'Inspired 3D:' Speaking with Sean Mullen

This is the second in a number of adaptations from the new Inspired series published by Premier Press. Comprised of four titles and edited by Kyle Clark and Michael Ford, these books are designed to provide animators and curious moviegoers with tips and tricks from Hollywood veterans.

The following excerpt is an interview with Sean Mullen, conducted by Kyle Clark. Sean, a one-time student of the CalArts character animation program, began his career with two short, embarrassing films, Oh Crappy Day and Horndog, for the Spike & Mike Sick & Twisted Festival of Animation. After working as a clean-up artist on various commercials and feature films such as The Lion King, he progressed to animating under his mentor, Dale Baer, on commercials featuring Tony the Tiger and Toucan Sam, among others, as well as films such as Space Jam and Quest for Camelot. Sean joined Sony Pictures Imageworks in 1998 to work on the original Stuart Little, for which he earned an Annie Award nomination for Outstanding Individual Achievement for Character Animation. Since that time he has been developing original feature film stories and artwork for Imageworks, as well as working as a supervising animator on Stuart Little 2.

Kyle Clark: I know you started your career as a traditional animator, when you made the jump to the computer, what did you find most difficult?

Sean Mullen: The most difficult thing was getting the technical and creative elements to work together. It got pretty bad in the first few weeks when I was learning the animation controls. I would start trying to do a simple piece of animation to get used to the controls and it would look horrible. I realized that I was focusing so much on the technical stuff that I was completely forgetting all the basic animation principles. I went back to a process I used to use when I first started doing traditional animation to get back on track. I made photocopies of all the lists of animation principles from The Illusion of Life and pasted them right by my computer screen. Id have to constantly be looking at them to remind myself of the basics. It seems pretty elementary, but it really helped me when a scene wasnt working.

KC: Was there a particular technical hurdle that you had to overcome?

SM: The biggest thing was my fear of the computer. Id never really worked on computers. I didnt own one yet and I hadnt even played around on one, really. It was pretty intimidating. Once I got past the fear, and realized all I had to concentrate on were the tools that Id always be using, it wasnt too bad.

KC: I know you have a CG working style that relies heavily upon the things youve learned in the traditional world. How about taking us through a shot youve completed and explaining your process.

SM: There is a shot in Stuart Little with the Stouts, Stuarts fake parents. It had some fun acting and involved multiple characters. I approached it exactly like I do a traditional scene. When working in 2D, Id do one drawing from the scene to establish the overall feeling of the characters emotional state and use that as the starting point. I do the same thing when posing in the computer. I do everything from the facial expression to the fingers. I get the whole pose worked out so I can look at it and see that Im on the right track. If its a subtle acting situation, youre pretty much going to be working in and out of that pose. If you get that one nailed down, the rest of the scene is going to be a lot easier. Action scenes will require more poses since the character will be moving more broadly, but even they can still be kept to a minimum. A good example of what Im talking about is any Norman Rockwell painting. One frame can illustrate an entire situation, from a characters movement all the way to their emotional state.

KC: That one pose is the main benchmark for the entire scene even though the scene may require many poses to communicate the emotions. That pose essentially summarizes that scene.

SM: Right. Even if there is a change of emotion, which most times there will be, usually there is an emphasis on one emotion or another. In the Stout scene, for example, when Mr. Stout goes from being preoccupied with a peanut to faking shame about having given Stuart up for adoption, the point of the scene is him faking shame. So Id focus on that second emotion.

KC: So theres one idea in a scene that you are trying to convey?

SM: Yeah. Thats what I was referring to in our discussion about commercials. The short deadlines really made me focus on that one central point. I still try to do that. Also, to a certain degree, and this probably just comes from experience and really knowing the characters Im working with, I dont put as much thought into a scene as you might expect. Once Ive got that central idea, I tend to go more with my instincts.

KC: So its similar to working straight ahead. Youre looking for that spark?

