Game development presentations - a useful reference


It's a good read, and may be why The Witcher 3 update doesn't seem to have raytraced reflections right now. They mention their reflections are WIP, seemingly taking ques from BFV's reflections.

Guessing both Elden Ring and Halo Infinite will use this as well. A GI system that runs in realtime on consoles is a big draw for most devs, though judging by the paper there's a lot of room for improvement. Probes can easily use generalized RESTIR in a temporal re-use sense, and reflections benefit highly from temporal and spatial re-use. They also need a worldspace cache scheme for volumetrics/translucency. Still it looks good for what it is.
 
It's a good read, and may be why The Witcher 3 update doesn't seem to have raytraced reflections right now. They mention their reflections are WIP, seemingly taking ques from BFV's reflections.

Guessing both Elden Ring and Halo Infinite will use this as well. A GI system that runs in realtime on consoles is a big draw for most devs, though judging by the paper there's a lot of room for improvement. Probes can easily use generalized RESTIR in a temporal re-use sense, and reflections benefit highly from temporal and spatial re-use. They also need a worldspace cache scheme for volumetrics/translucency. Still it looks good for what it is.

What’s the connection to Witcher 3?
 

Very interesting paper

We evaluate the performance of a wide set of combinations of traversal and voxel intersection testing of signed distance function grids in a path tracing setting. In addition, we present an optimized way to compute the intersection between a ray and the surface defined by trilinear interpolation of signed distances at the eight corners of a voxel. We also provide a novel way to compute continuous normals across voxels and an optimization for shadow rays. On an NVIDIA RTX 3090, the fastest method uses the GPU's ray tracing hardware to trace against a bounding volume hierarchy, built around all non-empty voxels, and then applies either an analytic cubic solver or a repeated linear interpolation for voxel intersection.
 
I am curious to see what they can do with sdf approximation. Seeing different papers it seems there is something to be done there.
 
Last edited:

Animated point based impostors! Another win, and since they can store animation states, well one could try to do something like arbitrary sized meshlet culling where instances start heavily overlapping due to tiny pixel coverage of tens of thousands of soldiers or distant animated grass or etc. Or just take inspiration from this and

979c7fd9bfb4ef7353f8a2d2972a1a67dc28ff5f8c9677fe36444fbc6c5ce44a.jpg
 
Back
Top