz to 75 hz, with still the same results. Also this is happening with osgreflect as well. I cannot see a reason why these two examples should be different from the rest of them. I am assuming its falling back to software rendering as Robert suggested, is there an easy well of telling if its software rendering? Going to try and roll back the driver shortly hopefully to prove the driver at fault. Regards Martin. -Original Message- From: [EMAIL PROTECTED] [mailto:osg-users- [EMAIL PROTECTED] On Behalf Of Paul Martz Sent: 05 July 2006 01:09 To: 'osg users' Subject: RE: [osg-users] OpenSceneGraph-1.1 Release Candidate this week I'm using NVIDIA's driver 91.31, their most recently released WinXP driver. Odd that you see this only with osganimate. Do you see a difference in the FPS if you change the refresh rate of your monitor? -Paul -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Martin Naylor Sent: Tuesday, July 04, 2006 3:48 PM To: 'osg users' Subject: RE: [osg-users] OpenSceneGraph-1.1 Release Candidate this week Hi Paul, I have set VSYNC to off in the display driver this makes all the other examples run 400 frames+, but osganimate still runs 22 fps to 72 fps with bad jittery moments. Which version are you running just out of curiosity? Thanks Martin. ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] OpenSceneGraph-1.1 Release Candidate this week
Martin, Oh, one more thing to try. If you have a laptop you might find something called SpeedStep or similar in your BIOS settings. Turning it off might help because the SpeedStep technology reduces the processor frequency to save batteries. I guess any scene with the particle system have the same problem. Brede On 7/5/06, Brede Johansen [EMAIL PROTECTED] wrote: Hi Martin, Did you uninstall the old driver before installing the new? It could make a difference. Regards, Brede On 7/5/06, Martin Naylor [EMAIL PROTECTED] wrote: Hi Paul, Yes it's certainly odd? Tried changing the refresh rate from 60hz to 75 hz, with still the same results. Also this is happening with osgreflect as well. I cannot see a reason why these two examples should be different from the rest of them. I am assuming its falling back to software rendering as Robert suggested, is there an easy well of telling if its software rendering? Going to try and roll back the driver shortly hopefully to prove the driver at fault. Regards Martin. -Original Message- From: [EMAIL PROTECTED] [mailto:osg-users- [EMAIL PROTECTED] On Behalf Of Paul Martz Sent: 05 July 2006 01:09 To: 'osg users' Subject: RE: [osg-users] OpenSceneGraph-1.1 Release Candidate this week I'm using NVIDIA's driver 91.31, their most recently released WinXP driver. Odd that you see this only with osganimate. Do you see a difference in the FPS if you change the refresh rate of your monitor? -Paul -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Martin Naylor Sent: Tuesday, July 04, 2006 3:48 PM To: 'osg users' Subject: RE: [osg-users] OpenSceneGraph-1.1 Release Candidate this week Hi Paul, I have set VSYNC to off in the display driver this makes all the other examples run 400 frames+, but osganimate still runs 22 fps to 72 fps with bad jittery moments. Which version are you running just out of curiosity? Thanks Martin. ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] Trialing the new OpenFlight as the default
Hi Paul and Robert, The a10 model is to small and pixel fill limited on my rig. I get 800 fps in fullscreen and 1600 fps in window mode. I assume the optimization pass is not included in the load time displayed by osgviewer? Yes, I believe that's correct. The new loader converts face records to Geodes to store record ID and comment. To get a decent framerate you have to enable MERGE_GEODES optimization. OSG_OPTIMIZER=DEFAULT MERGE_GEODES I think this should be the default, possibly within the OpenFlight plugin itself. For a quick fix that's probably the way to go. Even the old loader benefits from the merge optimization. I have an example going from 54 fps to 83 fps. Ideally I'd like a system where merged face records on the fly, without any recombining stages later. That's why there's a new option string FLT_preserveFace, alas it's noting more than a flag at the moment. Brede ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] Trialing the new OpenFlight as the default
Hi Robert, I'll check this in, unless there is an objection.. My only comment is on naming convention for reader options. I think it makes sense to put the filename extension in front of the keyword like dds_flip. If this is the template we should change FLT_preserveFace to flt_preserveFace. Thoughts? Brede On 7/3/06, Robert Osfield [EMAIL PROTECTED] wrote: HI Brede, On 7/3/06, Brede Johansen [EMAIL PROTECTED] wrote: That's why there's a new option string FLT_preserveFace, alas it's noting more than a flag at the moment. I've just so happen to have come across this and used it to add the MERGE_GEODES and MERGE_GEOMETRIES Optimizer pass internal to the ReaderWriterFLT.cpp for the case when there is no need to preserve the faces, and it certainly helps close the gap on performance. I'll check this in, unless there is an objection.. Robert. ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/ ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] Trialing the new OpenFlight as the default
