Kos

|          Image format (sized)         |      Unsized       | Compr |    Pixel format    |             Pixel type            |
|---------------------------------------|--------------------|-------|--------------------|-----------------------------------|
| GL_R8                                 | GL_RED             | False | GL_RED             | GL_UNSIGNED_BYTE                  |
| GL_R8_SNORM                           | GL_RED             | False | GL_RED             | GL_BYTE                           |
| GL_R16                                | GL_RED             | False | GL_RED             | GL_UNSIGNED_SHORT                 |
| GL_R16_SNORM                          | GL_RED             | False | GL_RED             | GL_SHORT                          |
| GL_R32F                               | GL_RED             | False | GL_RED             | GL_FLOAT                          |
| GL_R8I                                | GL_RED             | False | GL_RED_INTEGER     | GL_INT                            |

MulattoKid

Intro #1

I recently decided that I should start to work on my own path tracing "engine" and so here we are. I will try to share as much as possible without it just becoming too much to read. Hopefully what I write will be of some use to others who might run into the same problems I do during this process. Just as a quick disclaimer, here is my setup which I plan to use for throughout the project (hate it or love it, this is what I'm working with):

• Windows 10
• Intel i7-6700K @4.2GHz
• Nvidia GTX 970 (4GB)
• 16GB RAM
• Microsoft Visual Studio Community 2017
• CUDA Toolkit 9.0
• OpenGL 4.3 (I will occasionally work from my laptop which does not support v4.5)
• SDL 2.0

maxwellmlin

mv -v ~/Library/Application\ Support/.lpxuserdata ~/.Trash

jdupuy

// Helper function: sample the visible hemisphere from a spherical cap
vec3 SampleVndf_Hemisphere(vec2 u, vec3 wi)
{
	// sample a spherical cap in (-wi.z, 1]
	float phi = 2.0f * M_PI * u.x;
	float z = fma((1.0f - u.y), (1.0f + wi.z), -wi.z);
	float sinTheta = sqrt(clamp(1.0f - z * z, 0.0f, 1.0f));
	float x = sinTheta * cos(phi);
	float y = sinTheta * sin(phi);
	vec3 c = vec3(x, y, z);