The CGI 99 Conference at the University of Calgary
What should cyberspace look like when we can actually build realistic, immersive environments? Here's a short invited paper I delivered at CGI 99 on the topic.
More info
Much of what appears in this site is explained in this book. The second edition includes my new chapters on QAEB tracing, atmospheric models, and genetic textures. There's also new stuff in there from the other authors, too.
It's definitely worth getting the second edition, if you liked the first. Plus there's lots of nice new color pictures... ;-)
Surf's up!
Manuel Gamito visited us from Portugal for a year in 1998/99. He's been working on ray-domain distortion variations on QAEB tracing to render connected non-height field surfaces. I told him "if you make an image of a breaking wave by April 20th (1998), I'll put it in the 2nd edition of our book.
And so he did.
The next thing to do was to animate it. And so he did.
Unfortunately, Manuel doesn't have a web page of his own currently. So "watch this space" for a pointer to more of his work. Meanwhile, you can email him at mgamito@iscte.pt.
Some genetic images
I've done some work on genetic artwork...
A brief project description
We're working on cloud models with improved realism in both geometry and illumination.
A technical sketch submitted to SIGGRAPH 98.
A technical sketch submitted to SIGGRAPH 97.
Sonya Shannon, of the School of Visual Arts in New York City, and I submitted a technical sketch on the methods painters use to convey depth in their renderings. What's interesting is that painters have used these four perspectives for hundreds of years, yet we generally use at most two of them in computer graphics. This sketch sets out to set that aright.
(Unfortunately, it was rejected.)
"Emil"
Karen Phillips of Springer-Verlag has asked me to provide an image for the cover of a book tentatively titled "Beyond Computation," that's being commissioned by the ACM for the 50th Anniversary of the invention of the electronic computer. Such an honor blows my tiny mind! I figure that such an image has to represent the absolute state of the art in all of my capabilities.
The image above is my first cut at this challenge. It represents about two day's work as shown. It includes some of the items mentioned below: QAEB-traced mountains, GIT texturecoloring, and the Milky Way texture background. I'm pleased with it!
A moontrail on the water.
To quote Ian Hunter, "I've been wanting to do this for years."
It's far more easily said than done, as it requires geometry
(as opposed to a bump map)
all the way out to the horizon.
Such reflection is an important element in the next image I'm doing
for this project.
It was requested that the image for the book cover include a horizon, so my next cut involves both of the novel elements in the above two images. This new image features two QAEB-traced primitives, with different near and far clipping planes, in the same image. Accomplishing that required programming some new capabilities into my ray tracer. This image then represents a clear case of art driving technology in the development of my renderer. That, in fact, is pretty much how all of my modeling and rendering capabilities were developed: in service of some artistic visual goal. This image simply illustrates that process very clearly.
The full title of this image is "Parabolic Curves in the Plane of the Ecliptic." The plane of the ecliptic is the plane in which the orbits of planets in a solar system lie. Just as the planets' orbits all lie in prety much the same plane, for a given solar system its ecliptic plane is likely to lie nearly in the same plane its host spiral galaxy. Thus the alignment of the planets and the milky way is not simply a visual trick; it is an idealized conception of a possible solar system.
In several ways this is the most sophisticated of my images to date: It features rendering with adaptive level of detail via QAEB tracing, geometrically correct reflections off the water also via QAEB tracing, the complex procedural Earth and Moon models "Gaea" and "Selene," a spherical planetary atmosphere in the foreground and another around Gaea, a GIT texture on the mountains, and multifractal models in the mountains, Gaea and the milky way texture. The visual composition is also complex: there are parabolic V-curves in the mountains and Coriolis-twisted clouds of Gaea, and a similar exponential curve in distance from the near foreground up to the horizon, then to Gaea, to Selene, and finally to the galactic background -- the most complicated deliberate composition I've executed to date.
Click for a larger image.
I've been working on subtleties in the appearance of various planetary images. One thing I've always wanted, and that the image "Pleiades" called desperately for, was a background texture resembling the Milky Way. This is my recent attempt at creating such a beast. It has three parts: A white multifractal background glow, black multifractal dust lanes superimposed, and a weighted distribution of stars on top of that. I think it looks reasonably good; it will benefit from further refinement.
Project desciption
This is a project intended to bring some of the visual complexity and richeness of paintings, to computer graphics. It's the topic of a technical sketch that Myeong Lim presented at SIGGRAPH '95.
A brief project description
I've come up with a new algorithm for ray tracing procedural height fields. I call it quasi-analytic error-bounded (QAEB, pronounced "kweeb") ray tracing. If you're interested, see the technical sketch I presented at SIGGRAPH '95, or the paper submission. to SIGGRAPH 96.
This web page (or at least the art gallery sub-page) is featured in the new book "Internet Graphics Gallery" by Paul DeGroot and Dick Oliver. See pages 395-397. They have a CD ROM in the book that includes a few of my images, too.
Thanks, guys!
A bunch of snapshots taken during the week we were snowed in at home in the big blizzard (which wasn't all that big -- two or three feet of snow -- but enough to paralyze Virginia!)
In the January issue of Scientific American, pages 36 and 37, there's a short article about my planet building. They say they have some of the highlights of my home page on their home page, but alas, I've never seen it, as it's only available on America Online. If you want to know more, try SCAinquiry@aol.com.
"Ecliptic"