Good stuff! Also I just came across a 2022 paper that extends the polygonal area light paper that's referenced (link in desc). If anyone is interested, search for "Real-time Shading with Free-form Planar Area Lights using Linearly Transformed Cosines".
I want to explain one more point on shadow mapping for people.When you render the scene in the light source perspective,you get the depth texture that shares the same 2d size to your geometry because all 3d objects are represented in 2d coordinates.So that the shadow shapses on the texture will match up with your scene geometrically when you sync the view perspective with the light perspective again in vertex shader.This also means you compare the vertex z within the light perspective with the obtained depth texture and not the vertex z on your view system.
Thanks for another great lection! Im curious about one question: Can we render shadowmap from lightsourfce view volume, but with specific perspective matrix, which will have not straight near and far clips, but aligned with area of our "perspective volume"(if i can call it so) of spectator camera. And also in that case we can scale perspective matrix (of lightsource) for rendering shadowmap as trapezoid based on spectator camera far and near plane scales. Is that something that can probably work efficently or not?
And this area light will be in Touchdesigner soon!! Eheh 2023 and i still need to do this by hand since yeah render engine is what it is in TD!! 😂… but never leave the confy zone right :-)
Good stuff! Also I just came across a 2022 paper that extends the polygonal area light paper that's referenced (link in desc). If anyone is interested, search for "Real-time Shading with Free-form Planar Area Lights using Linearly Transformed Cosines".
I want to explain one more point on shadow mapping for people.When you render the scene in the light source perspective,you get the depth texture that shares the same 2d size to your geometry because all 3d objects are represented in 2d coordinates.So that the shadow shapses on the texture will match up with your scene geometrically when you sync the view perspective with the light perspective again in vertex shader.This also means you compare the vertex z within the light perspective with the obtained depth texture and not the vertex z on your view system.
Thanks for another great lection!
Im curious about one question: Can we render shadowmap from lightsourfce view volume, but with specific perspective matrix, which will have not straight near and far clips, but aligned with area of our "perspective volume"(if i can call it so) of spectator camera. And also in that case we can scale perspective matrix (of lightsource) for rendering shadowmap as trapezoid based on spectator camera far and near plane scales. Is that something that can probably work efficently or not?
And this area light will be in Touchdesigner soon!! Eheh 2023 and i still need to do this by hand since yeah render engine is what it is in TD!! 😂… but never leave the confy zone right :-)