Hi Robert, On 7/3/06, Robert Osfield [EMAIL PROTECTED] wrote: I guess the later is more in keeping with the OSG using lower case for namespaces. But perhaps we don't need the prefix at all. OK, no prefix it is. I think perhaps we need to start adding documentation for the options too Documentation is a good thing... Oops, that's going to backfire. Brede ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] Re: Trialing the new OpenFlight as the default
Hi, On 6/28/06, Robert Osfield [EMAIL PROTECTED] wrote: Brede, could you have a look at porting across to using the up to date std versions of istrstream, as you should be able to determine beter what functionality you are after, since there is a drop in replacement I don't want to go breaking runtime functionality with hacks. Done :-) Some testing required! Regards Brede ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] Catching notify output
Hi jOan, if some one has a link to osgConsole where this should be implemented I'm interested. I can't find it. Here's my notes on the subject based on posts by Marco and others. How-to forward cout and cerr to file: ofstream log(Message.log); streambuf* out = cout.rdbuf(log.rdbuf()); streambuf* err = cerr.rdbuf(log.rdbuf()); How-to forward cout and cerr to debugger using custom streambuf: Post on OSG mailing list by Marco and later patched by Ben D. #include streambuf void Log(const char *msg) { #ifdef _MSC_VER #ifdef _UNICODE wchar_t buf[1024]; MultiByteToWideChar(CP_ACP, 0, msg, -1, buf, 1024); OutputDebugString(buf); #else OutputDebugString(msg); #endif #endif } class OsgMsgTrap : public std::streambuf { public: inline virtual int_type overflow(int_type c = std::streambuf::traits_type::eof()) { if (c == std::streambuf::traits_type::eof()) return std::streambuf::traits_type::not_eof(c); char str[2]; str[0] = c; str[1] = 0; Log(str); return c; } } g_Trap; void main() { // Redirect cout messages (where OSG sends its messages) to our own log std::cout.rdbuf(g_Trap); std::cerr.rdbuf(g_Trap); } ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
[osg-users] Prerender textures
Hi, I wan't to prerender a number of textures. Once rendered the textures don't change so I'm looking for a way to generate the textures without adding prerender graphs to the scene. Regards Brede ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] ATI: Need testing for hitfixes
Hi Chris, Anyone have a chance to try this? I believe someone did a copy paste bug. Except for the title the text of the two KB items are identical. They both describe the VSYNC 100% CPU bug. At the moment I can't see the compressed textures at all regardless of driver version. I recently replaced my 9800 with X800 and I use OSG and Producer from CVS so I have to find the cause. Regarding the 100% CPU bug the new driver (hotfix_xp-2k_8-252-1-060510a-033101e.exe) was a great improvement. The CPU went from 55% with Catalyst Version 06.3 to 15% with the new driver. I have a P4 with HyperThreading enabled. Regards Brede On 5/17/06, Chris Hanson [EMAIL PROTECTED] wrote: Chris Hanson wrote: I am informed today through the grapevine (never from ATI themselves -- they don't speak to me ;) that there are hotfixes for the two remaining critical ATI driver bugs we pestered them about: https://support.ati.com/ics/support/KBAnswer.asp?questionID=22262 https://support.ati.com/ics/support/KBAnswer.asp?questionID=22263 I am not able to test these fixes and give feedback myself because my only ATI system is a laptop Radeon Mobility that normal ATI drivers won't install on. I have to wait and rely on OmegaDrivers. Could someone or several someones with access to normal ATI Radeon hardware please download these hotfixes and test them and provide feedback to ATI (via the Catalyst Crew Driver Feedback Program mentioned on the pages linked above) about whether it fixes the problems in question? I'd also like to know if the problem is resolved, so please reply here. Thanks. We're on the home stretch to getting this stuff finally working! Anyone have a chance to try this? ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/
Re: [osg-users] ATI: Need testing for hitfixes
Hi Chris and others, Following the generate mipmap+texture compression procedure on the conformance page http://www.openscenegraph.org/osgwiki/pmwiki.php/Tasks/OpenGLConformance I got different results with the optimizer set to default and off. I suspect something has changed in OSG because when I converted a file to .osg the texture file keyword was missing when optimizations was set to default. Regards Brede On 5/15/06, Chris Hanson [EMAIL PROTECTED] wrote: I am informed today through the grapevine (never from ATI themselves -- they don't speak to me ;) that there are hotfixes for the two remaining critical ATI driver bugs we pestered them about: https://support.ati.com/ics/support/KBAnswer.asp?questionID=22262 https://support.ati.com/ics/support/KBAnswer.asp?questionID=22263 I am not able to test these fixes and give feedback myself because my only ATI system is a laptop Radeon Mobility that normal ATI drivers won't install on. I have to wait and rely on OmegaDrivers. Could someone or several someones with access to normal ATI Radeon hardware please download these hotfixes and test them and provide feedback to ATI (via the Catalyst Crew Driver Feedback Program mentioned on the pages linked above) about whether it fixes the problems in question? I'd also like to know if the problem is resolved, so please reply here. Thanks. We're on the home stretch to getting this stuff finally working! ___ osg-users mailing list osg-users@openscenegraph.net http://openscenegraph.net/mailman/listinfo/osg-users http://www.openscenegraph.org/