The PDI/DreamWorks head of character animation helps bring one of animation’s classic duos to the big screen.
Let’s get it out in the open right away – Peabody’s Improbably History is one of my all-time favorite animated shows, a brilliantly conceived and hilariously written fixture on The Rocky & Bullwinkle Show that ran originally in the U.S. from 1959-1964 and then in various syndicated incarnations for what seemed like the next two hundred years. So several years ago, after reading the first announcement of a Mr. Peabody & Sherman feature in development at DreamWorks Animation, like many fans, I felt mixed emotions – curiosity and excitement to see a modern spin on a classic favorite tinged with a natural cynicism deep in my brain saying, “Two words - Boris and Natasha!”
But unlike many fans of classic properties who act as self-appointed protectors of the realm, disparaging any attempt to create fresh derivative work as sheer blasphemy, I devote my emotional zeal to more important things in life like family, career, disparaging local high-end gastropubs and my all-consuming quest to create the perfect rillettes. A new take on a cherished old animated property – who am I to object?
Which brings us back to DreamWorks Animation, whose army of artists have finally brought Mr. Peabody & Sherman to the big screen. I’ve known head of character animation Jason Schleifer for a decade or more going back to his Weta Lord of the Rings days. A couple times over the past few years we’ve spoken about my love for Peabody and Sherman and my hope the film delivers the goods. He assured me I wouldn’t be disappointed.
Indeed he was right - I have not been disappointed. Not in the least. Smartly written, wonderfully voiced and expertly animated, Mr. Peabody & Sherman, for me, was a hoot, sure to make the late Jay Ward proud as it introduces to a whole new audience the silly and entertaining notion of “a dog and his boy.”
I recently had a chance to catchup with Jason and talk in-depth about the film. He shared his insights on look development, scene scheduling, keeping things moving forward and most importantly, the challenges of keeping old fans happy while bringing a classic property to a brand new generation of fans.
Dan Sarto: Peabody and Sherman as you may recall are two of my all-time favorites. I’m sure it wasn’t easy resurrecting a beloved old property and making it relevant for a brand new audience who’ve never seen the original shows.
Jason Schleifer: I remember watching the original shows as well. It's one of those things when you get an opportunity to work on something you enjoyed as a kid, it’s really exciting. But then it's also pretty terrifying because you want all the fans to really enjoy it. It definitely can be kind of scary and kind of exciting all at the same time.
DS: There’s certainly a built-in fan base you have to try and please…
JS: Oh yeah! It was the same thing with Lord of the Rings. You have a built-in fan base that loves the source material and you better not mess it up. So we were thinking a lot about how the fans were going to feel. But we’re also fans ourselves, so we knew if we treated it with the same respect we’d expect to see if somebody was re-doing something we loved, then hopefully, the true die-hard fans would appreciate what we were able to bring to it.
DS: Aside from the story and animation, one of Mr. Peabody & Sherman’s strengths is the voice acting. Half the battle with Mr. Peabody, for example, is the acting. His character is so distinctive.
JS: When we heard Ty's [Burrell] performance as Mr. Peabody, we got the sense of the original Mr. Peabody sound, but there was so much warmth and depth in it. I thought, “Oh! This is going to be good.” Mr. Peabody is such a challenging character. He’s this amazing character who can do anything, who has no flaws, and you have to make him appealing and approachable to the audience. It was really hard.
DS: You have to make him human even though he is a dog.
JS: Exactly. Exactly! Trying to maintain his superiority but not make him off putting was really difficult. Ty's performance had so much warmth to it, so much heart. Plus Max [Charles] is just phenomenal as Sherman. He has such a great voice.
DS: Max did a great Sherman. Plus from a story standpoint, the film had the right amount of whimsy and wackiness that paid homage to the original cartoons. And the animation looked really great.
JS: Thank you very much. It was definitely fun for us to be able to push it.
DS: Bringing minimally designed 2D cartoon characters to 3D is not easy. Tell me about the early look development and motion studies. How did you guys first attack the fundamental challenge of bringing the characters to 3D space?
