Wednesday, November 22, 2006

The Collaborative Nature of Animated Acting

I'm making a presentation about my masters thesis this week at a graduate symposium at York University. Below is the basis for what my thesis will be when I finish my research and start writing it in December. I've already written a bit about this subject, but this is the longest and most organized piece I've done about it to date. I know that the things I'm writing about vary based on time period, studio, director, etc. but fundamentally, there's far more collaboration in an animated performance than in a live action one. Whether you agree or disagree, feel free to leave comments.

My name is Mark Mayerson. I’m in the studies stream and I’m doing a thesis on the collaborative nature of animated acting. My professional experience has been as an animator, as well as writing and directing for animation. I’m currently teaching animation at Sheridan College.

Within animation, there’s a cliché that an animator is an actor with a pencil. That makes it sound like a there’s one to one correlation. Animators are just like actors; they just use different tools. My experience is that animators are significantly different from actors in terms of how animated film production is structured, and my thesis is an examination of how and why this is the case.

In live action, we think of the actor as being central to a role. Marlon Brando is Vito Corleone in The Godfather. We understand that the script, direction, lighting, costuming, etc. all contribute to the performance, but while we can’t quantify how much Brando contributes, we have a gut feeling that he is responsible for the majority of the performance. Replace him with another actor and the role is different. It’s Brando’s body, voice and movement whenever the character of Vito Corleone is on screen and most of all it’s Brando’s brain driving it all.

In other cases, such as the James Bond films, we can easily see how changing the actor affects the role. Sean Connery is different than Roger Moore, Pierce Brosnan, etc. If you use the TV series Bewitched as an example, two different actors, Dick York and Dick Sargent, both played the role of Darren Stephens yet nobody confused them. They were different in same the part.

Imagine watching a dubbed film. The on-screen actor has acted the role in the usual way. The voice actor, adding his or her voice after the visuals have been created, is constrained in several ways. The timing and the emotions are dictated by what’s on the screen. The voice actor has to work within the limitations of the visuals if the dubbing is going to be successful.

If the dubbed version is the only version available to you, how do you judge the success of the performance? If the voice work is poor, is it due to the voice actor or the constraints placed on him or her? If the acting is successful, how much is due to the on-screen actor and how much is due to the voice actor? Who is responsible for the performance?

Animators are in this situation but in reverse. Where dubbing takes place after live action has been filmed, in animation the creation of the motion takes place after the voice tracks have been recorded. A performance is split between two collaborators who may never meet. Where the voice actor gets to interpret the script, the animator is forced to interpret the voice actor. Failing to do so results in a disjointed result on the screen.

This is only the tip of the iceberg for the ways that animated performances are fragmented. I’ve got to go into some history here.

Initially, animated films were the work of single artists. Winsor McCay, J. Stuart Blackton and Emile Cohl were among the earliest animators and they did all the important work themselves. Of the three, McCay was the one most interested in acting. In his Gertie the Dinosaur of 1913, he animated the dinosaur by himself. We can say that McCay is Gertie the same way that Brando is Vito Corleone.

Beginning around 1913, animation studios sprang up to produce series of films based on continuing characters such as Mutt and Jeff. This is where fragmentation entered the process. An animator who worked on the Mutt and Jeff cartoons recalled that they would pick a setting, like the beach, and then each animator would grab a section. By verbal agreement, one animator would deal with gags about a shark and another animator would deal with gags about mermaids. Each animator was responsible for writing his own section, drawing the background art and designing any characters he introduced in the segment.

At this point in animation history, cartoons were mostly about gags and acting wasn’t even on the radar. In a single cartoon, four or five animators would each take turns playing the roles of Mutt and Jeff. They wouldn’t see each other’s work until it was shot and cut together, so creating a coherent performance was not going to happen.

