• Advertisement
  • entries
  • comments
  • views

CulDeVu vs Photon Mapping

Sign in to follow this  


Heyy all!

Sorry for not posting for a couple weeks, but it's the beginning of the semester again, and somebody's got to represent the freshman class at parties cool.png

So the graphics professor at school here was talking to me and he was like "you know what would be a cool idea?" Turns out, it WAS a cool idea.

So! I present to you:

HBR: Human Based Rendering

(the photon mapping of the next generation)

The general idea is to either get the user that's looking at a screen to plot the photons themselves, or somehow crowd-source rendering an environment by manually tracing out light paths.

It's a novel technique, you see. Instead of you offloading the work to the GPU, the GPU is offloading the work to you. The former has been explored many times, but the other side of the symmetry has not. While HBR is not a silver bullet, we believe that, if used in conjunction with more conventional GPU usage, HBR will become a powerful tool in a graphics programmer's toolbox.

java 2015-01-31 14-06-56-40.jpg
java 2015-01-31 14-08-01-89.jpg
(all screenshots above were rendered with real humans!)

This has many advantages over normal rendering techniques. A few:

Aesthetically Accurate over Physically Accurate: the end result of a render will look like how the user thinks the environment should look. This weight on personal stylistic choice builds the illusion that they're crafting the environment themselves. This is why we believe that HBR, the act of manually placing and tracing photons, will increase immersion in games.

GPU Empathy: as previously stated, instead of you offloading the work to the GPU, the GPU is offloading work to you. When you have to manually render each frame yourself, you start being a little more okay with the thought of awful looking graphics.

Cheap for Static Geometry: this just comes with photon mapping. a pre-computation is all you need to cull offscreen photons, and BOOM! you're realtime.


Jesting aside, it would actually be really cool to have a crowdsourced rendered cornell box, where people come to place photons and render the scene themselves, and see how far off of the ground truth it is.

Welp, that's all for now. I promise, I'll be back next week with some more non-euclidean fun. I can't report on anything at the moment, though, because I have a project due in a few hours that I really should get to working on.

See yah!
Sign in to follow this  

1 Comment

Recommended Comments

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

  • Advertisement