JS: Very early on we had visual structure meetings where David James, our production designer, Philippe Denis, our visual effects supervisor, Kent Seki, our director of previs, Tim Lamb, our art director and myself all got together and starting putting up images of things we found inspirational, that we felt could work with the movie, whether from live action or animated films. Images from things that were happening during the 50s and 60s, the mid-century modern design aesthetic. We started trying to understand how we could apply some of those same rules to what we were doing in the film.
If you think about what was happening in animation and advertising at that time, you see a lot of really flat graphic imagery specifically trying to get across an idea and an emotion. Animation from the UPA period, it's all about trying to get a feeling, boiling out all the extra stuff to get the essence of what the idea and raw emotion is.
We decided to follow the same idea in 3D, trying to boil down the character designs, poses and environment so you got the idea of what they were really like. For example, our mindset for everything was to get rid of anything that was confusing about a character’s pose. We went back to the basics, such as clean silhouettes, very clear lines of action, clear ideas, very simple lines in the face, trying not to be distracting with textures and the reflection of glasses. Rob [Minkoff the director] really liked that approach. At the same time, we tried to add fun and whimsy to the designs as well as the modeling and texturing, to make it fun to look at. If you look at Jay Ward’s early designs for things like smoke, they have this fun hand drawn feel to them. So a lot of our modeling was geared toward making things simple but with fun and whimsy.
DS: How long did look development take before you were ready to start animating?
JS: I can't remember the exact number of months. My first meeting with Rob and David was February 2011. That was very early exploration time, when we got together and literally looked at tons of movies. We went to Stuart Ng’s Bookstore down in LA and looked at a bunch of reference material. David brought a whole bunch of art work to illustrate what he was thinking based on the script. Rob had been thinking quite a bit longer about the direction he wanted to take, but that was the first time we all got together to starting planning the look of the film.
DS: What does the head of character animation do on a movie like this?
JS: Drink massive amounts of coffee. It’s my job to make sure we’re keeping coffee growers fat and happy. [Laughs] My role sort of evolved through the course of the film. Animation is obviously about trying to bring characters to life and allowing the audience to connect with characters on a deep emotional level. Everything I’m working on is centered on what is the best way to make that happen. So early on in the show, I’m trying to understand emotionally who these characters are on a holistic level. Who is Mr. Peabody, how does he respond to any situation? Let’s say somebody attacks him with a sword and he responds by sword fighting, how would that look? What makes him unique from Sherman? I’m trying to really understand these characters down to their core essence so that every choice we make is unique and special. It’s an ongoing process.
We start very, very early, before we even get character designs. Getting designs starts to impact some of the decisions we make. We start listening to dialog tracks which also impact some of the decision making. Then we start animating and pushing on things, on situations that we hadn’t previously anticipated. We try and figure out how Mr. Peabody would respond, how Sherman would respond, what feels most like them. We do a lot of exploration to see what works and what doesn't work. Very early on in production, we do animation tests where we just throw out as many ideas as possible. We’ll say, “You know what, Mr. Peabody feels great when he’s posed this way and he doesn't feel so good when he’s posed that way.” We start making rules about how you pose Mr. Peabody, how you pose Sherman.
An example of that is dealing with their eyebrows and glasses. They both wear giant glasses with thick rims that cover quite a bit of their faces. The top rim of the glasses falls right where you would want a normal character’s eyebrows to sit. If you put their eyebrows right where those rims are, it becomes very difficult to read an expression. When we realized that, we said if we raise the eyebrows up and make sure they’re not touching the glasses, then all of a sudden we can understand what either character is doing. So there is a rule. Whenever either character is lifting their eyebrows, they must be sitting up above the rim of the glasses. When they frown, we have to drop them all the way down below the rim of their glasses so we can read the expression clearly. On Mr. Peabody, that's a huge distance. We had to figure out how fast we could move Mr. Peabody’s eyebrow so that it didn’t pop you out of the movie when you watched, that you could track his brows were dropping as opposed to sort of disappearing and reappearing down below the glasses. There is a lot of general character exploration that happens early on.
My job as head of character animation is to shepherd that process. Do animation tests, work with the character designers to make sure the designs will actually translate into 3D, work with the character TDs as they start to rig up the characters to make sure we can achieve the shapes we want and deliver the performances we want.
