Saturday, November 27, 2010

Behind the Scenes: Disney, Miyazaki, Paley

I’ve been reading around in Hayao Miyazaki’s Starting Point, 1979-1996, a collection of short pieces by and interviews with Miyazaki. Among them is a speech he gave to Japan’s Association of Scenario Writers in August 1994. He observes (p. 103):
So, why don’t we script the film in advance? Well, it’s mainly because this is the way my brain has been trained to work. I’ve tried writing the story out many times before, but even if I think a story works great in text, when we render it in continuity sketches it’s usually unusable. It just doesn’t work, and usually reveals how shallow my original ideas were.
A bit later (p. 104): “In the old days, we always thought getting the sketches first was essential, but the problem is that drawing the continuity sketches seemed to always get behind schedule…” [ellipsis in the original text]. What emerges from this discussion is that Miyazaki and his colleagues start animation before the story has been completely worked out in continuity sketches (that is, story-boarded). “If you keep working, making adjustments here and there, in the process you really develop a much better understanding of the characters you are trying to animate.”

Though Miyazaki doesn’t discuss either the voice acting, the musical scoring, or overall sound design, I assume that that isn’t done until the film is completely animated. Given his “work it out as we go” improvisational process, Miyazaki’s not going to have a speaking script until it’s all done, nor will he know the sound effects required nor the music.

What I find so interesting is that this process is quite different from what I had understood the “classic” process to be, the process that used by Disney, to name an obvious example. In that process full animation would be the last part of the process. Before that could happen a script would be written, characters and settings would be designed, a complete storyboard would be worked out, the actors would voice their parts and the music would be composed and recorded. Then the animators go to work, synching their drawings to the sound where necessary. Of course, things might not quite work out as planned up front, so some revisions might be necessary along the way. But the basic idea is to create a detailed plan – storyboard and sound – before starting to animate. That, I assume, is what Miyazaki had in mind when he asserted that in “the old days” they did the sketches first.

I have assumed that this classical process arose, at least in part, as a way of coordinating the activities of what, in a feature film production, would be a very large team of artists, animators, writers, inkers, actors, and so forth. What Miyazaki found, however, is that it didn’t work for him.

Is this merely a difference between his personal style and Disney’s, for example? We know that Disney stopped drawing in the mid-1920s and thereafter functioned strictly as a director, albeit as a director who demanded control over every detail and often micromanaged his projects into chaos. Miyazaki, however, has always done a great deal of drawing in his projects, often doing most of the story-boarding himself. But it’s not entirely obvious to me how that difference would itself translate into the very different overall management styles, the classic and the Miyazaki.

The fact is, nothing about the process of making an animated feature film is obvious to me. Frankly, it seems impossible, coordinating all that artistic work. But somehow it gets done.

Let’s consider a third animator/director, Nina Paley. Though she has done only one feature-length film, Sita Sings the Blues, it is of such high quality that I have no problem including her work in this discussion. That is to say, if the issue at hand in the process by which first-class animated features are made, then Sita qualifies Paley for consideration. Fortunately, Paley’s said quite a bit about how she made Sita, and evidence is scattered through-out the web – at her website and blog, the Wikipedia article about her contains useful links, and, of course, the interviews I’ve posted on New Savanna.

As is well-known, Paley wrote, designed, animated, and directed Sita Sings the Blues herself. The film emerged gradually as a response to the breakup of her marriage. Here’s her account of those early events:
Originally, I hoped to expel my demons of heartbreak with a single short film, Trial By Fire (2003). This set a pivotal scene from the Ramayana, Sita's walk through a funeral pyre, to Annette Hanshaw's 1929 rendition of Mean to Me. Trial By Fire won 2nd Place in New York's 2004 ASIFA-East Animation Festival, and screened in festivals in San Francisco, Latvia, and Red Bank, but I refrained from promoting it further. Audiences loved the design and animation, but were not sufficiently familiar with the Ramayana to really understand the story. Furthermore, my demons weren't adequately expressed; I was still tormented by grief and heartache. When another relationship failed in November of 2004, I saw only one course of action: I had to tell the whole Ramayana story from Sita's point of view. Sita Sings the Blues, a 72-minute feature, would be my salvation.