Animated acting came into its own in the 1930’s at Disney. The paradox is that Disney accomplished this by increasing the amount of collaboration rather than reducing it. The difference was that the control was centralized in Disney’s own hands. He used separate artists to design characters, to draw storyboards, to animate the characters and to re-draw the characters “on model” for the final frames of the film. Directors and musicians would plan out the tempo of every scene and voice actors would perform the dialogue prior to animation.
Every step of this process contributs to the performance. Character designers create the look of a character specifically to evoke a particular response from the audience, so animators, unlike actors, are forced to come to terms with an unfamiliar appearance that they don’t control.

(Powerpoint slide of early and final Queen designs from Snow White here. In live action terms, it’s the difference between Rosanne Barr and Angelina Jolie. Or in historical terms relative to Snow White, Margaret Dumont and Gale Sondergaard.)

Where a live actor can take his or her body for granted and is confident about how it moves, an animator never knows what age, size, shape or species will need to be animated in the next scene. In live action terms, it’s as if actors had their brains transplanted into different bodies. John Wayne becomes Don Knotts and Don Knotts becomes Marilyn Monroe.

In animation, it’s more severe. Several animators get their brains transplanted into a single body, and they have to grapple with how that body is going to move and how they’re going to create a consistent performance.

In live action, storyboards are mainly used for shot composition and continuity. In animation, because of the low shooting ratio, storyboards concern themselves with acting. On a feature film, an animator may produce as little as 5 seconds a week. In TV, the animator will probably produce between thirty seconds and a minute a week. At that rate, revisions take far longer than doing another take on a live action set. A production has to clearly communicate the emotional progression of each scene to the animators to enable them to get it right (or close to right) on the first attempt. But that means that the conception of the performance (both audio and visual) exists before the animator starts work.

(Story reel/Animation comparison from Mulan here).

In many ways, the job of the animator is more a synthesizer than a creator. All the work done prior to animation must be woven together into a coherent whole. This is different than an actor being free to create an interpretation with the approval of the director.

When an animator’s work is finished, there are still hands that shape it before it reaches the screen. At Disney, animators were encouraged to concentrate on performance more than drawing. Their assistants took their rough drawings and polished them into what finally appeared on screen. Assistants also often took care of secondary bits of animation on the character, such as hair, drapery, or costume details. (Slide of Sleeping Beauty rough and clean-up here.) Their contributions aim to be invisible, but there is the potential for introducing problems. In the 1939 animated feature Gulliver’s Travels, there are scenes of the Princess Glory character where the assistant work detracts from the animation. The character’s hair shifts around on her head, giving the impression of a loose wig. (Princess Glory clip here.)

The problem I’m dealing with is determining where an animated performance is centered. In the case of a character like Bugs Bunny, he had at least three character designers, nine directors, eight writers, one voice actor and more than two dozen animators in a fifteen year period. Is there anybody whose influence is as large as Brando’s on Don Corleone?

In feature films, animators are sometimes cast by character, but the amount of screen time a character has often requires that the animator supervise other animators in order to hit the deadline. On Disney’s Beauty and the Beast, Glen Keane was the animator in charge of the Beast character, but he supervised seven other animators, in addition to assistant animators, breakdown artists and inbetweeners who all had to draw the character consistently as well as merging their work into a coherent performance. There’s a scene of the Beast learning to eat with a spoon that was animated by Tom Sito and based on actions he observed when his cat tried to eat from a spoon. Sito pitched the idea to Keane who accepted it, but the scene would have been done differently had another animator handled it.

This thesis will also examine live action reference, whether rotoscoped (which is tracing live action to turn it into drawings for cel animation) or motion captured (which uses hardware and software to capture a live performance and convert it into a computer animated character). Both these approaches are another layer of collaboration.

Not a lot has been written specifically on this topic; because people who write about animation are rarely animation professionals, they don’t understand the process or the repercussions of it. I’ve collected studio documentation from Disney, Warner Bros, Lantz and MGM that lists who animated specific scenes in a film. This has given me insight into the styles of individual animators as well as directors, who have very different approaches to assigning animators to characters or scenes. In addition, there is a large body of interviews with animation professionals that reveal the behind-the-scenes working methods.