SM: Yeah. I will think about the scene when its first given to me to figure out the overall performance and how I want to achieve it. But thats maybe in the first day or so. After that I really just focus on sticking to that one idea. I try not to over-think things while Im working on the scene. So with Mr. Stout in this example, I took into account the fact that he was a life-long, second-rate hustler who was trying to convince Stuart and the Littles that Stu was actually his son. And the real trick was getting that across somewhat subtly so that I wouldnt give away the fact that it was a scam too early in the film. Mrs. Stout was also involved in the scam, but she approached it differently. First of all, shes pretty ditzy, but she has a truly loving personality. She liked the idea of being a mom so much that she had essentially convinced herself that this was actually real. So I played her much straighter. Shes not trying to act as much as Mr. Stout. So keeping these things in mind, I listen to the dialogue over and over while I picture what their performances might look like in my head. Once I decide on something Im happy with, I create that first pose. Then its just a matter of following through with the rest of the idea and not getting off track or second-guessing myself.

Mr. Stout admires his snack. All Stuart Little images © 2002 Columbia Pictures Industries. All rights reserved. Courtesy of Columbia Pictures.

KC: Once youve created a pose for each of those characters, what is the next step?

SM: The next step is setting what you would consider to be normal key poses. Just an overall road map of where I want the scene to go. Both spatially, where they are going to start and finish within the confines of the frame, as well as emotionally. This is where Id start establishing that change of emotion if there is one. So focusing on Mr. Stout in the scene, we first see him as hes about to take a bite of the peanut. Stuart asks Mr. Stout why they didnt want him, and Mr. Stout pulls his attention away from the peanut to answer him. Then he goes into delivering his sob story. Some of the key extremes would be the pose where hes about to bite the peanut, the poses where he looks at Stuart and his mind is scrambling to come up with the sob story and a few poses of extremes, such as pointing at Stuart with the peanut to accent a word and slumping over, covering his face in shame at the end of the scene.

At this point Im still not really worrying about movement. Of course, Im keeping it in mind, but the actual movement of the character is focused on later, at the point where Im doing breakdowns.

Mr. Stout thinks of an excuse.

KC: When you are setting these poses, are you keying facial features, fingers, everything?

SM: At this point, I usually included all of the details in the pose. If it was a more action-oriented scene, I probably wouldnt. However, in a subtle acting situation I definitely do.

KC: Would you show this stage of the shot to your supervisor or the director?

SM: I actually start doing some breakdowns before showing it. I want to get some idea of the timing in addition to just the emotion. Ill bring it to a point where its clear what Im going for. Thats actually a pretty important thing. A lot of people show scenes at a point where the lead or animation director has to ask, Where are you with this scene? or What is it youre trying to get across here? and the animator has to explain what is and isnt done yet. I dont think thats the best time to show a scene.

KC: I might disagree there a bit. If you are dealing with a director or supervisor who has trouble projecting what they might want, you can often get some information by showing them anything. Ive had several directors who wouldnt tell you what they wanted until you put a roughed-out scene in front of them. You didnt want to invest a lot of time producing an idea that was incompatible with the directors vision.

Mr. Stout looks away in shame.

SM: A director like thats going to be difficult to work with no matter how you work. I would personally always tend to lean toward not showing something you have to explain. Ive found that it leaves the scene too open to interpretation and then you risk the chance of being directed along a totally different path before your ideas have been given a real chance to prove themselves. And really, setting your keys and adding some breakdowns shouldnt take that long to do. For example, the scene Im working on now in Stu2 is 95 frames long. Between keys and breakdowns, I probably had no more than about 9 or 10 poses and it only took a few hours to rough out. Thats when I showed it for the first time. I agree with you on the idea of not investing too much time in a scene before showing it to whomever you need to show it to.

KC: Getting back to the shot, when you do get ready to show a scene, dont you initially present it with step-curves turned on? And if so, why?

