Why do we assume the rate of energy consumption is constant for graphics? For animation settings wouldn't there often need to be changes in lighting that would require changing rates of energy consumption?
StaffJakeHoles
I don't it is the norm to model the process that is creating the light. Maybe for some sort of super accurate fire sim or flickering incandescent light you might account for how much energy is going into the light as apposed to heat.
My understanding is that usually CG lights are parameterized by an exposure that would define the amount of energy emitted by the light. In animation, if you needed to you could set this value to change overtime to fit the shot.
Other places where energy conservation could come up are modeling materials and volumetric effects. If your light was passing through some fog for example, you could approximate its effect by having it lose energy as it travels.
sandykzhang
How would we consider the case where light is passing through other materials that aren't quite objects? Such as modeling light under water, where the effect on light seems to be more chaotic
willyj2k
I think that cases such as simulating light under water aren't necessarily tied to varying flux; instead it seems like those effects would be more related to reflection and refraction. Perhaps a way to think about this independence is that the presence of water doesn't necessarily affect the light source — a 60w lightbulb is still 60w whether it's shining into a tank that's empty or full of water.
Here's a quick stackexchange page that links one paper on the topic: https://computergraphics.stackexchange.com/questions/1848/what-is-the-basis-of-rendering-light-caustics
peterqiu1997
I agree that the light source itself is fixed, and going from the previous lecture, would we use recursive ray tracing to deal with transparent/reflective objects (relying purely on geometric properties)? For intensive applications of something like water where it seems repetitive, it could be rendered offline and then looped.
Caozongkai
If the rate of energy consumption is constant, are they shine evenly distributed to the space? Or the space in the center would be more illuminated?
michaeltu1
How the radiant energy is distributed from the light build into space is probably dependent on the actual light bulb itself. For instance in the light bulb on the slide, light would not emitted from the part where you plug it into the wall.
Why do we assume the rate of energy consumption is constant for graphics? For animation settings wouldn't there often need to be changes in lighting that would require changing rates of energy consumption?
I don't it is the norm to model the process that is creating the light. Maybe for some sort of super accurate fire sim or flickering incandescent light you might account for how much energy is going into the light as apposed to heat.
My understanding is that usually CG lights are parameterized by an exposure that would define the amount of energy emitted by the light. In animation, if you needed to you could set this value to change overtime to fit the shot.
Other places where energy conservation could come up are modeling materials and volumetric effects. If your light was passing through some fog for example, you could approximate its effect by having it lose energy as it travels.
How would we consider the case where light is passing through other materials that aren't quite objects? Such as modeling light under water, where the effect on light seems to be more chaotic
I think that cases such as simulating light under water aren't necessarily tied to varying flux; instead it seems like those effects would be more related to reflection and refraction. Perhaps a way to think about this independence is that the presence of water doesn't necessarily affect the light source — a 60w lightbulb is still 60w whether it's shining into a tank that's empty or full of water.
Here's a quick stackexchange page that links one paper on the topic: https://computergraphics.stackexchange.com/questions/1848/what-is-the-basis-of-rendering-light-caustics
I agree that the light source itself is fixed, and going from the previous lecture, would we use recursive ray tracing to deal with transparent/reflective objects (relying purely on geometric properties)? For intensive applications of something like water where it seems repetitive, it could be rendered offline and then looped.
If the rate of energy consumption is constant, are they shine evenly distributed to the space? Or the space in the center would be more illuminated?
How the radiant energy is distributed from the light build into space is probably dependent on the actual light bulb itself. For instance in the light bulb on the slide, light would not emitted from the part where you plug it into the wall.