Bill Desowitz interviews Softimages vp/gm Marc Stevens and special projects manager Michael Isner about the business ramifications of the companys release of Face Robot, the industrys first software dedicated to the creation of believable facial animation for high-end productions.
With last months launch of Softimage | Face Robot, the industrys first software application dedicated to the creation of believable facial animation for high-end film, post and games productions, Bill Desowitz discusses the business ramifications and challenges with Softimages vp/gm Marc Stevens and special projects manager Michael Isner.
Bill Desowitz: Talk about the business plan for Face Robot and the significance for Softimage and potentially for the future of facial animation.
Marc Stevens: Weve been in the character animation business for over 20 years, we think our talents are focused and can make the most difference in terms of added value to the industry. We were doing some soul searching back in the fall, going back over the last couple of years about the importance of character creation, where the tools arent quite where they should be in terms of delivering great performances. It all kept pointing back to the face. Michael and his team for the last couple of years did research and we came to the realization that theres a real opportunity here to make an impact on the industry. And the face is such a broad and complicated thing. Theres a reason why there arent amazing solutions out there. Solving that problem is very difficult. And so thats a place that we thought about focusing our attention, and not just from a business side. We didnt want to just throw out another feature in a general-purpose product. Thats why we decided to separate out the face with a separate product that became Face Robot: a separate streamlined UI to deal with the face and really a separate team dedicated to some extent to just worrying about problems around the face.
Michael Isner: I think its an interesting question you pose about the future. Its that idea of the future of a face product that is the birth of this whole thing. Really, if you look at the way 3D software has existed traditionally, it has evolved from the kind of computer graphics and mechanical, logistic problems involved in both making a movie and getting a 3D software package together. So its a lot of tools that a movie transforms around, how transforms move over time, and F curves and all these things are coming from the world of programming computer graphics and trying to meet the entertainment industry half-way. And so Face Robot is looking at the problem the exact opposite way. What if the entertainment industry was a mature industry in terms of computer graphics? What would be the main focus point based on the needs of entertainment? Well, if you look at something mature like the film industry, the most important thing is acting, and the most important part of acting is the face. Like Angelina Jolie: if her face is not in the scene, they use a body double. And so its really clear that these performances need a lot of specialized energy and assistance from people who can improve the process. So thats really the vision of Face Robot: a team of people and an environment dedicated to making this platform to improve 3D acting.
BD: What has been the response since the launch, including at GDC?
MS: Were overwhelmed. So theres more interest in the product than we have capability to respond to it. And so were spending all our time coming up with complex deployment plans to try and meet all the interesting surrounding the product.
BD: What about the reaction to the high price? Theres been some sticker shock with the full Designer package selling for $95,000 and the Animator selling for $15,000.
MI: It depends who youre talking to. For people who are already investing a lot of money in entertainment, theyre likely spending a lot more than that working on the face. For the student or someone who is casual about the 3D, it is a lot of money. But looking forward and what 3D really is, its not just the work of enthusiasts its also a profession. And like construction, there is a need for professional, industrial tools that address the hard needs of production.
BD: Yes, of course, but I spoke to an animator at a major studio who thinks it will be hard to justify such an investment. He estimated that every td would require the Designer and every animator would require the Animator.
MS: Well, I think if you [limit it to] those tds and animators that concentrate on the face, it comes out to one td working on the setup and several animators animating one or many faces. There are a couple things in terms of the price. This product is very much focused on quality and processing large amounts of facial animation. For the people who have those two problems, it makes a lot of sense for them. And for those who only have a job here or there to do or dont require super high quality, it doesnt make any sense. But we are looking at some alternatives like a rental program. Give a comparison: Lets say you do motion capture youre not going to go out and buy a Vicon system to do that one shoot; youre going to go to a service bureau and get it done. Thats what were looking at in terms of Face Robot too. Some type of rental model would make more sense for you if you only have one job to do.
BD: So its important for studios to know there are different ways of investing in Face Robot.
MS: Absolutely. From just our kind of experience and what we see people doing, theres a lot of time and manpower that goes into setting up these complicated facial rigs that are used and then have to be thrown away and redone for the next thing that comes along. So I think people are very optimistic about the idea of having a tool that they can quickly get set up and running and gets them 85% of the way they want to go, and they can start working on the performance right away. In the past, there would be anywhere from a couple months up to eight months of setup, depending on the quality and complexity of the facial problem they were trying to tackle. What Face Robot does is completely flip around that model where you can get your head in and get a working face in what, 15 minutes?
MI: Its doing something and you hit levels of refinement its accelerated refinement.
MS: So were not building a tool that just takes everything away from the animator. Were doing all the hard preparation and setup so they can start their animation very quickly.
MI: Weve got some really bright people who are very passionate and believe in advancing computer graphics and so its a question of figuring out the right point of focus where we can advance the core entertainment problem. I think when big studios start analyzing Face Robot, what they get excited about isnt the fact that we can get a face moving really well, given time and manpower they think they can get there, what excites them is templating the system. And the fact that 60% of the development effort behind Face Robot is the way that its able to express itself across any head, and thats an incredibly difficult problem and a valuable problem that relates to time saving.
BD: How would you gauge the interest between film and gaming?
MS: The problem is keeping up with the demand and its coming equally from film and vfx as it is from games, sometimes for a variety of different reasons. Theres a lot of evaluation and a lot of work going on right now.
BD: What kind of work?
MS: Its a new product, its a bit of a new way of working too, so we want to make sure those people have a good experience, so we come in, we get them trained, we get them up and running, we help get their heads in and teach good results once theyre comfortable with the product. We help them work with it on their own pretty quickly.
MI: I think the big moment when people are looking at Face Robot is when they get their animation on their head working very quickly. Then sort of a light goes off of what the implications are.
MS: Or what it took them to try to get something of that level in their own methods of the past. Right now that short compression of time gets people very excited. Its a huge time and money saver for people in the long run.
MI: Its a different model. If you think of the traditional model for CG, youre always having a lot of pre-production work when youre tackling face projects, so this, again, flips that model around, if that time during pre-production you really push the emotional expression of a character rather than pushing [the project] off the ground. Again, we want to push this more from a technical domain to an artistic domain.
BD: Where do you go from here?
MS: Were responding to the most critical things and there are tons of new ideas coming in. We dont have time to do them all so its pretty obvious that were going to follow-up with new ongoing releases of the product and moving forward.
MI: Its going to be a better production tool for doing faces. If you look at the price, you can do return on investment analysis and analyze it in terms of time saved during production. The other way you can analyze it is in a global level, and I think it becomes clear that were making space for a dedicated team focusing on this important problem. So if you cycle this forward, were making room for an investment thats going to grow, and the early partners that were working with, we are very in tune to their needs and very in tune to how this is going to change the entertainment industry. Its kind of like the space program in the 60s in that theres this race and were determined to get there first.
Bill Desowitz is editor of VFXWorld.
Bill Desowitz, former editor of VFXWorld, is currently the Crafts Editor of IndieWire.
