ABOUT THE COVER
Vol.33 No.2 May 1999
Covers Represent Industry's Broad Reach
Long time Computer Graphics columnist Mike Milne has provided spectacular art for this issue's cover. His introductory text on the project is followed by technical information from Sophie Lodge and Daren Horley. Sophie built all the models for the project and Daren has done a lifetime study on possible dinosaur skin patterns and has designed the "paint job" for all the dinosaurs in the series.
For more than two years, FrameStore (London) has been working on a major CG animation project for the British Broadcasting Company. Walking with Dinosaurs consists of six 30-minute films for the British Broadcasting Corporation's Horizon spot (the U.K.'s flagship science series). The films are "wildlife documentaries" covering the natural history of dinosaurs in the Triassic, Jurassic and Cretaceous eras. The basic idea behind the project is that the audience should forget that they are watching animation - that they actually think they're watching film of real creatures.
The realism we've achieved (though I say it myself) is excellent, and partly this is due to the fact that the project is grounded in real science. All of the live-action backgrounds have been filmed in locations that closely resemble the environments of the appropriate era, and the design and movement of the dinosaurs has been based on the guidance of a group of experts from both sides of the Atlantic.
The animators' task has been to recreate the animals of that world, and to show what it must have been like to walk amongst the great reptiles.
All our dinosaurs begin as physical sculptures. They are commissioned for and created by sculptors Jeremy Harris and Nigel Booth.
The torosaurus for the series is an excellent example of the modeling process. The torosaurus came from Jeremy Harris as a 24 inch resin sculpt. This sculpt was lazer-scanned to produce a hefty 140 megabyte polygonal mesh as well as another, more malleable, 30 megabyte version. Because the hi-res mesh, though intricately detailed, was too big to use practically, a simpler more efficient model had to be built. The aim was to create a light and manageable patch that preserved the artistry of the original sculpt.
The first stage was to align and lattice the lo-res version so that it could be sliced lengthways and mirrored to form a complete and symmetrical dinosaur. This involved an enormous amount of point pulling, examination and reexamination to make sure her spine was straight and she didn't have any holes. Once this was done, the patches were built by lofting B-spline cross sections. By slicing the mesh where I wanted my curves, I was able to use each mesh-slice as a guide when creating the B-splines. This ensured an accurate build. Although there are off-the-shelf software packages to render this process, building the patches by hand ensures an economical model with splines and points precisely placed.
With her extra horns, beaks and fully toed and clawed feet, the torosaurus is made up of 51 separate patches - slightly more than normal on our average dinosaur build.
When the patches were built we used a process involving proprietary software to add missing detail and render the model seamless. The dinosaur, in its default grey material, was then ready to be mapped. This is where Daren Horley begins work.
The process of creating the dinosaurs began with a series of talks about dinosaurs in general by a team of paleontologists including Kent Stevens (University of Oregon), David Norman (Cambridge University), Ken Carpenter (University of Colorado) and Tom Holtz (University of Maryland). This covered many aspects of scientific theory, including anatomy, brain size, eyesight, possible behaviour, environment and skin appearance - the part relevant to me.
Ken was kind enough to leave me some casts of skin impressions from stegosaurus and hadrosaurus. The thing that struck me was the small size of the scales; on a large animal they would be all but invisible at a distance. This is assuming that the scale size was even all over the body, which I think unlikely - I guessed that maybe they increased in size towards the spine becoming heavier, more scute-like for protection. But mentioning this to Kent, he said that, for instance, carnotaurus had just the opposite with larger scales underneath. Anyway I discovered that to make the scale texture show up at TV resolution, I had to make them reasonably large anyway; it became a compromise between what was scientifically accurate and what looked right.
As there is no evidence at all (as far as I know) on colouration, designing the skin colours meant looking towards modern animals for inspiration. The BBC directors wanted a look that utilised a lot of colour, but to go over-the-top risks a fake toy-like appearance. I favoured a muted colour palette. In the end we met half way, using pattern to create an interesting look and splashes of colour as display features. I drew my references from reptiles, birds and mammals. Some people may not agree with a mammalian look, but an animal evolves a skin colouration to suit its lifestyle and environment. So, if a dinosaur fills the niche that a modern mammal does, it may well have been similarly coloured. A leopard gekko is so-called because its skin pattern closely resembles a leopard - mammal and reptile evolving similar colouration.