SM: I prefer it that way and its basically due to my traditional background. It looks exactly like the traditional pencil tests we used to do. I also prefer to see only what work Ive done with no drifty computer in-betweens getting in the way. A lot of people will set keys and allow the computer to do its spline work between those keys. With splines, 90% of what you are looking at is the computers work, and the majority of the time that work is going to be wrong. The stepped method lets the scene speak much more for itself.

KC: Youre taking the computer out of your work and just presenting the decisions youve made.

SM: The only things the director, supervisor or myself see are things that I intentionally put there. This is just what Ive been used to for my entire animation career. Although I have strong convictions about working this way, I can legitimately see people working in other methods and being successful.

The pointing of a peanut.

KC: Everyone has a style thats influenced by his or her prior experiences. However, I think theres plenty of room to bring in elements from other styles. Its a constant refinement process. I have a different approach than this but I might be able to take small parts of that process and work them into my own routine.

SM: Thats definitely one of the advantages of the computer.

KC: So at this stage of the shot are you just going to continue adding breakdown keyframes between your initial pass of poses?

SM: Ive only in the last few months settled into a comfortable working method for this stage. For me, its always been the most painful portion of the process. I used to take my scenes down to fours, sixes or eights, meaning that Id save a pose on every fourth, sixth or eighth frame, depending on the action. From there Id switch to spline curves and start fixing all the inaccurate in-betweens the computer was inserting. This was so tedious and painful for me that I would have trouble concentrating on the scene. So now I like to bring the scene all the way down to twos or threes while still using stepped curves. This way, when the curves are switched to splines, theres much less room for error in the computers in-betweens. Theres not much tweaking of things necessary at all when I do it this way. The only time I dont bring it down this far is on something like a moving hold that doesnt need to be broken down very much.

The slumping state as Mr. Stout finishes his speech.

KC: By tweaking things are you setting additional keys or are you adjusting the curves?

SM: More often it would involve working with the poses themselves, interacting directly with the character. I tend to avoid any of the technical editors and curves until I need to use them for something specific. Its in the same line of thinking as only wanting to see what Ive intentionally done when showing a scene. Id rather work with the pose itself than mess around with some curve and then have to go back to see how it affected my pose. The way I look at it, no audience is ever going to see my curves on screen, so I dont really care what they look like. All anyone is ever going to see is the character, so thats where I direct my attention. I will use the graph editor, but its just usually to help me figure out whats causing something weird to happen, such as a pop in the arm.

KC: Youre using those tools as a clean-up device.

SM: Yes. Its really just that. Ill take the scene down to twos or threes by interacting directly with the model rather than refining movements by working with the curves.

KC: Youre setting keys every two or three frames for every part of the body?

SM: Not necessarily every part, but every part that needs something done to it. Say the characters body is standing relatively still but they are doing a broad sweeping motion with their arm. The body might stay on fours but the arm will end up going to twos to keep it on arc. Its a matter of only setting as many keys as you need.

KC: Its essentially putting keys where you need more control.

SM: Right. And Ive found that when I work things down to twos or threes and switch to the spline curve, 95% of the time its essentially done.

KC: How do you deal with changes once youve gotten to this stage? For example, Im the director and look at your scene in a 98% completed state. I decided to have the character jump up and down in the middle of the shot as opposed to just standing there. What happens now?

SM: If its a significant change, I basically go in and start blowing away sections of keys. Its easier just to start over on a section of the scene than it is to try and rework what you already have there.

KC: Its difficult to salvage something with that many keys. Its often just too laborious.

SM: Right. But thats why its important to get the blocking approved. Ideally, you would show your scene in the early blocking stage, and then a second time at a point about halfway through, when the idea is completely clear. You would ensure that the direction you are moving in is correct, and then finish the scene. Hopefully, you get any director comments in those first two stages before breaking it down to that many keys.

KC: So what advice can you offer to people looking to become stronger at posing their characters?

