RAY TRACING

Ray tracing is the technique used to generate highly realistic images of 3-D geometry.

Some tech specs on the ray-tracing code I have written

User interface provides a real-time display of solid shaded objects with a software Z-Buffer polygon drawing algorithm that I developed to draw directly to dibs.
Illumination and shading using diffuse reflection, specular reflection, coloured lights, atmospheric attentuation, shadows, transparency, chrome reflection, phong normal vector interpolation.
Planar/Cubic/Spherical texture mapping including point sampling through to full antialiased mip-map texture mapping
Bump mapping.
Noise/Turbulence solid textures, marble etc..
Importing of 3-D geometry stored in Lightwave Object File Format, also imports texture definitions.
Exporting of 3-D geometry to Lightwave format or to my own proprietry format.
Adaptive super sampling and sub-division of pixels to shoot more rays through pixels where needed to provide full scene antialiasing
Adaptive object hierarchy creation before rendering to speed up and avoid polygon/ray intersection calculations with view rays, shadow rays and reflected rays.

Screen shots and history of the ray-tracing applications I have written

Started writing the algorithms for ray tracing in 1994 on the Amiga. Switched to the PC around about 1995. The dos version of my raytracer which I wrote in 1996 (using Borland Turbo C++) had an entirely text based interface. There was a main menu and objects were created by typing in x,y,z co-ordinates ! You could load and save objects to disk once you created them. A scene containing 500 polygons would be considered large due to the memory allocated would reach the 640K limit. This version contained many features which I still haven't implemented in the windows version yet. It included high quality mip-map texture mapping, bump mapping, csg (constructive solid geometry), solid textures, adaptive pixel supersampling (full scene antialiasing), and many other features.

Here is the first windows version which I developed over most of 1997. One of the main new features added to this version was that you could load objects that were stored in lightwave file format. You could define a curved camera path over a series of points and generate an avi animation. It also contained a simple text based animation language. I wrote a separate utility that would create 3-d letters from true type fonts (eg the gallery title). Some of the images and animations on this page were created with this version.

Here is a screen shot from early 1999. I had just implemented fast display of solid shaded objects using a z-buffer polygon drawing algorithm. The software z-buffer polygon drawing algorithm is very fast, I spent a fair bit of time optimizing it with the profiler.

Heres a few snaps from 1999. It has many advanced GUI features for manipulating objects.

These are some of the images generated from the programs I developed.



	I created the mesh for the lettering on the left with a utility I wrote that allowed creating solid objects from true type fonts.
		The two cars here were added onto an existing image that I captured from my video camera. I added support for alpha channel compositing and post processing so I could render the cars onto the image slightly out of focus. I also added support for alpha channel shadows.

Back to http://www.megaderek.com/