Jump to content
  • Advertisement


  • Content Count

  • Joined

  • Last visited

Community Reputation

294 Neutral

About glaeken

  • Rank
  1. glaeken

    Revert back to the old GDNet

    Any word on bookmarks being imported from the old forums? Or at least being able to browse the archives of the old forums (for areas that I don't think were carried over, e.g. IOTD)?
  2. Thanks for the help, but that's not the one I'm trying to find. I'll try to think of more identifying features.
  3. Probably a long shot, but I'm looking for a terrain/nature rendering demo that was on someone's portfolio site. I'm pretty sure I came across it here a few years ago (perhaps on iotd). From what I remember there was terrain, a creek running through the middle, may/may not have been grass, and a fairly slick ui that I believe blurred the background when you were in the menu. Yeah, I realize that pretty much describes every demo out there . The portfolio website was centered with scrollable "news posts" I believe. And there were different tabs for resume, demos, contacts, etc. It wasn't a wordpress/blogger site. (again I realize this describes almost every portfolio website ). I also remember the dude commenting on his portfolio website that he took a job at Epic Games. I've been scouring my backups and bookmarks from the last few years, but no nice yet. Any help would be awesome!
  4. Not OpenCL, but along the same track: http://research.nvidia.com/publication/high-performance-software-rasterization-gpus
  5. glaeken

    DX11 Beginning Radiosity

    If you happen to go with instant radiosity with VPLs, I've got an example of that with Nvidia's ray-tracing library Optix and DirectX: http://graphicsrunner.blogspot.com/2011/03/instant-radiosity-using-optix-and.html
  6. glaeken

    Generate Lightmaps

    IIRC there's a fairly complete example in Core Techniques and Algorithms in Game Programming.
  7. Here is how my setup basically works: 1 - draw geometry to g-buffer 2 - draw lights to light buffer 3 - combine g-buffer and light buffer and render to backbuffer 4 - draw light geometry to backbuffer with depth testing turned on As froop said, you don't need a seperate mesh for each light, but a seperate mesh for each light type ( e.g. one sphere for all point lights ).
  8. You don't have to add them to the scene. You could simply run through all your lights in your renderer and draw them after you've combined your light buffer and g-buffer.
  9. Maybe I'm misunderstanding, but why don't you just draw a mesh that represents the type of light (e.g. point light = a sphere, spot light = a cone + sphere, direct light = plane ).
  10. Thanks for your reply. I think I know what is view frustum culling. But the problem is: If I have a scene with 100 trees, I want to cull those invisible ones and I don't want to do any preprocessing. Is there any scene management algorithms can: 1. Add new objects dynamically. 2. Cull a set of objects each time instead of check the bounding box of each object one by one? [/quote] You can go along way with arranging your trees in a hierarchy and performing frustum culling on the hierarchy. This would let you cull large groups of trees that aren't in the view frustum. If you still need more culling then taking a look at the presentations mentioned above should help you out.
  11. glaeken

    Did dragonage use deferred shading?

    Yep you could easily determine the closest N lights for the current geometry. You could have different shaders for 1, 2, 3, 4 lights, etc. And you could approximate lights after the closest 4 by using a spherical harmonic.
  12. I would probably cover a good portion of the various lighting models: Phong/Blinn-Phong, Cook-Torrance, Oren-Nayar, Ward, Image Based Lighting, etc.
  13. As far as billboard reflections/refractions go, these aren't really new. It's just been slightly impractical to use them until DX11, although they're still quite expensive. You can extend these with depth imposters to get motion parallax. I was involved in research that extended depth impostors to use camera models that capture more than front facing polys. With non-pinhole cameras it's possible to see almost an entire object in a reflection from a single viewpoint. Non-pinhole impostors paper and video: http://www.cs.uiowa....ors/npa-cga.pdf I also have a demo of billboard reflections on my blog: http://graphicsrunne...-impostors.html
  14. I guess a potential follow up question would be: If RG32F is not available, then what format do people use for variance shadow map textures? When I try 16bit floating point formats I get HORRIBLE artifacts... Is the only option then to manually filter bi or trilinearly in the fragment shader? If this is the case, then I don't see the advantage over regular shadow maps and PCF... [/quote] One of the benefits is that you can run a separable filter over the shadowmap, like a separable gaussian blur.
  15. Haven't looked at the code much but are you sure you video card supports floating point filtering for RG32F? Check the caps to make sure.
  • Advertisement

Important Information

By using GameDev.net, you agree to our community Guidelines, Terms of Use, and Privacy Policy.

GameDev.net is your game development community. Create an account for your GameDev Portfolio and participate in the largest developer community in the games industry.

Sign me up!