The textures were all painted using Photoshop. We decided not to use a 3D paint package because none offer the versatility of Photoshop. I began by painting a greyscale image over the top of a side-view rendering of the model, which incorporated all of the wrinkle and scales detail. From this I created two maps; one a bump map; the other a colour map. The pattern was painted using the dodge and burn tools to lighten and darken areas of skin. I also used quick mask to paint a highly detailed and subtle mask, then used the levels function to alter the tones. The airbrush tool, set to colour mode, enabled me to lay in colour without altering the tonal values. I also found the sponge tool incredibly useful for fine tuning the saturation of the colours, essential to achieving a natural, organic look. Animals pick up dirt, stains and scuff marks which leaves the skin with a subtle uneven randomness, a constantly changing mottle of cool and warm tones, light and dark. There is no such thing as a flat, even surface in nature!
After the side view was painted, additional views from the top, bottom, front and back were painted to ensure that all areas of the model were streak-free. UV-space versions were then created, and the maps applied to the model. Test renders were produced to reveal any errors which were then cleaned up, bump levels worked out and - voila! A painted model!
For close up shots we decided that a displacement map achieved a more realistic effect than a bump map. This was derived from the bump map, but I also painted additional displacement maps to create warts, veins and more wrinkles. I also painted a map for the ribs which was faded in and out to create a breathing effect. I painted the maps at a large enough scale to ensure a high level of detail; the standard was 6K pixels horizontally, but Diplodocus was all neck and tail so I painted his map at 18K. The maps were then shrunk to half or quarter scale depending on the distance that the animal was from camera.
The BBC should be granting permission to enter some excerpts for the Electronic Theatre at SIGGRAPH 99. The series begins airing in Fall of 1999.
Two Space by Larry Cuba. 1979, 8 minutes, 16mm, B/W.
For spring semester 1999, the Ringling School of Art and Design Department of Computer Animation began offering a freshmen course on Historical and Contemporary Issues in Computer Art, with an emphasis on animation and time-based media. As part of this course, Ringling freshmen conducted research on historical research facilities, companies and people in computer art and animation. Each team of students created a small Web page based on their data. This included biographical information, portfolio of pieces and when possible, an interview. This research is the basis for a small database on historical figures, places and events in computer art. It is the department's hope that it can eventually get the copyright permissions that will allow for this historical information to be available to other students and educators. SIGGRAPH Computer Graphics has agreed to publish interviews conducted by the students, with proper permissions, when there is space.
The first interviews are with unquestionable pioneers in the field: Larry Cuba, Alvy Ray Smith and Lillian Schwartz. A special thanks goes out to these influential pioneers for their help with the students.
Description: Two dimensional patterns, like the tile patterns of Islamic temples, are generated by performing a set of symmetry operations (translations, rotations and reflections) upon a basic figure or tile. Two Space consists of 12 such patterns produced using each of nine different animating figures (12 x 9 = 108 total). Rendered in stark black and white, the patterns produce optical illusions of figure-ground reversal and afterimages of color. Gamelan music from the classical tradition of Java adds to the mesmerizing effect.
Larry Cuba is recognized as a pioneer in computer animation and art. Producing his first computer animation, First Fig, in 1974, Cuba was at the forefront of the computer animation artists considered the "second generation" -- those who directly followed the visionaries of the sixties: John Whitney, Sr., Stan Vanderbeek and Lillian Schwartz. In his other films, Objects and Transformations, Two Space and Calculated Movements, Cuba's interest in pure abstraction is clearly defined. "Because the images are generated with algorithms written in computer language, there is a paradox in trying to use words to describe images for which words to do not exist." (Cuba, http://www.well.com/user/cuba/Biography.html)
This interview with Larry Cuba consists of a series of questions and answers conducted by Glenn Royer, based on research also conducted by Hyunhee Park and Leonidas Savvides. The answers are unaltered and in their original state. Used by permission of Larry Cuba.