I worked a lot with Rob to understand what he liked and what he responded favorably to. Once we got into production, we’d have anywhere from 35 to as many as 50 animators working on the show. My job is to make sure that they deliver the performances that Rob likes. When we show him work, we want him to say, “Yes, that's exactly what I want.” I have to really understand what he likes in terms of performance and animation style. It's kind of daunting trying to manage the animation process.
DS: I can imagine the chaos if you’re constantly chasing his approval without insight into his sensibilities.
JS: Once we start getting characters and start going into production then my job is mostly centered around making sure the animators have a clear sense of what’s needed on the scenes they’re working on. It involves a lot of talking story, talking to the director, talking to the supervising animators and really clarifying the emotional arcs of the characters and what's happening with them within the story. Then, as we work on sequences, every day I do draw overs and reviews with the animators, making sure they keep moving and we make all the right choices for each of our characters. I also help the animators bring their idea to the production. They have so many great ideas. Every single one of them is a filmmaker and they all have great ideas. My job is to help them bring their ideas to the show and communicate those best with Rob and the producers.
By the end of the show, my job mostly is making sure everybody is working as smoothly as possible, doing draw overs, giving notes, taking notes and then when I have time, animating a bit myself.
DS: So you got a chance to do some animation yourself?
JS: Yeah, definitely. That's one of the challenges being the head of character animation. You start off as an artist animating every day. That's what you got into this industry to do. As you move up you start supervising and then as head of animation, sort of bridging animating and managing. In order to be an effective head of animation you have to animate as well because you have to understand how to move the characters and make them unique.
DS: Regarding scheduling such a big film, do you work fairly linearly through the story? Do you leave more difficult sequences for later in the production or tackle them early on? How do you determine when you work on stuff and how long you can work on it before you need to move on?
JS: We work on a sequence level. Sequences are usually 2 to 3 minutes long and each department has a generic set amount of time they will work on a sequence. We have a generic schedule, sort of our best case scenario, for each sequence. Given an average complexity, an average length of time, and an average number of people on it, we expect previs to take six weeks, animation preparation to take one week, animation to take eight weeks, lighting and setup to take a couple of weeks and visual effects to take however long they take. We start the film with this general waterfall schedule. We know we have this many sequences, and this is our waterfall. We need to get the movie done in 18 months, which will mean this many teams of people based on this number of sequences in the timeline.
Based on complexity, all that changes as we bring sequences into production. In the beginning, we give ourselves a bit more time with sequences since we’re trying to get the pipeline working smoothly and have everybody get used to the show. We know it's going to take a little bit longer to do stuff. We know our first sequence is going to take maybe twice as long as what a normal sequence will take by the end of the show.
The producer, director and story department decide what the strongest piece is to start production with. What really feels like an achievable first sequence to get out the door. We started with the bedroom sequence after the principal’s office, with Peabody and Sherman talking about Sherman getting into trouble at school. For animation, it was a perfect sequence to start on because it was all character. There is one scene, one environment, Sherman is lying in bed and it's really about the connection between the two characters. So for us that was an amazing first sequence to get into because you got to really understand who these guys are.
Sometimes on a film, you’ll say, “OK, we know this is going to be a very effects-heavy show and we need to try out our city system first.” Let’s get a sequence in early that has a lot of city stuff going on, lot of crowds, so we can figure out all the potential problem issues. Sometimes we want to save that type of work for later because we know it's going to be really hard and we have a lot of pre-production work to do. So it really involves looking at where we are at in the pipeline in terms of what we can handle. We don't want to take it too easy in the beginning so that we have all the hard stuff at the very end and it's like, “Oh! My God! We can't staff this. It's impossible, because there are other movies we’re trying to do too!”
DS: How long do you work on a challenging sequence before questioning whether or not you can actually pull it off as planned? How do you arbitrate and make decisions regarding when to keep pushing along the current path or deciding to change tactics?
