Modeling and rendering a Lava Lamp

We are trying to use metaballs to model the material in a lava lamp, using subsurface scattering with photon mapping to simulate the diffusion within the metaballs. This will require us to create a new metaball shape in lrt as well as a new material for translucent objects. We will also need to fix the photon map integrator to allow for subsurface scattering.

A SkyLight light source for LRT

This project involves adding a new SkyLight light source to lrt. The Skylight can be used to render realistic images of the sky given a particular location on Earth, day of the year, time of day, viewing direction, and atmospheric conditions.

Adding light polarization to LRT

In this project we will implement the effect of the polarization of light reflecting off of water. If you look at clear water, usually the glitter of the sun and other objects prevent you from seeing the bottom. However, the light reflected off of water is generally polarized horizontally. So with sunglasses that filter out horizontally polarized light, you should be able to see "into" the water.

Modeling the appearance of Jade

The goal of this project is to accurately model the appearance of Green Jade artifacts such as statues. This effect will be simulated using rendering texhniques for subsurface scattering.

Rendering phosphorescent materials under UV illumination

I will define a new light similar to arealight that will produce ultraviolet rays. I will also define a new phosphorescent surface material. The blacklight will produce a soft purple colored light to simulate the fact that actual blacklights do put out some visible light. I will also add a member to the Spectrum class which will hold the UV intensity. When a UV ray hits a phosphorescent material it will produce visible light in all directions. I need to be careful here so that energy is conserved.

Rendering an image of an M.C. Escher drawing

Inspired by a project from the 2001 offering of the course in Stanford, I decided to render a painting by Escher as it would look in a photorealistic environment. The image is shown below, and consists of Escher holding a glass sphere in his hand. In order to render the hand, I plan to implement sub-surface scattering and maybe model the micro-geometry on the surface. A possible extension to this would be to use a black / dark crystal ball in place of the glass ball to see refraction as well as reflection in the ball. In either case, some caustics might appear on the hand, which will be rendered using photon-mapping.

Rendering Furry Objects

The goal of this project is to render furry objects from reference photographs using a traditional fur rendering algorithm or deep shadow maps.

Fake Fur Rendering

In Dan B. Goldman's "Fake Fur Rendering", he describes an algorithm to model fur where individual hairs are much smaller than the size of a pixel. The goal of this project is to use this technique to render dog and zebra fur.

Volumetric rendering

At the very least I intend to have volumetric rendering implemented via photon mapping, so as to illustrate volumetric caustics. Beyond that I'm looking to implement either: 1) wavelength dependent scattering, 2) varying volume density based upon some kind of flowing particle field.

Ray tracing continually varying index of refraction

I intend to model index of refraction as a function of heat in 3D space. This is what causes mirages and the shimmer in the air behind a jet's engines.

Thin film interference

My goal is to render bubbles with the effects of thin lens interference. My plan is to do this procedurally by simply using spheres and evaluating a noise function on the sphere to determine the thickness at that point. I will use an extremely low frequency noise function to get just slight variations in thickness. To make the rendering look cool I will use image based lighting.

Long-exposure images of fire

I intend to implement as much of the flame simulation as I can, or at least enough to generate realistic looking flames with some tweaking. Note that a long exposure will place less importance on the internal detail of the flame, and more emphasis on the overall color and boundary imperfections of the flame; this should make it easier to hack up a decent-looking implementation without getting it perfect. I'll implement volumetric lighting, or if I run out of time try using well-placed point sources.