Glenn Royer (GR): First, as an art student, I'm interested in knowing what other artists influenced you early on that helped shape the direction you went in.
Larry Cuba (LC): Jordan Belson's films opened my eyes to how film could be a medium for fine art far removed from its mainstream narrative function. When I first saw his film Allures, it was the epiphany that shifted the direction of my career from architecture to film art. He's still the master.
John Whitney's work introduced me to the concept of mathematically generated form and movement via computer graphics. My interest in form and geometry had led me to architecture in the first place, but I found that computer graphic based art was the most direct way to experiment with that relationship between mathematics and form and movement.
At Calarts, I started seeing Oskar Fischinger's work, a true expression of how musical this art form could be. Fischinger's influence on me goes deep -- as does Norman McLaren's, but in a different way.
GR: I read that you did 3D photography while in college. Do you still do any artwork outside of the computer?
LC: I've been doing stereoscopic photography since I was in college --that's makes 25 years worth of it. Last year, I presented a retrospective show of this work at the Exploratorium, a science museum in San Francisco. I've always done some photography, but I started the 3D work specifically to learn about stereoscopy which I wanted to use in my computer animation. The 3D computer animated film has been a long time coming, but in the interim the 3D photography has had a life of its own.
GR: What is it specifically about the computer that helps you to better create your 'visual music'?
LC: I don't particularly like the term visual music, but unfortunately no one has come up with a better one so far. My work consists of experimenting with algebraic structures to generate visual material that is unpredictable in its form. By that I mean that it was not designed, imagined or otherwise previsualized. The results are only known after the program is run, the computer executes the computations and produces the animation.
This is what gives the process its element of exploration, discovery and surprise. However, this kind of thing can never be 100 percent unknown, nor would you want it to be. The structure of the program does determine the visual results so I must have "some" knowledge of what's going to happen (since I wrote the program). It's just that this knowledge is of a procedural nature rather than a visual one. It's the visible results of each experiment which forms the basis of the next experiment. Turning the mathematical expression of the work into a visible one is the role of the computer.
GR: I enjoyed seeing Calculated Movements. The idea of visual music is evident there both in product and in process. What similarities do you see between mathematics, music and art? (the piece does speak for itself but I wanted to ask you this).
LC: The connection between these three is at the very core of my work. They are so deeply intertwined in my mind, that I couldn't begin to enumerate similarities.
GR: It seems to me that in 1985 when you produced Calculated Movements, there were very few artists that were producing computer animated sequences not designed to exemplify the newest technology. Do you think that type of situation is present today as well?
LC: Judging from the work I've seen over the years, there seems to be a confusion between what constitutes an interesting and worthwhile work of art and an expression of a technological feat. This is a consequence of considering computer art as a unique art form, putting the focus on the technology employed independent of any aesthetic concerns. There's also the problem that animation has never been considered to be a legitimate medium for fine art except by a very small minority.
GR: What sort of work did you do for ZKM during your time there?
LC: The ZKM offered me the facility to do my experiments on workstation level machines. I continued developing my new software in the Python computer language and produced a number of sequences on the road to a new film.
GR: What do you think of how computer graphics and animation courses are generally approached in schools?
LC: I'm not that involved with schools at the moment, but I believe that programming is the key to unlocking the power of the computer for image making. It's also key to understanding systems in a way that facilitates learning new software. The use of a particular software package is an essential skill to learn, but will be short lived compared to the general knowledge that prepares a person to adapt to new situations. If a programming course is offered to art students at all, it's usually as an elective rather than part of the core curriculum.
GR: What influence do you think abstract animation has on the art world?
GR: You've described your work as a whole to be a continuous experiment, or more precisely, an ongoing research project consisting of many experiments. What information from your past experiments led you to form your newest experiment-in-progress?