SM: I would put a really strong emphasis on getting classical training. Learn life-drawing, learn what a strong silhouette is, learn design, and all the basics that are constantly being harped on. (See Figure 10.10.) To me, the most important aspect of a good pose is that it should look natural. Dont do anything with the character that looks like it would really hurt or maybe even break a bone if you tried to do it in real life. The Illusion of Life is another great reference for posing tips, too. Also, keep in mind that 3D animation presents some new challenges, at least for 2D animators. One of the biggest for me to get used to was the idea that your character is actually three-dimensional, so even though it will only be seen from a specific angle in a scene, you should do your best to make it work from all angles. For example, if the character is leaning over toward the camera, the pose will tend to look more natural if the character is balanced when you look at it from the side as well. If it looks off-balance from the side, chances are it will look a little odd in the camera view. And remember, just because the computer appears that it can do a lot of stuff for you doesnt mean that its doing the right stuff. You really have to control it.

KC: If a person is not good at performance, do you think he can learn it?

SM: To a certain degree I think you can. In the larger sense, I believe its a gift. For some people it definitely comes much more natural. Some people just have a knack for that. But there are a few specific things you can keep in mind to create an interesting performance. I feel its really important to know the characters backstory. Not just knowing that a character acts a certain way in a particular situation, but knowing why they would act that way in a situation. Its like knowing your wife and why she reacts to things the way she does. If you know the character that strongly, its like animating a friend. I also think its important to put yourself in the characters situation. You can definitely tell the people who are doing a scene thinking, What would a cartoon character do here? versus the people who are thinking about what they might actually feel if they were in that situation. The people who apply the characters emotional state to themselves deliver much more natural performances.

KC: The result is a less clichéd approach to the shot.

SM: Absolutely. Thats a huge thing. Once you have an idea how you would act if you were in the characters shoes, you can go on to exaggerate it however you want. This ties in to my belief about getting classical training to help learn animation. Obviously, movement in animation is a caricature of movement in real life, and you need to know the fundamentals of real life before you can caricature it. Its the same with the emotional performance. Figure out an emotionally realistic approach to the scene and what would happen if you were in that situation, and then you can caricature it as much as you need to. Also, keep in mind where the character is going in the context of the overall film. The character should be growing over time, so theyll most likely act differently at different points in the story.

KC: Any last words for our readers?

SM: To sum it all up, the most important single word I can stress is natural. I personally believe that the most important elements of a successful animated scene are a natural emotional performance and natural movement from an entertaining character. These few, simple elements rely heavily on each other to truly work. Nice movement alone is only a visual novelty and gets boring pretty quickly without a feeling, interesting character behind it. A strong emotional statement can get lost behind unnatural or elaborate but unnecessary movements. And even the most interesting character can be uninteresting if its not clearly emoting or moving well. But if you can get all these elements working together naturally, the result will be a performance that audiences will believe and enjoy.

To learn more about character animation and other topics of interest to animators, check out Inspired 3D Character Animation by Kyle Clark; series edited by Kyle Clark and Michael Ford: Premier Press, 2002. 268 pages with illustrations. ISBN 1-931841-48-9 (US $59.99) Read more about all four titles in the Inspired series and check back to VFXWorld frequently to read new excerpts.

Author and series editor Kyle Clark (left). Series editor Mike Ford.

Series editor Kyle Clark is a lead animator at Microsoft's Digital Anvil Studios and co-founder of Animation Foundation. He majored in Film, Video and Computer Animation at USC and has since worked on a number of feature, commercial and game projects. He has also taught at various schools including San Francisco Academy of Art College, San Francisco State University, UCLA School of Design and Texas A&M University.

Series editor and author Michael Ford is a senior technical animator at Sony Pictures Imageworks and co-founder of Animation Foundation. A graduate of UCLAs School of Design, he has since worked on numerous feature and commercial projects at ILM, Centropolis FX and Digital Magic. He has lectured at the UCLA School of Design, USC, DeAnza College and San Francisco Academy of Art College.