Radiance change with light distance

[Gadersd]

I am trying to solve the rendering equation with monte carlo integration. I gather all of the incoming radiance from light sources and plug it into the rendering equation. A reference I am using said that radiance does not change with distance. Does this mean that a point light source will illuminate objects from a infinite distance away since the radiance does not change?

 

[Drakkith]

Not sure. Can you provide a link to your reference?

 

[Gadersd]

Here is one of my references. http://www.graphics.stanford.edu/courses/cs348b-01/course29.hanrahan.pdf Read the first page.

 

[Drakkith]

I can't be sure, but it appears to say that whatever quantity of light you have associated with a ray will stay with that ray. I wouldn't say that a point source can illuminate an object an infinite distance away, since light is not physically composed of rays.

 

[mic*]

In this modelling scenario you are modelling the path of individual rays which may as well be regarded as individual photons for the purpose of intuition, although the amount of energy associated with each ray is likely to be much greater than an actual photon.

Yes the ray will travel infinitely if running the simulation does not introduce an interaction with another object. But what probability is associated with this outcome in your scenario?

The model appears limited by the quantity of light that can be associated with a ray, and also by the physical accuracy of modelling reflections between ideal flat surfaces separated by large distances.

The first limitation is due to the increase in computational cost of tracing more rays, each having less energy associated with it (as ray energy approaches photon energy resolution is improved). The second limitation is due to measuring and encoding accurate curvature.

The first limitation is obvious and tangible. The second is barely of concern for basically all pracital applications.