LC: This information comes in a wide assortment of forms from the mathematical to the perceptual. If I could put it into words, it would take another career to write it all down. Instead I choose to put it into images.
GR: Why stereography?
LC: Stereoscopic 3D is usually used to enhance the realism of the artificial experience of film. Like perspective projection, 3D is another cue which fools our visual system into perceiving depth. Since I'm not seeking photorealism, I don't think of it as an enhancement, but as a basic perceptual parameter. Just as perspective projection adds to the sense of depth in a scene, an orthogonal projection diminishes the depth effect and gives a more graphic result (like a mechanical engineering drawing). What visual world is created when you subtract depth by using an orthogonal projection and add depth with stereoscopy? It's questions like this that drive my research.
GR: Do you think there may be a renewed interest in stereography in the world at large?
LC: Everyone enjoys seeing 3D images. Whether it succeeds in the image marketplace as a standard form or not depends on technological, historical and social questions. I believe the success of 3D is inversely proportional to the additional cost (both in dollars and effort to use) of 3D over traditional methods. If the cost ever goes to zero, then 3D could replace 2D completely.
white.sands by Alvy Ray Smith. 1983.
Alvy Ray Smith
Dr. Alvy Ray Smith designed this piece as a tribute to his home town of White Sands, New Mexico, and to his father, who is a botanist. The plants are an example of Smith's "graftal plants," and the grasses were done using a particle system. He included his Chinese translated name, Chop, in this piece. Chop means "White Light Smith" which is both Smith's name and profession.
Smith is a graphics fellow at Microsoft. He cofounded four centers of computer graphics excellence (Altamira, Pixar, Lucasfilm, NY Tech) before joining Microsoft as its first graphics fellow. Smith received a technical Academy Award for alpha channel concept and a second technical Academy Award for digital paint systems. He invented, directed, originated or otherwise was instrumental in the following developments: first full-color paint program, HSV color model, alpha channel, Genesis demo in Star Trek II: The Wrath of Khan, first Academy Award winning computer-generated short Tin Toy, first computer-generated film Toy Story, Academy Award winning Disney animation production system CAPS and the Visible Human Project. Smith was a star witness in a trial that successfully invalidated five patents that have been plaguing the digital imaging business. He writes and speaks extensively and is on the Microsoft Art Committee.
This interview with Alvy Ray Smith was conducted by Jessica Sances and Danielle Antonini. The answers are unaltered and in their original state and are used by permission of Alvy Ray Smith.
Sances and Antonini (S&A): We found it very interesting that you received your Ph.D. in computer science and yet you are so involved in the art field. In class, we are discussing the separations between the artist and scientist and how the computer might reunite them. Given your experience in both art and science, how do you feel about this split and do you feel that they will ever be united?
Alvy Ray Smith (ARS): I have been riding the line between art and technology all my professional life. In one sense I stopped making lots of art in order to get the digital tools ready. That brief break has extended to a major chunk of my life, but I felt it was necessary to get the tools I wanted (and presumably other artists would want too). One of my constant themes to other artists is this: Please embrace the new medium of computation. It is the most malleable medium/tool we humans have ever created and it is in dire need of interpretation by those explorers of our culture whom we call artists. To say it negatively, during most of my professional life, artists have been offended or frightened by the computer rather than recognizing it for what it is - the most powerful creative tool ever. I have attributed this to the painful early days of the technology when everything was slow and crude. But we are getting beyond that now, and it is time for artists - at least those who see themselves as masters of the culture - to tune in and help us understand just what this new stuff, called computation, is.
As to your question: will art and technology ever be united? Notice that this is a separate question from whether artists will ever become comfortable with technology (for which I am convinced the answer is, Of Course!). This is rather about the question of whether technically creative people will ever be comfortable with artistically creative people and vice versa. There seem to be two quite different kinds of minds at work although both should be described as creative. I have seen two very different work environments up close that give two very different pictures. Pixar is an excellent example of how the two types can and do work harmoniously together - with equal respect, dignity, salary, promotion opportunities, company ownership and mutual admiration. This is probably due to the self-selection of the employees there to those who actively look for harmonious relationships with "the other side."