JS: Lot of that decision making is based on what else is happening on the show. Every week we look at the status of what we’re currently working on, how it's going, making sure we’re making progress. If things feel like they’re stagnating or not moving in the right direction, we very quickly try and figure out what's causing the problems. There is a lot of communication. The leadership team on this show pretty much all worked together on Megamind. We have a great short hand for communication. We all trust each other implicitly and know that everybody has the best interest of the film in their minds. At our meetings, if anyone is having any troubles, it gets communicated early on. Let’s say I’ve got animators who are trying to animate a shot and the camera is just really difficult for them to work with because previs setup the camera work before they had the performances. Now that we’re doing the performances we see we need to change the camera work. Instead of struggling and ending up wasting time and being inefficient, we will just go right over to the camera guys and say, “Hey, here is what we are trying to do, can you help us out with that?” They are always more than willing to make any necessary changes we need.
Our normal pipeline is quite efficient. We’re really good at figuring stuff out. But there are certainly sequences that are extremely complicated and fall outside our normal pipeline. Like for example the sequence where Mr. Peabody is going through his entire history with Sherman and we actually see Sherman getting younger and younger and younger. The look and design was completely different from what we normally did in our sequences. We knew that going in. It was going to be unique. We immediately sat together and said, “This is going to take longer than a normal sequence will. We don't know how we’re going to do it yet, so we’re going to spend a lot more time in pre-production figuring out how we can get this to work.” We had weekly meetings on it, we tried things and we made sure the art department had a very clear idea of what they wanted so we knew where we were headed. They did a lot of upfront design. Then we just talked about it every day, every week, what’s the status, what’s needed to keep it moving forward.
DS: Even with all the planning and all your collective experience, were there some areas that proved more challenging than you anticipated? Were some things easier than you expected?
JS: That's a really good question because having done this for a number of years now, I know going into every film there is going to be stuff that's easier than I expected and stuff that's harder than I expected. Probably the biggest challenge for us was Ms. Grunion, really understanding her character, her motivation. Just from the animation standpoint, she was really challenging for us, trying to get a real sense of her. That’s one of the things you want with your villain. You must make sure they make sense, that they have a clear sense of motivation that the audience can understand. She wasn’t very present in the film to keep things going like a traditional villain who is there all the time. We had to figure out how to make her have an impact that sticks around long enough to be interesting. I didn't expect her to be as much of a challenge as she was.
In terms of the production itself there weren’t that many challenges that blew me away, where I was like, “Whoa, where did that come from? I didn’t expect that at all!” Having that exploration in the very beginning was so key to identifying potential challenges before they ever occurred.
A good example is our relationship with layout. The way the production pipeline works, the camera is done first and then we animate to the camera. But, we know we’re going to make changes with the camera. I got together with the heads of layout early on and said, “Okay, we know when we’re in production, we’re going to need camera changes because the camera is going to follow the characters.” So let’s say an animator is going to show their shot to the director at 4:30. It’s 11:00 in the morning and they need to make a camera change before 4:30. What's the process? Do they go to layout and say, “Hey, can you change the camera?” Or, do they change the camera themselves? How do we handle this? First of all, we want to keep animators moving. Second of all, we need to make sure the animators don't mess up any of the cinematography. We need to make sure we track everything that's going on so we don't accidently lose a request during production. We don't want people changing the cameras and then not telling effects, because effects may already have been doing their work to the camera. And, if lighting has already gotten started we have to make sure any changes we make are tracked all the way through the pipeline.
Even though we know they were going to happen, in the past, camera changes had the potential to be devastating. Now, going in, we just make sure we have a solid communication flow for those type of changes.
DS: It sounds like you devoted a tremendous amount of upfront attention to making sure you knew how to handle most anything expected or unexpected that might happen on the production.
JS: Exactly. We knew that a whole bunch of stuff could come up like that. So, we just defined our processes early on so that when we were in the thick of production and everyone was firing on all cylinders, working as fast as they could, everybody knew exactly what the process was for anything unanticipated that might get thrown at them.
DS: Any tech innovations or new tools you used on the film that made things easier or more efficient?
JS: Definitely. There were two things specifically. Because the film is based on a 2D show, we had a real desire to try some of the same techniques that were done originally in 2D that make the animation feel special. One of those is using multiple limbs. If you look at old classic 2D animation, the characters freak out and start scrambling their legs to run off the screen. If you pause on any of those frames, you’ll see 5, 6, maybe 7 legs and lot of stretches and smears, things like that, that gave it a great sense of frenetic energy. We wanted to do that as well.
