Nick McDonald

I finally released my article "Simulating Wind on Procedural Terrain using GPU Accelerated Lattice Boltzmann Method". I explain in detail how #LBM works and how you can implement it on your #GPU! Read it here: nickmcd.me/2022/10/01/pro… #procedural #terrain

I made an adaptive image triangulator! Here is "Las Meninas" by Diego Velazquez. source: github.com/weigert/t-pose

Rotational Field for the Iterative Circumcenter Map of a non-convex 6gon. How I made this nickmcd.me/exploring-the-… #geometry #shaders #GPU #mathematics

@HetiDev @ANSYS Definitely the boundary condition. Yeah I think that the initial and boundary conditions are not strictly compatible here, which causes this "overcorrection". I'll try something like that thanks.

@weigert__ Isn't it a matter of boundary conditions? It start with a very unstable state. When @ANSYS Fluent initializes the simulation it preinitializes the finite elements with some analytically computed values. Try to add gradient to the initial velocities and the effect should vanish.

I Implemented some #LBM cylinder vortex-shedding in only a few hours after reading the excellent book by @TimmKrueger. Does anybody have any idea why the bounce-back BC causes an initial "sound wave pulse" to be emitted from the cylinder? #CFD

Simulating wind over procedural terrain #procgen #geomorphology

Real-Time Voronoi Cubemap Shader (just for fun)

@shafikyaghmour Of course this is <cstdint.h> so I guess this makes no sense. lol

@shafikyaghmour Also there is no situation in which it would cast uint32 (i.e. B) because this would be an ambiguous operator definition, since a cast to uint32 and int would both be defined. I guess if "uint16_t operator-(const uint16_t lhs, const uint16_t rhs)" is defined, then A, otherwise C.

#172 Given the following in C++: #include <iostream> #include <cstdint> int main() { uint16_t x1 = 1; uint16_t x2 = 2; std::cout << x1 - x2 << "\n"; // What will this output? } Without checking: A. 65535 B. 4294967295 C. -1 D. Show me the results #Cplusplus #Cpppolls

@jbaker_graphics how do you make these shapes?

@lwansbrough Thanks Loch, I really appreciate that! :) Been working on my blog less recently but I'm getting the itch again. I have some stuff in the pipeline. Stay tuned!

@weigert__ just wanted to say your blog posts on terrain generation are incredible, maybe the best on the internet. please keep writing!

A 3D non-linear ODE of the form d[x,y,z]/dt = M*[x,y,z] + Q*[x*x,y*y,z*z] + R*[x*y,y*z,z*x] + c, (M,Q,R are matrices) can have a stable orbit passing through a saddle point along a separatrix, splitting it and recombining in another saddle point, with plenty of spirals. Nice.

A closeup of the action near the middle when it is still converging

Learning compute shaders - visualizing the dynamics of a 3D non-linear quadratic ODE with a particle system. Check out this sick stable orbit. It has everything you could want! #opengl #particles #shader

@marcosalvi MCMCMCMC -> Multi-Core Markov Chain Monte Carlo Marching Cubes

MCMCMC --> Markov Chain Monte Carlo Marching Cubes

Aerial shot of some terrain generated with SoilMachine. #opensource #geomorphology #procedural #terrain #procgen

I'm working on a new tool for geomorphology simulation and procedural terrain generation! SoilMachine: Multi-Layer, Multi-Sediment Hydraulic + Wind Erosion. Below with 3 sediments on a 1024x1024 grid. Src: github.com/weigert/SoilMa… #opensource #geomorphology #procedural #terrain

@Toms_mo Hey Tomas, do you still need anything? Just curious.