My other experience is Microsoft which, frankly, just doesn't get it about artists. The technically creative people here are awesome and Microsoft is the best-run company I've ever seen, but the people here don't respect artists (in that deep way I just described at Pixar). They seem to believe the really good talents in the world are technical and if you can't cut that then you do other things, like art. In other words, the culture here doesn't (not yet, anyway) welcome in the "other side." I'm trying to change this, but it isn't so yet. So it's like I've seen both ends of a spectrum of possible harmony between the two camps.
On the personal level, I've had two tremendously successful collaborations in my artistic life, one of which you might have heard, the other probably not. The public one is with Ed Emshwiller. He and I worked together for half a year or so in 1979 and created Sunstone, a piece I am still very proud of which is in lots of museum collections around the world. In this collaboration, one could say Ed was the artist and I was the technologist, but in fact, it wouldn't have worked if that had been the strict division of labor. Instead, Ed brought a profound love of machines and machinery and technology to the table, and I brought a deep love and need for artistic creativity. We both reached across the (low) barrier between us with full admiration, support and contributions, and the results were wonderful and the camaraderie of the highest order. My other "collaboration" is/was with my very dear friend David DiFrancesco. Here again, he was an artist with a deep love for things technical - in fact, he now builds the laser printers that Pixar uses to record movies directly to film. So I know that it is possible for people who are willing to admit the other side to have wonderful cross-boundary relationships - so fine, in fact, as to belie the existence of a "boundary" at all.
S&A: In your Web site there is mention of your dissertation: The Cellular Automata Theory. What is this theory, and how does it pertain to art and/or science?
ARS: Cellular automata (CA) is a mathematical discipline in the science of computation. I was drawn to it because (1) the 2D versions of them make spatial patterns (my Scientific American cover of 1971), and (2) there is a possible relation between them and how living things "compute" their own bodies (my Ph.D. thesis was on self reproducing machines, among other topics in CA theory). So I chose this topic because it was not "cold" - it had hints of art and life involved as well as beautiful mathematics (with an altogether different esthetic). There is a cross-over paper in my writings: Plants, Graftals, and Formal Languages, listed under CG papers, is an application of the theory of formal languages I learned in CA theory applied to growth of plants - using the lifelike aspects obviously.
S&A: When did you first recognize that computers could be integrated with art, and what inspired you to incorporate the two?
ARS: I can remember the day fully when I got the crucial insight - not the date but the day: My good friend Dick Shoup, in about 1973 or 1974, told me about a machine he was building at Xerox PARC that would help artists. He was describing what turned out to be the first serious paint program in the world. I'm embarrassed to say that I didn't get it at first, from his verbal description. However, I eventually dropped out of academia (NYU) and made my way to California on a strong whim that I would find my destiny there. My destiny turned out to come via Dick and his wonderful machine/program at PARC. When I actually saw that I could paint electronically, I was immediately hooked and understood both my future and that of computer art (in broad outline anyway). What I had intended to be a visit of a few minutes to his lab turned into about 12 hours - an obligatory visit became ecstasy. I got myself hired at PARC to artistically exploit the new medium.
S&A: In your opinion, what do you think is the next step for art and science in technology/animation?
ARS: I've already answered part of this: It is time for artists to embrace computation and start teaching the rest of us what this new medium really is. Please don't leave it to the nerds! Another answer is that we still have a long way to go in crafting, or sculpting, the UIs (user interfaces) between artists and the underlying technology. They are (in animation and 3D) still too hard, not graceful, overwhelming in complexity, etc. In fact, sculpting UIs is an art form in itself.
S&A: Finally, is there a theme behind your work that represents you?
ARS: I think your set of questions has been well-aimed directly at what "represents" me: Knocking down the barriers between art and technology, fostering a mutual admiration between the technically and artistically creative, making these early days of digital technology (yes, these are still the early days) less painful for artists, and to make precise the characteristics of the new creative medium of computation. I have tried to think of the one central theme in all this that would also describe what I think of as my art. It might surprise you. It is the magic and beauty of controlled complexity. What happens when an artistic temperament is aimed at the creative modulation of unbounded but controlled complexity - for example, living things, embryos, plants, brains, computation, language, pattern, "reality"?