With the exception of writings by Michael Barrier, who argues that casting by character results in better performances than casting by sequence, and Ed Hooks, who never deals with the issue of fragmentation, there really isn’t any critical writing on the nature of animated acting. In part, this thesis is an argument against Barrier’s position, as I don’t think he understands the degree of fragmentation that exists. Even when casting by character, I don’t know if animators ever have the same level of control that a live actor takes for granted.

I am familiar with an animator’s thought processes from personal experience. I am not familiar with an actor’s thought processes. I’m curious to know how similar they are and if differences are due to how theirjobs are organized or due to their nature of their crafts. I need to read more about acting, though I’m not sure that the material will turn out to be relevant to this thesis.

In writing this thesis, I have to be sensitive to how I use jargon and how much explanation is needed. If I was writing about what a film editor does or what a steadicam is, I know the readers would understand what I was talking about. However, I don’t think that the readers are going to understand what a inbetweener does or what an exposure sheet is, so I’ve got to make sure that I explain things so that readers can follow my argument.

10 comments:

Steve Schnier said...

Hi Mark,
Great presentation. Very clear and concise (as usual). Good luck on your thesis.

Benjamin De Schrijver said...

This looks great - and interesting, just as your previous posts about this topic.

However, I wouldn't agree if you said that animators aren't actors with a pencil. The cliche image of it is wrong, but in essence, the phrase isn't. You just can't fall into the trap of having a too narrow image of what an actor is. Stage acting is in many ways different from film acting. The same with animation... It's very different from both of those. It's more distinct than those two, but it still is acting.

I've always been an advocate of the casting by character system, instead of the casting by scene one. Supervising animator on one specific character in a feature film is the position I aspire to. With good collaboration with the voice actor, and good supervising of your team, you really do create a character. One of your animators might animate a scene differently than you would, but they still feed on you, on your interpretation and creation of the character. Ideally, you'd be part of the storyteam as well, so that, just as when live action directors rehearse (or an actor produces/directs), the story could be altered to ideally fit that creation/interpretation.
You could think of your animators sort of as your understudies in a play. Their performance won't be the same, but their choices still have to come from/relate to how you perform it.
For similar reasons, I'm also an advocate of actually holding script readings/rehearsals between the supervising animator and voice actor. Though I have no idea how that collaboration between works now. But the ideal scenario would, at least to me, seem to be that the supervising animator does the first big job, than works very closely with the voice actor, and the rest of the team are "understudies" of that duo.

Once this is done, you can still cast by scene - like Keane did with Beast (and I assume most of his characters), so the performance at least is perfectly consistent in that scene. It's my observation that when films cast by shot or scene, it's usually by shot (or a series of shots), instead of scene. Which can't really add to the overall performance, can it? And even if they were cast by scene, I don't believe you really have a chance to get to that ultimate core of the character. You can create very distinct, strong characters, but I don't think it could ever be of "oscar-winning" calliber.

I guess what it comes down to, for me, is that character cast acting is closer to "real" acting. Where acting can be a real form of self expression. Which is what I really want animation to be for me. And which is what I don't think you can *really* get when you cast by shot.

Cooked Art said...

Hey Mark,

Great post and great work on the presentation. A really great topic and definitely an interesting read.

I definetly have to say from the outside looking in that scene casting has more going for it than character casting. I personally think that supervising animators for each character can exist to ensure continuity of character, but where the different really lies is when multiple characters are to be animated on the screen and must coexist and/or contact each other in the scene. It is possible, although a major headache, to animate these characters with two seperate animators, but if the character within a scene is a secondary character than you get the 'extras syndrome' where the animator wants to play up the movement and action of the background character which can potentially take away from the main action of the scene. If you are casting by scene, such ego warring would not occur - the animator would simply hand each character appropriately as it is needed unto the scene. I would also think that the directors would cast scenes according to what scenes they have already completed - that is, they wouldn't have an animator work on a character for the entire film and near the end of production animate a completely different character that they have never handled before, especially if the shot contained only this character. Personally, for the scenes that would contain multiple characters, I would that this would aid in the distinct personalities of each character being played off each other by one person, who gets to the big picture of the scene. There is definetly a valid argument, however, for getting into the 'core' of the character which is definetly better achieved by animating only one character throughout the entire film - but having said that, it seems that the advantages definetly point to casting for scenes.