I began production in December 2004. I have since completed 20 minutes of animation, comprised of 6 musical chapters. In April, a popular weblog called BoingBoing reported on my work-in-progress; within hours, thousands of viewers were downloading the movie clips I posted online, temporarily shutting down my web site. Reviews began appearing on hundreds of other weblogs, all positive.
At roughly that time, with only 20 minutes of animation complete, she applied for a Guggenheim Fellowship and was successful in the 2006 round of funding. In her Guggenheim application she described a general plan for the film that finally emerged. That plan mentioned four visual styles and the running parallel between the events of her life and the events of Sita’s life. I have no idea how much she had planned out in detail at that time, but here’s my guess: not much.

The film emerged over time. There was never a detailed plan. Since Paley didn’t even start with the idea of creating a feature film, her overall process seems even more diffuse and improvisational than Miyazaki’s. Yet the final result is as tight and as well-formed as films by Miyazaki, or Disney. The demands of first-class art are unyielding; but the process by which one gets there, that can be flexible.

Finally, I note that Paley couldn’t have made Sita herself without computer technology. That brought the workload within the capability of a single person. Is it possible that Miyazaki’s relatively fluid process depended on decades of industry experience with the more rigid classical process used by (indeed, invented by) Disney and others? I don’t really know, but I suspect the answer is Yes.

Beyond that there is the question of aesthetic possibility: How is that affected by the process used to create the film?


  1. Essential incompleteness of animation and other cartooning

    Interesting stuff, Bill. Let me add some thoughts based on my experience working with my wife Martha Cornog on a graphic narrative – not animation, but a comic – called The Adventures of Princess Adele.

    My point is not to suggest that our work equals Miyazaki’s, but to say that he is quite accurate when he says that you can’t script the whole thing in advance. Let me explain.

    The plot of Adele centers on Princess Adele of Utopia. Utopia is a lovely modern-world realm, but Adele runs away from home (the palace) at age 18 to become a motorcycle courier. She rides Bikey, a talking motorcycle, not an anthropomorphic cartoon motorcycle, but a robot motorcycle who is a laid back kind of dude. We have written the whole thing – a set of short stories plus a full-length graphic novel, with detailed descriptions of the artwork and continuity. Here’s an example. Please note what we have and have not included in each panel description. Each of these descriptions serves as an armature for the artist to draw the scene, which, when finished, will contain a good many more details.

    Adele is sitting in a strange little bar listening to some quite loony dragons talking about the Cthulhu with a team of British and American explorers. (That’s H.P. Lovecraft’s Cthulhu – real little monsters, in other words.)

    (1) The dragon Bilpin, as the other dragons finish their tubs of beer and even the furry men are listening to him: “That there sneeze puts me in mind of another story about my old Uncle. One day, he notices a passel of them Thooloops [ = Cthulhu] bothering some sheep they found in the mountains near where you’re going. Now, I gotta explain something to ya so ya comprehends this here story.”

    (2) Adele and everyone else is listening. Bilpin: “Us dragons – ” here Bilpin tries to look modest, an impossible feat for a dragon “ – we doesn’t admire them Thooloops too much. In fact, to tell the truth, we disadmires them.” Murmurs of agreement from the listeners. Sir John, Todd, and the other explorers are looking dismayed.

    (3) Bilpin: “Accordingly, my old uncle, he swoops down on them Thooloops intending to cause some disarray in their ranks.” The listeners, including Adele although not the explorers, are starting to grin. Image: a dragon, swooping down, wings fully extended, on a troupe of Cthulhu attacking some sheep, which are bleating and whimpering.

    (4) Bilpin: “Now, turns out that some verminous little critter got into my old uncle’s nose hairs – ” A semi-close up of a verminous little critter tangled up in the dragon’s nasal feelers, with the dragon, somewhat cross-eyed, glaring at it.

    (5) Bilpin: “And that makes him sneeze, it does.” Other dragons: “Sure does!” “Hates that, I do!” The other listeners are wide-eyed with some difficult-to-describe emotions; it’s not every day that you hear a dragon talking about verminous little critters in their nasal feelers.

    (6) Bilpin, imitating his uncle: “AH-CHOO!” Image of uncle sneezing: The verminous critter in several pieces being tossed aside by a vast burst of flaming green dragon snot. Burning lumps of dragon snot are colliding with Cthulhu down below, igniting some of them.

