Interview with Stewart McSherry of Xfrog and More

I have been using a program called Xfrog for more reasons than it has a great name. Many years ago I was told about it by several people, and one of them was Saty Raghavachary, a programmer and educator at DreamWorks Animation. I only regret it has taken me so many years to start using the program. On the company website this program is described as a “procedural organic 3D modeler.” It is much more. All of the creations you make in the program can also be animated. Recently people have started 3D printing using Xfrog models. If you go to the web site, xfrog.com there is an amazing collection of images and animations that have been created using the program.

Xfrog has been used in everything from feature films (recently, "Avatar") to multimedia projects. A stand alone program, it also has a plug in for a number of 3D packages. It can also be used in conjuncture with Terragen, a landscaping software that the company owns. I have been using it to create animations involving birds' nests and strange creatures for an immersive dome project I am working on called “Migrations.” I was pleased to find that not only could things be generated right out of the box, but the tools were also there to create very individual, original and complex animated forms. The interface is such that expressions can be used and they also  have some excellent tutorials online.

Stewart McSherry is a CG artist and the project manager/business director of Xfrog.  In his words, he "studied computer science, fine art, printmaking, and experimental film in the bayous of southwestern Louisiana ... gradually fused it all together and made abstract large-scale computer artworks and a few films here and there."

I am a board member of  Iota, a non-profit organization dedicated to experimental and abstract film work, both past and present. It's a great organization, I invite you to peruse the website. Stewart McSherry recently became a board member and when I heard he was involved with Xfrog, this program I loved using, I asked him if he would speak about it. Below is the interview I conducted with Stewart McSherry. I talked with Stewart at his home in Malibu. A friend who had never seen Malibu came with me, and kept asking me where Malibu was. I had to explain there is no real “there there” in Malibu, just a strip of homes, a few restaurants and bars sandwiched between the Pacific Ocean and the Pacific Coast Highway. As we sat on his balcony  and were almost blown away by the brisk ocean wind Stewart told me that the city council had never allowed a Malibu unified sewage system to be created so that they could keep Malibu from being overdeveloped. It worked, wonderfully well. (Note, in all fairness Malibu does have some hills you can live on, and a beautiful state park!)

Interview with Stewart McSherry

1. Tell me a little about your background as it pertains to Xfrog?

Well I was a computer artist, fairly successful for a number of years, and I stumbled across a "shareware" called Xfrog for Silicon Graphics machines, and I began to play with it and realized it was extremely powerful and unique tool for 3D organic procedural modeling. It was a huge leap for me - I was using tools that basically were able to draw polygonal shapes and extrude them in 3D,  I went from that paradigm to this tool that was consisting of Nature - programs that could branch, spiral, create Natural forms - so i suddenly realized this program Xfrog was really special in the history of 3D modeling.

2. What inspired you to start Xfrog? What aspect of the company is the most fun for you?

So I saw the software, played with it, then met the author, and on the day I met him, after he showed me his new version, which was again a huge leap forward from the one that I was already bowled over by, on that day, I dropped my artist hat and put on a business hat and have never looked back since. I went from artist to businessman instantly and we ported the software to PC and began selling it right after that, and soon every major visual effects studio was buying it, and lots of hobbyists and small production companies bought it after they saw what studios were doing with it. The aspect of the company that is the most interesting for me is the challenge of competing with other companies who are doing plants, grasses, flowers, trees, etc, as we hired very early on professional botanists and taught them how to use Xfrog and built really complex trees and other plants, and so I enjoy basically the fun of competition and strategy in keeping us the hottest company in the organic modeling (plants) arena. We have been doing okay for the past 9 years and continue to grow every year.

3. What are the challenges,  programming and otherwise you or your partner have faced with Xfrog?

It is a good challenge to keep current with our software and our plant models. Everything must constantly improve. I never really considered that we would have to improve the plant models. I thought, we could build, for example an adult English Oak with very high complexity, and that would be it. but the reality is that the technology advances very rapidly and what no one could actually use inside their software two years ago, is easy to work with today. Now we are again pushing the limits with trees some of which nowadays have millions and millions of polygons. Which is still nowhere near actual reality, mind you, but I do like seeing the hardware continue to improve because it gives us the opportunity to crank up the precision in our procedure and yesterday's tree of 100,000 polys becomes today's effort of 10 million. We really look forward to our first billion polygon tree someday. or a forest in which we don't have to be concerned with hardware or software limits. "Avatar" in particular, was hard to find a balance between complexity and the limitations of the render farm. Because the trees were needing to be IMAX quality yet not completely crashing the network of computers rendering the film. Avatar is still our most important film work though we look forward to the release of new work in "The Hobbit" and we very recently enjoyed "Rise of the Planet of the Apes" which has the very nice shots of the Northern California Redwood forest created using XfrogPlants.

4. Any relationship to how hair and fur are programmed?

Hair and fur are created using procedural algorithms and we do the same, only our focus has been to try and mimic some of the building blocks of Nature, especially we are proud of our complex approaches at "branching" algorithms.

5. What previous work by others did you or your partner gain helpful insight from that you were able to use in Xfrog?

There is programming work by William Latham that directly influenced our horn component and work by Karl Sims influenced things too.

6. What are the biggest business challenges?

Everyday someone tries to write a better approach to plant modeling. even ourselves. we are experimenting with Lidar backscatter and nobes, we are experimenting with fluid dynamics as branching simulator, etc. but the biggest business challenge is actually coming out of eastern europe, as there are a lot of hungry technology-enriched folks who can model: plants, cars, furniture, and for very low rates. So I don't really have a good answer to that one - at some tipping point soon most computer graphics work will probably come from those regions, as they are highly skilled and very talented artists, and although we have literally tens of thousands of plant models, we see very nice plant work coming out of Russia and Poland. two traditional hotbeds of highly trained and motivated artisans.

7. What do you think the future holds for your company? In which direction or directions will you be steering it?

Although we've created plants from all over the world, we would like to at some point known for having built many of the most important plants of the world - we have a scattering from each continent but if we could go further and have really nearly all the important plants represented in 3D computer graphics, that would be a nice legacy for our company. as both a historical effort and as a way to getting closer to representing reality inside movies, computer games, online worlds.

8. Any advice for developers?

Move to India, China, or other hotbed of development! That's what I would do if I was graduating college today.

9. Any thoughts about the future of CG, different platforms, social media?

Hardware improvement has been staggering . Software has not kept up. Social media is very primitive, look for very intense improvement, live chat, hd video, everyone plugging their iPhone into Facebook and having walls of talking heads in chat rooms . But we are getting very distant through it all aren't we? I mean, who calls each other anymore? I send 200 texts a day, and maybe make 1 phone call, if that. So it is all a bit disconnecting, isnt it?  I mean I am basically sending people telegraphs...

10. What do you think about the art that is made by people using your program, what would you like to see people do with Xfrog?

I'd like to see more abstract films. Some great efforts in Japan using Xfrog to make them, but i'd really like to see more.

11. What is your wish list in terms of hardware that would help you improve your program?

Everything is great with personal computers, they keep doubling in speed every two years, e.g. Moore's law.

 

For more complete information about compiler optimizations, see our Optimization Notice.