Benjamin De Schrijver said...

It's true that some scenes require that one animator does all the work, especially when characters are touching. But having a character-based casting system doesn't mean you have to do that with every single shot. Belle and Beast/Prince kissing was completely animated by Glen Keane. So was Jim playing with his mom on the bed in Treasure Planet. In other cases, the 2nd character is often "pre-animated" by the animator of the lead character of the shot. An example that comes to mind is Tarzan surfing through the trees with Jane in his arms.

And in the other cases, in which multiple character shots/scenes are cast by character, it's just a case of collaboration, professionalism, skill and good direction. Which should be present at all times, shouldn't it? It's the animator's job to have the professionalism and skill to not steal a scene, or anything like that. Being scared that the animator doesn't do what is basically his/her job shouldn't be an argument against a system.

So when you cast by character, you really get the best of both worlds. Your animators get the chance to really get into the core of the character, and when needed, an animator can take on one or more extra characters for that scene. Having a scene-bases system obviously gives you that last advantage when it's needed, but it might keep you from reaching the full potential of other shots or scenes.

Though I must say I'm approaching everything from a handdrawn feature standpoint, as being an animator in that field is what I hope to achieve. For TV, that level of acting obviously isn't usually required. And in CG, the way of blocking out a scene really differs from animator to animator. Though possible with efficient collaborative planning/thumbnailing, if you're an animator who blocks in stepped keys, à la 2D, it'd be really hard to match that with the other animator, if he blocks in splines, or worse, layers his animation, starting with animating the hips, then legs/upper body, etc. Small changes in eg. timing could be made more easily and quickly than in handdrawn, but overall I think it'd be a more difficult task.

Nancy said...

I agree with Mr. De Schrijver. Just because a cliche exists does not make it true.

We are not actors. We are CREATORS of LIFE.
An actor interprets a role. An animator brings an imaginary character to life. We are responsible for its existence, not just its attitude and appearance.

So animators are control freaks, God-manque's , but not actors. I always thought we had more in common with dancers than actors; dancers also work from the imagination to create forms that may have little relation to the human body.

I've worked on the per-scene and on the casting system at various studios. Really, one works as well as the other. If the animator is good enough, they can contribute to the total, rounded personality of a character when there are no leads.

BAMBI had no lead animators; the cartoon and realistic animators were in two divisions, but the cartoonists shifted around on the different characters, and that film has some of the most outstanding acting of any Disney film. Ollie Johnston and Milt Kahl have great scenes on Thumper in the same sequence, only a shot away from one another--the drawings don't change, but each scene reflects a different attitude toward the rebellious small boy-rabbit's acting and actions.

Likewise the Warner brothers cartoons are excellent; in that studio, the animator did all the characters.

Casting by lead can 'lead' to repetitive performances when the same people are cast every time. I can identify schtick from different animators on some roles--and it is no coincidence that the secondary characters in some of the later Disney drawn features have more interesting acting than some of the leads.

Of course leads keep features on track. But again, there is BAMBI to show us that it wasn't always necessary to have this system in place.

Nancy said...

I would also like to add that breaking up a scene by character, so that multiple animators worked on one shot, was a very inefficient system especially if the characters had to contact each other. It was sometimes workable, but I had to animate Hades and the Fates together in HERCULES for one shot, since the unit animator did not understand the gag action in the scene.

Hans Perk said...