We’ve done multiple limbs in the past but it has always been a very, very time consuming process. We’ve had to pop outside of our pipeline to make it work. “On this shot, we want this penguin to have five limbs instead of the normal two. How are we going to do that?” It takes a lot of wrangling.
So on this show, I wanted the animators to have the option of making that choice whenever they wanted without having to pop out of the pipeline and deal with a bunch of stuff. So early on I did a test with Sherman where I animated him on twos, in Maya, jumping up, scrambling and running off screen. Even though Maya is outside our normal pipeline, it’s a great tool for testing ideas. I just duplicated one of his legs and used it to help sell the idea of multiple limbs. I showed it to Rob, and he said, “That's fantastic, we should do it.” So we developed a system within our pipeline where the animators could turn on three extra limbs for each limb, like three extra left arms, three extra right arms, three extra left legs and three extra right legs, which would then automatically flow through the pipeline. The system would generate geometry when needed, lighting could pick it up and effects was aware of it and wouldn't necessarily simulate clothing around it. We could use it wherever we wanted. Initially the idea was to use it just for big scrambles and takes, but animators started using it in general.
So with a fast character arm move, they’d throw in extra arms to help describe that arc a little bit just for a frame or two. It would add some thickness to the motion blur which helps the audience follow the action a little bit better. That was a really fun way for us to push our animation chops.
The other innovation involves the big challenges seeing our characters lit for the first time, trying to make sure their eyes are connecting properly. Usually what happens is when we animate, we’re seeing our characters lit with very low detail lighting. There are no real shadows. The textures aren't really finessed and the refraction is not there on characters in terms of their eyes. Then when we see it lit, we often go back and adjust the eye lines to make sure they really work in lighting. As soon as you shade the characters and have deep shading in the corner of the eyes, for example, it pulls the eyes together so you often get characters who look slightly cross eyed. The eye lines just don't look quite right in stereo. And making those changes means a rerender.
So I wanted to try and catch that earlier and not do as many rerenders, knowing our two main characters had glasses and glasses have refraction on them which tends to pull the eyes out and change what you can see. For example, if you take Sherman in three-quarter view, the far eye with glasses is pulled forward off the surface of his face. If you don't have refraction turned on, sometimes you can’t even see that far eye. Turn on refraction and all of a sudden you can see the eye. So it's important to pose it correctly. If we didn’t see any of that stuff until things were lit, we’d be going back quite a bit to adjust things in animation.
We developed a technique to turn on refraction inside our animation software so that we could see the character’s expressions before we sent things into lighting. That ended up saving a ton of time in terms of rerenders and having to go back and finesse things.
DS: Last thing. Looking back on this production, what gives you the most personal sense of satisfaction? What makes you say to yourself, “Okay, this is why I do this?”
JS: There are little tiny things throughout the production that give you that feeling. When I was a kid, I asked my dad once, “What’s the meaning of life? What’s the point of it all?” And he said to me, “Simple pleasures. Try and look for something every day that makes you happy.” So for me, during the course of such a long and difficult production, it's the little successes. It’s the little things we figure out about a character that make them a little bit more real. That keeps me going. Early on, when you figure out hey, Sherman looks really cool if you just skew him a little bit. He has a real pleasing look on his face. You can feel it emotionally. That to me is really satisfying and keeps me going.
At the end of the show, it’s when you are sitting in the audience and you are watching the film with people who haven’t worked on it, who have no idea what you’ve gone through to create it, and you hear a great reaction from them. That is really genuine. It’s an amazing moment. I went to Colorado last week and I watched the movie with a bunch of animation students. The moment at the end of the film where Mr. Peabody says, “I love you, Sherman,” the entire row behind me went “Awwwwwwww.” A moment like that just feels so good. The girl next to me actually burst into tears. It feels so wonderful to know all the hard work, all the trouble and exploration and ideas that worked and didn’t work, all the stuff you put into the show for three years, you can have an audience actually feel this inanimate object, this character that does not exist in any way except for the work you and your team did pushing points around frame by frame, for them to feel like this character really feels love for another character, it’s phenomenal! It makes you feel absolutely jazzed and excited about the next project and the project after that.
Dan Sarto is Publisher and Editor-in-Chief of Animation World Network.