EPISODE TRANSCRIPT

Morgan Williams: Even though MoCap has actually been around for quite a while, I saw a very early setup at Windlight Studios in Minneapolis in the '90s. MoCap has recently really come of age, in particular with Andy Serkis's revolutionary performance as Golem and Lord of the Rings. That really kind of brought MoCap to the forefront as a really remarkable means of creating amazing character performances. 

We since have had, of course, Planet of the Apes and a lot of other examples of MoCap being used to tremendous effect in movies and TV. And now, with the new creative cloud 15 Adobe is introducing it's character animation tool which will be a part of after-effects and this is a simple MoCap system using your WebCam or built-in camera for facial recognition and it also has a kind of ability to do recordable real-time limb control kind of with click-and-drag kind of an approach. And it's pretty amazing, it's a pretty exciting tool. I'm really excited to see what people do with it. I has some great possibilities. 

I had a student recently come to me and was sort of asking about MoCap and his question was basically, "Isn't MoCap just gonna replace manual let's say character animation, and by manual I mean an animator making the poses and doing the tweens, and just manually workin' that out rather than using a tool like MoCap. Isn't that just gonna replace character animation?" And he was basically saying it to me in that situation in sort of a way of saying "Well, why do I have to learn this? Why do I have to learn this thing if eventually it's just all gonna be MoCap?" And to a degree it's a valid question and I think it's a question that is worth thinking about. 

And the other question that I've been asked by people or an attitude that I've encountered with people, "Is MoCap cheating? Is it essentially a cheat in place of character animation, or is it a kind of way around it or something like that?" And I think these are questions worth thinking about as we enter an age where Adobe has given us built-in MoCap or we can hack a connect to do MoCap with a connect pretty easily these days. It's getting so that we are more and more in control of technology that can allow us to do some pretty amazing things. 

So let's think a little bit about how MoCap fits in, what is it's place in the world or character animation. And I'd like, for the sake of this conversation, to kind of once and for all just do away with any distinction between puppeteering and animation. I truly believe they are the same art form at the root and I've talking about that in previous podcasts, and I think it's significant with MoCap because I think quite obviously MoCap has a more direct parallel in a sense with puppeteering than with sort of classical "manual animation." But I think at their root the goal is the same, to bring a character to life, to help tell a story or communicate an idea. That's what we're all trying to do here. We're all trying to bring a character to life, communicate a story or an idea. I think we can boil it down pretty well to that. 

So to me, MoCap just seems to be another tool for the animator or the puppeteer or the animateer or the puppimator, or whatever you want to call it. But it's just another tool, it's just another tool in our tool kit. And if it helps us to achieve that goal of bringing that character to life and telling that story or communicating that idea, then hallelujah it should be used. I believe that animators have always been technical innovators and they shouldn't stop being technical innovators, and this is a technical innovation that we can use. 

But, I think we can also look at it in a broader context of just being a part of a set of tools that's always been available to animators and puppeteers also. For example, a Muppet with a body suit like Sweetums, if you're familiar with the Muppets, or Big Bird from Sesame Street. These are all body suit Muppets where basically a human climbs into a suit and performs as that character in that suit. I think you can easily see that MoCap is essentially a virtualized version of that same concept. Andy Serkis is essentially wearing a Golem suit or wearing a King Kong suit. So it's a pretty direct parallel there. 

And in terms of animation I think the parallel is Rotoscoping or using very closely following video reference which animators have been doing since the beginning of animation. Disney used Rotoscoping like crazy, especially in the early films. And they used it not only to create great animation but they also used it to teach their animators and it's was a great way for the animators to learn. It still is, Rotoscoping is still a great way to learn about animation. 

I tell you guys all the time to research, look at video reference, videotape yourself doing the action. Now that's certainly not seen as "cheating." Sometimes Rotoscoping is sometimes argued as being "cheating" but I think that's bogus when you look at the work that Disney and other animators did with Rotoscoping, some of it absolutely breathtaking. 

We can also say to a lesser degree that using an oscillating expression, a group of oscillating expressions to create a walk cycle or using a rig with ann auto-walk system is in a similar vein, it's an assist for the animator. And all of these things can be of great benefit. They can also be misused. There are certainly examples in every one of these cases of these things being misused. Now Sweetums or Big Bird or any of the bodysuit Muppets, those are performed by amazing, gifted, talented physical performers. But you've also seen, say in a cheesy old movie you've seen a very weak performance of somebody in a gorilla suit let's say who puts nothing into it and doesn't seem to be even trying particularly to move in a very interesting or entertaining or charming way. It's the quality of the person in the suit, right? Who's in the suit? If the person in the suit is a great performer and knows how to move and be charming and tell a story through their motions then you're in good shape. But if it's just some random in a suit, not so much. 

