The sample program creates a deep G-Buffer using depth peeling, which is straightforward but slow.

This directory contains a reference implementation to generate a two-layer Deep G-Buffer in a single shading pass using the technique from:

> Mara, McGuire, Nowrouzezahrai, and Luebke, Deep G-Buffers for Stable Global Illumination
> Approximation, submitted to ACM Transactions on Graphics, January 2016

If your GPU has the NV_viewport_array2 and NV_geometry_shader_passthroughextensions (NVIDIA Maxwell and later GUs), then these shaders implement the "Reproject Optimized" variant described in that paper. Otherwise, they automatically fall back to the "Reproject" variant.

To use these reference shaders in your own system, implement all of the "TODO" lines that are application specific. 

From CPU code, set up a G-Buffer with each render target having two layers 
(allocate each texture with glTexImage3D), and bind it 
to the framebuffer. Set up two viewports (via glViewportArray); the second one can be 
lower resolution. Bind all necessary uniforms and vertex attributes (see the top of the 
vertex and pixel shaders).




<!-- Markdeep: --><style class="fallback">body{visibility:hidden;white-space:pre;font-family:monospace}</style><script src="markdeep.min.js"></script><script src="https://casual-effects.com/markdeep/latest/markdeep.min.js"></script><script>window.alreadyProcessedMarkdeep||(document.body.style.visibility="visible")</script>