As you may have gotten the impression, I'm not the world's most prolific artist. These images take a lot of time to refine, and I have a full-time job and a life to maintain as well. So I generally only get a small number done in a given year, and even that's not very consistent.
Recently I've been on a role. Here are some of my new images.
In October I received a commission to create the cover image for Springer-Verlag's book "Beyond Calculation: The Next Fifty Years of Computing," which was in turn commissioned by the ACM to commemorate the 50th anniversary of the digital computer.
I created three images for their perusal. "Emil" was the first. "Emil" bears the name of my father-in-law, Emil Ladyko, who had a stroke in October of 1996 (he's recovering quite well). Emil is a wonderful man. He played football for the Philadelphia Eagles, so he's not one to trifle with!
The people at Springer saw "Emil" and asked for an image showing the horizon. I countered with "Moontrail," which features something I've always wanted to do: make a moontrail that goes all the way to the horizon. This is possible with a variation of the QAEB ray tracing primitive.
I'm always surprised at how much people like this simple, elegant composition.
In April, 1997, at Digital Domain we're beginning the home stretch of production for Jim Cameron's $200 million effects-extravaganza film "Titanic." One of our main tasks, of course, is to render CG water. We were arguing about the geometry of moontrail-type reflections on water; I was maintaining that they're always straight and of constant width, and Jim Rothrock said "I'd have to see it rendered in a ray tracer to be sure." So here's my rendered-in-a-ray-tracer "proof" of my correctness. (Of course, this "proof" is no better than the model it's based on, the geometric optics of ray tracing is a pretty reliable model of Nature, and the distortion inevitable in mapping a scene onto an image plane didn't cause any strange distortions, though it easily could have.)
Given that the commission was to commemorate the past and coming fifty years of computing, I thought I should create an image that represents the absolute state of the art in image synthesis. My personal research goal is to capture the complexity of Nature in synthetic images, a goal that no one has yet achieved. This image represents my best attempt to date, as of the beginning of 1997.
This image comprises most of the visual elements in my repertoire, in a scene almost as visually complex as a simple scene in Nature. I call it "Parabolic Curves in the Plane of the Ecliptic" because the view is in the plane of the ecliptic (the plane in which the planets orbit) of an imaginary solar system, and because it is suffused with parabolic curves: the terrain is a fractal sum of parabolae, and there are parabolae in the clouds on the planet, and a parabola in depth in the composition, from the near foreground to the planet, the moon behind, and the galaxy behind that. Fore these reasons, I consider it the most complex and sophisticated of my works to date.
Imagine, then, my horror when I started work here at DD in Venice Beach and discovered that down on the beach promenade, there are legions of guys making pictures almost identical to this, using hubcaps, paint can lids, and cans of spray paint! So much for my squalid hubris.
On March 27th, 1997, Mandelbrot called me up and said that he was filming a segment for the PBS show "Colors of Infinity" on the April 4th. He said that if I got some new images to him, he'd try to get them on the show which aired on April 9th. So I whipped up the following three images over the weekend of April 5th & 6th: "Benoit," "Beneath the Rule a Landscape Lies," and "Aerial Perspective."
"Benoit" is simply a horizontal reframing of "Emil."
This image is simply a realistic rendering of the multifractal terrain that I first used in an artwork, in "Emil." The idea here was simply realism and beauty.
I lecture a lot on the importance of aerial perspective--the loss of contrast and change in color with distance--in making realistic landscape renderings. (Landscape painters have known about this for hundreds of years; we're still catching up in computer graphics.) I wanted a definitive illustration of the effect for a technical sketch on the four perspectives of Renaissance painters that I coauthored with Sonya Shannon of the School of Visual Arts in New York City. (Unfortunately, it was rejected.)
Here we have the atmosphere present on the right, and absent on the left. This illustrates it's constribution to overall realism and sense of depth.
Again, for purposes of that sketch, I made this illustration of how color changes with distance in aerial perspective, or as Sonya points out, more properly, in the color perspective part of aerial perspective. This image consists of two vertical, parallel planes, receding off to infinity, with an atmosphere between them. Note that the black plane becomes blue, and the white one red, with distance.
In Nature, color perspective is due to Rayleigh scattering. For production image synthesis, we need a simpler model than a full-blown physical model of Rayleigh scattering. In practice, it's very simple to do. (See the sketch.)
The atmospheric parameters in "Aerial Perspective Test" are identical to those in "Aerial Perspective" and "Beneath the Rule a Landscape Lies."
I've always liked my little visions of Hell that I rendered for our Guggenheim performance in 1990. I always felt that the images were unfinished, though--the skies were too boring. So I revisited my private Hell and rendered it in peach tones (don't ask why; it was just one of those random inspirations) and put a gratuitous planet in the sky. This image is technically simple by my current standards, but I like it nevertheless, as a color study.