** No image - waiting for legal aproval **|
Famous Moviestar - Copyright 1999, Lillian Schwartz - no rights to copy or reproduce.
Lillian Schwartz is a pioneer in computer art including film/video, graphics and the World Wide Web. Her work is in major collections around the world including The Museum of Modern Art, The Metropolitan Museum of Art and Georges Pompidou Center in Paris. A Fellow of the World Academy of Art and Science, Schwartz authored, with Laurens R. Schwartz, The Computer Artist's Handbook (W.W.Norton, 1992).
In her extensive work at Bell Labs, she has collaborated with computer scientists, psychologists in visual perception, engineers and musicians to expand the vocabulary of computer generated images. Perhaps best known for her piece, It is I, Schwartz documents the striking resemblance between the Mona Lisa and Leonardo himself. The implication is that the true identity of Leonardo's model is himself.
In her latest projects, Schwartz and a team from Rutgers University have created a model that predicts when the Tower of Pisa will fall and provides a means of testing approaches for saving the tower (September 1998).
This interview with Lillian Schwartz consists of a series of questions-and-answers conducted by Ryan Duncan via email and telephone, Monday, March 22, 1999. The answers are unaltered and in their original state. Used by permission of Lillian Schwartz.
Ryan Duncan (RD): I am interested in your motivation for creating computer art. Could you relate some of your influences and sources of inspiration?
Lillian Schwartz (LS): I have always been interested in experimenting with new technologies, to use whatever is around us to create new imagery. I have had a number of muses including Leonardo, Rembrandt, Picasso and most recently, Piero della Francesca.
RD: In the past, you have used the computer to create your own original pieces, but also to manipulate and examine the work of other famous artists. How has the computer helped you to better understand what these people were trying to do?
LS: Artists have always learned from the masters. It is a common occurrence, even today, to see artists in the Metropolitan Museum of Art in NYC, for example, copying great works of art as a learning experience. In drawing from a masterpiece, the artist learns a bit how the artist worked, the composition used, how the artists used materials. I find the computer an excellent tool for study. I can take apart a work of art and study the composition alone, the palette alone. I can then reorder a work of art using the same compositions or palettes to create another work of art. Picasso, at times, relied heavily on others work, to abstract the work, break the elements down to their simplest denominations.
RD: I understand that you started working first as a fine artist, and later added the computer to your set of tools. I am wondering if you have worked much in traditional media since, and what impact the computer has had on your more traditional work.
LS: I work almost solely with the computer, but many times end up with a traditional media such as silk screen, lithograph, film or video. I find the computer, coupled with more traditional tools, is one of the more sophisticated tools of our times. I'm allowed to think of creating in ways that were not possible before.
RD: I am interested in the different reactions from the art community when you began to create art on the computer. Was your work readily accepted, or did you face obstacles in generating public understanding of this art form?
LS: All new art forms provoke criticism. Just as photography took years to become museum worthy, the computer has also taken time. The public today more readily accepts the results of work created by computer when it is more in the commercial realm. I continue to work no matter what the public approval or disapproval is. There are more outlets today for exhibitions of computer art. This helps public understanding.
RD: You have been at the forefront of computer graphics and technology breakthroughs for several decades now. Having witnessed (and contributed to) the evolution of this art form, do you have any predictions for the computer and its influence on art in the future?
LS: A misconception is that the computer will replace other art forms. This is not true. It is its own medium with its own inherent look and feel. Again, the coupling of this tool with video, for example, makes the tool an extremely powerful one.
Karen Sullivan is faculty at Ringling School of Art and Design in the Department of Computer Animation and Foundation Studies. Her major focus of research is in concept, narrative and literacy for media and animation. Her video installations and single channel pieces have been shown nationwide. Karen received her M.F.A. from the University of Massachusetts, Amherst, and her B.F.A. from Indiana University, Bloomington.