    [The Adventures of Princess Adele, © 2010 Timothy Perper and Martha Cornog.]

    continued below

  2. Continued

    I want to point out two things. One concerns the details of what these various beings look like, and the other concerns what this episode is really all about. Neither can be specified in any detail in advance, because the artist will synthesize his or her own experiences and skills with the armature we’ve provided first to create rough then increasingly finalized artwork. At each step, details and events will change when – or if – it turns out that something isn’t working.

    For example, we have given no descriptions of the dragons, the Cthulhu (= Thooloops), the verminous little critter, the sheep, the explorers (Sir John, the archetypical British heroic type or Todd, an equally intrepid American hero type guy), or even Adele. Nonetheless, certain things come across loud and strong: Bilpin is a slangy kind of dragon with a wicked sense of humor; the audience, itself obviously fairly strange, thinks the story is very funny; and Sir John and the other explorers are appalled.

    And as a result we, the readers, are somehow aware that Bilpin is having a bit of wicked fun at the explorers’ expense – and so the sequence becomes satiric. It’s the “somehow” that no one can pre-plan or describe before the fact. That isn’t because we’re incompetent – and neither are Miyazaki nor Nina Paley. It’s because the real punchline is an emergent from the descriptions: it occurs not on the page but in the reader’s mind. So the description is essentially incomplete, not accidentally abbreviated or shortened using a known short-hand, but essentially so. It leaves out what a purely prose writer might consider to be the heart of the matter because in a drawing script we cannot specify how to draw or invoke that “somehow” – a “somehow” the scene is really about.

    We might add a sidebar as an “instruction to the artist” explaining the satiric intention and the point of Bilpin’s joking story, but the reader will never see that sidebar. Rather, the reader will see how the artist interprets it into lines on paper. Until those lines on paper actually exist, the work is necessarily incomplete.

    That conclusion holds for all scripts purporting to describe animation or a cartoon-drawn narrative.

  3. One additional point. A drawing script, like a film shooting script, is a working document, not the final artwork. It is crafted with certain features and characteristics that suit the medium but do not exist in the final form of the work. So analyzing a drawing or shooting script requires that we understand the medium - and that is what Miyazaki and Paley are talking about.

  4. Interesting, Tim. Art is sensitive to details, and the details don't exist until you produce the full work. They can't be guessed-at in advance.

    Mike Barrier just put up a post that gives a look at Disney's Alice in Wonderland late in the game, but not quite done. Børge Ring descibes a work print that was sent to Denmark for preparation of the Danish dub:

  5. True – artists add and change details as they work, and perhaps overwork, any piece of art. But that process might reflect simple disorganization, lack of self-discipline, or lack of directorial decisiveness – it doesn’t have to be innate to the process of animating or cartooning. Michael Barrier’s quote about Børge Ring’s animation suggests a laissez-faire, easy-come-easy-go attitude on the part of the Disney organization that would drive a time-and-motion efficiency expert up the walls – because there doesn’t seem to be any reason for such chaos.

    But I am suggesting something else. As my example shows, there are two kinds of material that are not specified in the script (or in later preliminary drawings as well). One of these represent details added later by the artist(s) and perhaps changed many times – in Adele, the nature of the “verminous little critter” who gets blown apart when the dragon sneezes. But that kind of incompleteness is of interest only to people who study the micro-history of animation, e.g., of Disney’s Alice, and will disappear by the time the final print is produced. The viewer does not know or care about such incompleteness; it’s an issue only to the people doing the work.

    The other kind of incompleteness centers on issues that cannot be defined so easily so early in the project, no matter how disciplined the artist is or how dictatorial the director. These deal with how the audience reacts to specific images. These are issues central to the artwork, but no one can specify them beforehand. In Adele, these center (for example) on our realization that Bilpin is pulling the explorers’ collective legs with his story of green flaming dragon snot. Neither the writer nor the director can say “Draw this scene so it’s clear that Bilpin is having fun at the explorers’ expense” – or, more precisely, a writer or director can say that, but it won’t do any good. An audience can see the final or semi-final artwork and still miss the point completely – which means that the whole scene has to be redrawn from the start.

