Jump to content

  • Log In with Google      Sign In   
  • Create Account


Member Since 16 Aug 2007
Offline Last Active Aug 14 2013 04:20 AM

Topics I've Started

Pipeline for tesselated character

03 January 2013 - 04:42 AM

Does anybody have an experience with a good working pipeline for DX11 style tessellated character?


I mean from the selection of the tessellation method, the assets creation tools, a tool to create and bake the displacement/normal map etc.


I'd like to include some (in fact, a lot of) tessellation in an engine, but I keep running into dead ends (lack of supporting tools, glitches and artifacts etc), so I'd treasure an info from somebody who actually got it working.





DX10 motion blur

04 March 2012 - 01:16 PM

I decided to implement the object based motion blur the 'DX10' way (with geometry shader, vs the velocity texture).
I read the docs for DX10 SDK sample; they were clear enough - first you amplify the geometry "on both sides" using geometry shader, then you have to blur the texture for said geometry using anisotropic filtering.

However, when I took a look at the sample, it looked fishy. I checked the source, and found out they were not amplifying geometry AND bluring the texture; they were EITHER amplifying the geometry, OR bluring the texture, which doesn't make much sense?

Does anybody know what is going on here? What is the right way to do this?

Or rather, what is the best way to implement object based motion blur (real-time) for DX11 hardware?

Paraboloid shadows mapping

10 February 2008 - 11:40 PM

Hi, after trying out several Shadow mapping algorithms, I settled for Paraboloid/Dual paraboloid shadow mapping. However, I have a problem - all the shadows I get are a bit "curved". This is not much of an issue for figures etc., but it is pretty noticeable for vertical columns casting shadows on the walls. I'd like to ask: Are the curved shadows a normal feature of Paraboloid shadow mapping? Or is it just a bug in my code? Thanks.

Cube PSM

16 August 2007 - 08:30 PM

Hello, I'm implementing a rendering system that will use quite a lot of point lights in an in-door scene. So I decided to use a Cube PSM (I excerpted the "PSM - Care and feeding" heavily :-). However, I encountered some problems with the algoritm. The biggest one so far is that in the original algoritm, they use the Zfar/(Zfar-Znear) and -Znear*Zfar/(Zfar-Znear) factors for rendering the shadows from ALL FACES of the cube, assuming they are constant. Provided that you set Znear/Zfar to be constant, the first factor is indeed constant for all faces. But the second factor is not. Now I'm thinking if it's worth the trouble to somehow find out which face is being rendered from, and choose the appropriate Znear , or store the original post-projective coordinates in the cube mape...etc. Is there any easier way out? Does anybody know how was the original algorithm supposed to work? Thx.