Apparently, Bland 2D Grey Squares Are Important to My 3D Renderer

The furnace render presents an albedo-1 sphere within an environment with a radiance of 0.5. At first glance, the sphere appears invisible to the observer. The furnace test entails placing a sphere with an albedo of 1 (a surface that absorbs no light) inside an environment with radiance L in all directions. Camera rays either miss the sphere or hit it.

If the ray misses, the rendered pixel reflects the environment with a radiance of L. When the ray hits, the light bounces until it eventually misses, so the pixel also reflects L. Thus, no matter the path, the pixel consistently reads L, resulting in a flat grey sphere with no discernible edge.

The equation P: a miss reads L. Q: a hit bounces until it misses, so it reads L too. Thus, both P and Q lead to the same outcome. Simplifying, R states that the square is flat grey, implying that every pixel, reflecting L, results in a sphere lacking any perceived edge. Essentially, both camera rays in the furnace scenario read L, causing the sphere to vanish from view.

The test proceeds in reverse, indicating that if the sphere becomes visible, either P or Q is incorrect; however, P cannot be true as it would result in the sphere being lost. Therefore, Q must hold true, meaning the renderer is losing energy if the sphere is darker, while it gains energy if the sphere is brighter. When the furnace test runs correctly, no image is produced.

Julien Guertault's "The White Furnace Test" showcases this concept, starting with a ski-slope photograph that aligns with the predicted results.