Having seen 32 features go through my studio, I have gotten the distinct impression that the truth (if any) lies somewhere in the middle. The number one important thing is, that the animator in question must be good enough to deal with whatever is thrown at him. This seems to be simple, but it isn't: not all animators can be of the caliber of say Nancy, or Glen Keane. And when that is the case, you as director need to work around this. You cast the animators to whatever you think they can handle. We see this at Disney's in the 30s. In the early days, the scenes were pretty much dealt out to whomever was free, but then 'preferred' animators got certain bits. Later, when everyone seemed on par, the films were just cut up in pretty much equal bits, but still with an eye on, which animator was most appropriate for action, and which for acting. The system of casting was not a rigid system: it was based on production needs. We are not in a perfect world where we can make up a team of animators like 'give me three Milt Kahl's, three Frank Thomasses, a Kimball etc. We deal with the situation at hand, as best we can. And how good is the director in analyzing this? On our earliest films, we did a lot of character casting, and cross-overs, and found that the scenes that were cast by character had more personality, as the other scenes were so bogged down by drawing issues that they were mechanical more than acting. Later, we found that many animators could well deal with all characters, and they handled whatever scenes were finished in layout for them to work on. Still, it became obvious that certain animators had more flair for certain characters...

And talking of directors, how good are they at communicating their needs? This is where I want t o insert the notion of the good oldfashioned hand-out, the way Dave Hand described it in the lecture on my blog. The ONLY time I have seen a satisfactory hand-out on sub-contracting work has been Bill Kroyers video hand-outs to us on FernGully in 1990! And that was still only one-way! It is a forgotten art and needs looking in to as a way of making a consistent product. It is all part of a whole...

Boris Hiestand said...

agreed, it must be somewhere in the middle.
A wonderful sequence comes to mind in 'Aladdin' where the princess and him are on a roof having a conversation. Absolutely brilliant acting. Glen Keane did Aladdin in Burbank and Mark Henn did the princess in Orlando. Ludicrous, insane and stupid. But it's one of the strongest moments in that film! Beautiful acting in animation.
Of course story came up with that rather than the animators, then there's layout, and the voice acting is half the job done. Milt Kahl said this as well about Geraldine Paige who did Medusa's voice. Half the job done, or more.

It's ironic though that all the examples in these comments are examples from Disney films, where politics and hierarchy have always determined wether or not an animator gets to do any 'acting' in the first place. Keane and Henn could do scenes like that becasue they were at the top.
A couple of posts ago you discussed the phenomenon of being a 'tool'!!!

http://mayersononanimation.blogspot.com/2006/11/you-are-tool.html#links

Hitchcock saw his actors as 'tools'. After all actors are hired to (like Nancy suggests) bring to life the vision and ideas of the director and scriptwriter.

It'd be nice to go into independent animation as well, where a director is often also the animator. They have total control, but are they actors?

I always had a big interest in seeing the drafts from Disney features, and I don't think I'm alone. It would be amazing if you could post them, so we can see exactly who animated what on some of those films...

Anonymous said...

One problem that I can see in the Character based casting, is that there are a limited number of animators working on every character throughout the film, while this may help with keeping the performance consistent, it could also hurt the film by limiting the diversity of the acting which you would obviously get from more animtors who get involved.

Each animator by personal experience in life brings something unique to the character, like Tom sito's Beast scene, perhaps no other animator had that particular inspiration from life, and there are many more examples in which an animator had drawn from a personal experience. And when you limit the number of animators, you automaticaly limit the end result.

I personaly think what Pixar did with the Incredibles is the best system, a hybrid of the two, where you cast by scenes and so bring in all your animation power to every character, and yet have a few supervising animators who, by being involved on a daily basis with animators, create a general guideline for keeping the acting consistent.

Stephanie Ciccarelli said...

Hi Mark,

I am extremely impressed by your thoughts, interpretation and research.

Coming from the other end of the spectrum (I work for Voices.com in London, ON), I found your thesis material to be quite interesting.

It is quite rare that one comes across something of this nature and in such detail online where voice overs and animators are concerned.

I'd love to know more about your thesis so that we might share your ideas with the voice actors everywhere.

Please contact me via email to discuss:

stephanie (at) voices.com

Cheers,

Stephanie Ciccarelli