I want to project a movie onto a curved screen which could be modelled as an inner portion of a sphere. I want to do the geometry correction on the graphics card in real time and have a whole number of questions.
I have worked through the gpu morphing tutorial and it looks good.
What I think I need to do is start from a flat rectangular grid mesh that is textured using the movie. I then need to define the curved target mesh and write (?) a vertex shader to map between the two.
q1. How to get the input vertex buffer - I was looking at the grid nodes and Grid (DX9) is the one that takes the texture in. So what I want to is get the mesh out. However the output of the grid is a layer and not a mesh. The Grid (EX9.Geometry) outputs a mesh but I can’t see how to get the texture in.
q2. How to get the output vertex buffer - This is a similar to q1. I have been looking at the Sphere (DX9) node and it again produces a layer not a mesh. On top of this I’m not sure how I can isolate the target area of the sphere. The situation is that the projectors will be positioned about 10% of the radius from the top of the sphere and projecting onto the inner surface that is, from the point of view of the centre of the sphere, maybe 45 to 70 degrees from the vertical.
q3. Vertex shader programming - The tutorial was good but only touched the tip of the proverbial iceberg. I couldn’t find a decent reference on how to program the vertex buffers. Most of the links to the microsoft pages are out of date.
Any help would be very much appreciated.