    Barrier’s quote from Børge Ring actually gives an example of this kind of intrinsic incompleteness – his comment that the image of Alice crying in the preliminary film was unpleasant. No one planned to make her unpleasant or aversive; that happened because the early drawings went that way and had to be changed. The problem cannot be solved by telling the artist to make her dress green, not blue. It is not a problem of details, but of what the art means and communicates. Then we find out about these problems only when the art is field-tested, so to speak, and when an audience says “I didn’t like Alice at all!”

    Now, however, we are much closer to the actual craft of animation and cartooning. The writer and artist most know their craft sufficiently well so they avoid such mistakes from the outset, or at least can avoid the most egregious of such mistakes. You can recolor Alice’s dress all you want; that detail is not important. Something is amiss of how her character was portrayed. That problem cuts to the heart of these artforms, and represents a considerable step upward from selecting a color for her dress. It is also the step where Miyazaki shows his greatest mastery of his craft. But it is not a matter of details. It is much deeper.

  6. Miyazaki talks of how something makes sense in a scenario, but when you try to make it work in production, nope! it doesn't work.

  7. Yes, that is part of the phenomenon I'm discussing. It is much deeper than details, but centers on form and emergent characteristics of the art. Resolving such problems needs a great deal more knowledge than of how to draw faces.

  8. Well, if the problem is with a face, then, in some sense, the problem must be one of drawing faces, no? Some aspects of drawing faces are easily isolated and so are relatively easy to talk about. Other things are all but impossible to isolate and discuss, but are nonetheless quite real.

  9. Hmmm. What I meant concerns the ability to draw anatomically accurate faces. If one cannot do that, then one is incompetent as an artist, but I'm not discussing such failures. I'm talking about the abilities to create an illustratable sequence of events that convey not only a plot (= a narrative) but also character, and that contain/communicate a series of points not all of which can be illustrated (= drawn) directly. Let me explain because the answer takes us far beyond mere anatomically accurate drawing.

    The reason I included the scenes from Adele is to show by means of a real example exactly what is involved. If the artist can't draw a recognizable dragon, then that's it -- the story of Adele is done for, gone, dead. It next takes a higher level of artistic skill to draw a dragon trying to be modest, and an even higher level of skill to draw a dragon trying to look modest but failing, as happens with Bilpin. But even assuming that the artist can do all those things, then we still aren't home free.

    IMO, it isn't a matter of being able to isolate and discuss these problems. That can be very easy, for example the Alice picture mentioned before that is so unpleasant. Very easy to isolate and discuss; much harder to solve! With that kind of problem, we reach a new level of the discussion -- the ability to draw character. The artist and writer must decide at some point in the process what kind of character Alice or Bilpin are. "She's a young girl" or "He's a dragon" doesn't cut it, because there are too many kinds of young girls or dragons. Instead, we must decide that Bilpin is a slangy dragon with a wicked sense of humor OR that Bilpin is a slow-speaking ominous creature who frightens most everyone who meets him OR -- and so on (and on and on) through a long list of possibilities. Martha and I opted for the first. You or someone else would make a different choice, but next problem is still the same. How do you convey those character choices in a sequence of drawings?

    Miyazaki, among others, has said that it is easy to create (write) a script with a sustainable plot and characters only to discover when trying to draw them that the whole thing bombs out. But since I’m trying to illustrate the principle concretely, not in glittering generalities, let me return to Bilpin the dragon. We can easily draw a Disneyesque funny dragon, but that bombs out also: it is not what we intended. Bilpin isn’t a Disneyesque funny dragon. So let me be concrete. The script sequence I gave is actually the second one we tried, and the first didn’t work. (More coming.)

  10. It involved Bilpin explaining in his slangy fashion that old Uncle Gilpin first dropped some sheep on them Thooloops (picture of a mangy sheep being dropped on some Cthulhu jumping up and down) but the sheep squashes the Cthulhu when it falls. Sound effect: Skwoosh-splat. Then Gilpin picks up a Cthulhu and drops him on the sheep, so this time the Cthulhu is squashed, and the sheep run around bleating and whimpering and stomping on the fallen Cthulhu.

