jb
@jbakergraphics.bsky.social
Previously Engine Programmer @idSoftware
interested in: C++/OpenGL/Vulkan/Voxels/SDFs/Pathtracing/GPU Compute/Photography/CpE/Electronics
Co-organizer of Graphics Programming Virtual Meetup
Project writeups, blog posts:
https://jbaker.graphics/index.html
interested in: C++/OpenGL/Vulkan/Voxels/SDFs/Pathtracing/GPU Compute/Photography/CpE/Electronics
Co-organizer of Graphics Programming Virtual Meetup
Project writeups, blog posts:
https://jbaker.graphics/index.html
"playing supercomputer" - it's been fun to learn more about atomics and threads in the more recent versions of C++, some really cool utilities in there now
November 16, 2025 at 4:21 AM
"playing supercomputer" - it's been fun to learn more about atomics and threads in the more recent versions of C++, some really cool utilities in there now
screenshot utility is very cool too, same job system as the particle sim, using the assignment of an atomic int to zero to cut over and start dispatching rendering work to all the threads, basically just treats the image as a linear array of memory
November 16, 2025 at 4:17 AM
screenshot utility is very cool too, same job system as the particle sim, using the assignment of an atomic int to zero to cut over and start dispatching rendering work to all the threads, basically just treats the image as a linear array of memory
Reposted by jb
The state of the art report on Monte Carlo volume rendering by Jan Novák et al. [2018] gives an excellent introduction:
www.jannovak.info/publications...
www.jannovak.info/publications...
Monte Carlo Methods for Volumetric Light Transport Simulation
Jan Novak
www.jannovak.info
November 10, 2025 at 6:48 AM
The state of the art report on Monte Carlo volume rendering by Jan Novák et al. [2018] gives an excellent introduction:
www.jannovak.info/publications...
www.jannovak.info/publications...
problem is it's losing so much detail relative to the splat renderer... but it does have occlusion and lighting using delta tracking
enough infrastructure to start playing with some more meaningful variation in sim parameters now... still the same DLA sim, placing constraints on how a given particle can attach to the set of attached points to model a crystal lattice... renderer still needs work, youtu.be/otULkFxmTLc
November 11, 2025 at 2:18 PM
problem is it's losing so much detail relative to the splat renderer... but it does have occlusion and lighting using delta tracking
server CPUs scale very differently than I am used to, it's awesome
November 10, 2025 at 2:10 AM
server CPUs scale very differently than I am used to, it's awesome
working on brightness scaling, framing
November 8, 2025 at 10:02 PM
working on brightness scaling, framing