Rotoscoping also can be good and bad. Rotoscoping can be brilliantly again. Once again, you have to have a skilled performer performing that Rotoscope but you also have to have the ability to finesse and stylize that Rotoscoping. And a good example of this would be the wicked stepmother from Disney's Snow White film. That was created using a lot of Rotoscoping because it was extremely realistic and you have to recall that at that time Disney had not really done a lot of very realistic characters. Most of their characters were Mickey and Pluto, and they very broad, very cartoony characters. So when they had to animate the wicked queen they settled on a character design that was very realistic and so they had to create more realistic animation. And to be quite honest they weren't all up to the task artistically or creatively yet. That came later at that studio. 

So they used Rotoscoping and it was party how they learned. But if you look at the performance of the wicked queen in Snow White you can see the Rotoscoping, you can see how closely it is drawn from life. But you'll also see that it has been stylized and exaggerated and finessed by a skilled animator who understood the principles of animation. So it wasn't just used raw essentially, and so the performance is striking and very beautiful. 

Now, an example of Rotoscoping being used poorly would be the original Lord of the Rings animated film that was directed by Ralph Bakshi. Ralph Bakshi was quite outspoken at the time and made a lot of animators very angry because he very loudly proclaimed that Rotoscoping was gonna completely replace traditional character animation. Well, part of the reason that didn't happen is because this film was horrible, and it was horrible for a lot of reasons quite frankly. But the animation in particular, which again was all Rotoscoped, really doesn't work. Well, why not? It was traced from real life. They got a bunch of actors to dress up like the characters and jump around in a warehouse basically, and then they were Rotoscoped by the animators. 

Well, here's the thing. We've been talking throughout this class about good poses, good staging, good silhouettes, careful timing, and we've been talking about how you have to exaggerate wit animation because you have to communicate more broadly because you don't have the subtlety of the actual live person. Well, guess what? In the Lord of the Rings film they just exactly traced the performances of a bunch of human actors and they were decent performers and decent actors. But guess what? Bad silhouettes, bad staging, no exaggeration. Those characters are transformed into cartoons by being flattened, by the drawing and painting process, and the stylization of the characters that was done by the artists Rotoscoping. As soon as that abstraction is put into place, the merely natural movements of the actors weren't enough. They fell flat in the guise of this stylized drawing. They needed exaggeration and stylizing in the movement itself in order to make it communicate through those characters. So when you watch that film, the animation is painfully flat and lacks a lot of attitude and dynamism and action that you would see in an average Disney movie. 

So, I think that's really a significant thing in thinking about tools. Tools can be used well. Tools can be misused. Rotoscoping can be a great tool. Rotoscoping can be a disaster if it's seen as a complete end and of itself rather than just a tool that will help you get to an end. And I think that's the important distinction there. 

So, MoCap like Rotoscoping, can be used well and it can be used poorly. 

So let's kind of talk about some of the pros and cons of MoCap and I'm basically going to answer the question, "Is MoCap gonna replace manual character animation?" And I'm gonna say no, it's not going to and I think for a lot of really good reasons.  

So, one of the first reasons we've already sort of been talking about with the Rotoscoping and with the puppeteers in the body suits, which is that in all of those cases those were performed by highly skilled actors and physical performers. Andy Serkis is an amazing artist. We are not all amazing artists in terms of physical performance. We may be amazing artists in other ways, but we are all not Andy Serkis. So, for example, if you go to the Adobe website and you look at their little video introduction to the character animation tool, the character animation is horrible, I mean really, really horrible, horrible and bad. First of all it's a fairly simplistic tool and secondly because the guy performing the character he's just kind of waggling his head around and kind of wiggling the arm around and there's no sense or timing there's no sense of story, there's no sense of expression. Okay?

So, again, if you're gonna use this great tool that Adobe is gonna give us all, if you don't care at all about the performance of your character it might work great. But if you care at all about the performance of your character and you want some control over that performance, you better start taking some movement classes or start taking some acting classes. Because if you want your character animation to look good but you want to use motion capture technology then you have to become as great a performer as you can be when you have 100% control over every pose and every frame and every but of timing which you get with manual character animation.