    Now that certainly works in the script – but then we asked ourselves, “Yeah, so what’s the point? Does that sequence show that Bilpin has a wicked sense of humor?” The answer is No, not really – it merely shows a dragon dropping sheep on Cthulhu and vice versa. If we bounce that story off the explorers and others in the audience, they will merely look blank or say “Gee, why hurt the sheep? This must be a really vicious dragon…”

    “OK,” we said to ourselves, “no good. Back to the drawing boards.” And this time we concocted the verminous little critter and flaming green dragon snot. And we added something else: this time, we included in the story itself a mirror image of our own audience of readers, out here, who can have two reactions. One is to think THAT is very funny and the other is to announce that their delicate sensibilities have been OFFENDED by this Bilpin’s tale.

    In the story, those two readerships are represented first by the other dragons, by Adele, and by the bears – all of whom think this is really very funny - and, second, by the explorers, whose sensibilities are upset and who are appalled by the tale. So we, the readers, are now IN the story as invisible but real members of its audience. Thus, if you, as a reader, are offended by green dragon snot – you are not alone. You are with the explorers, but not with the dragons, Adele, and the bears. Or you too, as a reader, think it is very funny, and then you are with them but not with the explorers – and you, again as a reader, realize that Bilpin’s story is a send-up of the explorers and of a certain feigned delicacy that doesn’t want to deal with dragons, that is, with the real world. So now the story has a sous-texte: if you are going to live out here, you’d better be able to deal with dragons and green flaming dragon snot because otherwise reality is gonna getcha real good. And this version is now much closer to showing that Yes, Bilpin has a wicked sense of humor.

  11. I'd only add, Tim, that you're talking about the use of a written scenario/script in the course of producing a manga while Miyazaki is talking about producing anime. It's not just the written scenarios, though a part of the process, are not fully adequate, neither are continuity sketches (storyboards). He also observes that the fact that a story exists in an excellent manga realization doesn't mean that it is suitable for anime. Maybe yes, maybe no, but even it so, the manga obviously cannot be treated as a plan for the anime. The story must be reconceived anew, specifically for anime realization.

  12. You're missing my major point. Miyazaki (and I) are discussing a problem that has nothing to do with the translation of manga into anime, or of a novel into a film for that matter. In the course of creating a new work, whether from scratch or from another written source, problems arise that cannot be foreseen. Those problems deal with the nature of the ending medium, as I have been trying to illustrate. To put that differently, if you "reconceive" the story anew, you will still have these problems. Since you haven't dealt with my own example -- of the Adele story -- perhaps you would prefer to deal with the problem that Barrier's blog mentions, of the image of Alice being unexpectedly "unpleasant." That problem, which is precisely defined, has nothing at all to do with translating a novel into animation or a written script into manga. It has to do with the nature of animation and skill in creating it.

  13. ...if you "reconceive" the story anew, you will still have these problems.

    Of course.

    ...problems arise that cannot be foreseen.

    And the use of scripts and storyboards is an attempt to conceptualize, to plan, and so, in part, to foresee.

  14. Yes, but I think you've left the thought unfinished. Miyazaki is saying that the process of scripting and storyboarding can readily fail -- from which it follows that we cannot readily or easily make such prophecies. Instead, we have to try it, and discover where it goes awry, as I am trying to explain with my example of Adele. And then we discover that we didn't think through the story carefully enough, not because we're incompetent, but because stories have emergent properties that depend, among other things, on how readers react. Alice being "unpleasant" is an excellent example, and so is Bilpin's story. Handling these problems is part of the craft writers, artists, directors, and so on learn and slowly master. The writer or artist is in continuous reciprocal engagement with the work, in a kind of dialectics that eventually produces something new at the end. But the novelty is not at the level of colors of dresses or how faces are drawn, but at the level of structure of the particular piece of artwork itself. If that isn't clear, then take a second look at the Bilpin sequence, because it's quite clear there.

  15. Given what you're saying, Tim, the question becomes: how is it possible to plan out a film in advance, script and storyboard, and the execute to plan? I assume that, to some extent, that that's an ideal and that, in practice, the plan gets modified.

    But I also assume that, when Miyazaki says he's given up on even giving lip service to the ideal, that that DOES represent a change in procedure. Miyazaki's more "organic" certainly makes more sense to me. And it also suggests the proceeding according to the classical procedure, even if one must deviate from the plan, imposes some kind of limitation on what you can do in a film. What kind of limitation?