If you are using motion capture, you need Andy Serkis, or a dancer, or a skilled actor or a skilled performer to perform if you want the performance of your character to be good. If you understand a little bit about character animation you probably will already be able to perform a little better because you'll understand maybe to exaggerate your anticipation to maybe hold on poses a little bit longer. Your study of character animation is the study of performance. So, it's gonna help you with the performance anyway. So there's a reason why we still need to study character animation. 

The other reason is that MoCap needs to be finessed. They almost never use that MoCap information raw, and for many reasons. And the people that finesse MoCap data have to understand movement and animation. MoCap data has to be cleaned up. It has to be tweaked to reflect the sometimes radically different portions between the performer and the character, just for simple things like making sure the knees don't pass through each other or the feet go through the floor or the hands go into the chest because the fingers are longer. If Andy Serkis taps his chest at Golem, Golem's got those long crazy fingers, you have to reprogram that so that it accounts for Golem's longer fingers. So you can't just use the data raw anyway. Someone has to take that data and translate it, and that person better know something about movement and better know something about animation. 

The other reason is, we talked about with the Rotoscoping too. MoCap data sometimes needs to be exaggerated or stylized in order to communicate properly through the character. A tilt of the head for Andy Serkis as King Kong may not in every case be entirely enough and the animators who clean up that MoCap data they may choose to exaggerate it a little or hold on it a little. It's another opportunity for the director to tweak the performance by tweaking the MoCap data. Now they could call up Andy and fly him back, and put him all back in the suit and get the whole crew all out and redo the scene. But if they've got it there in the computer and they can just say "Well, just hold on that posed a little bit longer," they can do that without going through all that expense. And again for that to happen there has to be some understanding of movement and animation. So MoCap is always gonna have the need for people to have that kind of a background in my opinion. 

And the other big reason is that there's gonna be sometimes where MoCap just don't work. Andy Serkis and any other human performer you hire is a human. Their bodies will only do so many things. And what if you need a snake, or what if you need an eight-legged insect, or what is you need some other crazy creature, some fantastic amazing creature with 8 heads and 16 arms, and slime pods that it squishes around on? At a certain level what can Andy Serkis do with that? Maybe the face, maybe the facial expressions, maybe part of the movement? But again you're gonna have to have an animator for some of that really extreme motion. 

And when it comes to animals too, animal training and animal performers they are expensive, you have to deal with tremendous regulations dealing with animal performers. Getting them to wear sensors and perform just so it gonna way out of the reach of most productions. In most situations you're not gonna be able to MoCap an animal. So just being able to deal with an animal is another reason why MoCap isn't gonna work in every situation. 

So, no, MoCap's not gonna replace manual character animation. Will it be a tool that will help us with our work? I think it already is, and it is already helping us and we can already see that, and I think it will be as a character animation tool is gonna be a great addition to our tool kit. I would use it in a heartbeat for a really fast turnaround job. I know enough about animation that I think I could perform something pretty well pretty quick. It would be great for background characters. It would be a great way to very, very quickly get background characters moving potentially. 

The other potential I see for this tool is maybe live puppet shows. That could be really show. Live digital puppet shows. What an awesome idea. Honestly as I say this out loud I'd be willing to bet $1000 somebody is already doing this with a Connect with a hack to Connect. I'm sure it's out there, I probably just need to be pointed in the direction. 

So I think there's a lot of amazing opportunity there. And I think again we need to remember that all tools for animators are good if they're used with an understanding of our fundamentals and if they serve the story or the idea being communicated. If they bring that character to life in an appropriate way, if they serve the story, the serve the idea, and they help get the job done then those tools are a good thing.

Animators again have always been innovators. It's part of what I love about being an animator. As I've said before we're like a squishy blend of artist and technician, geek and painter, drawer and computer programmer. How cool is that? I think it's one of the greatest things that animators have always had and have always brought to the table.

So let's not shy away from the tools. Let's embrace them, let's discover them, let's push them. Let's make them make great work. But, let's also not use them as a crutch. Let's not let them lower our standards. Let's keep our standards high. Let's make the tools rise to our standards rather than lowering our work in order to let a tool simply do our job in a lazy way and let our standards just sort of fall by the wayside.

So use the tools but use them to be a force for good. Use the tools to fight for the truth of the character, for the justice of high quality winning over mediocrity, and for the animation way. I don't really think there's such a thing as the animation way, but it sounded cool.