In this reprint from the April 1998 issue of Game Developer magazine, 3D artist Josh White outlines some best practices for keeping lines of communication clear and cordial in your art department.
Art direction is the communication between an original designer and the executor of the design. It's not the design process, and it's not the final design's execution. This month, we'll peek at the perils and delights of game art direction.
For example, let's say you want a portrait of your pet lizard. Your idea is to put the lizard on a nice branch with a desert scene as a backdrop (that's the design). You hire a professional photographer to actually take the photograph. You spend 15 minutes explaining how you want the shot to look and drawing a simple layout sketch (that's art direction). The photographer sets up the shot, takes some prints, and shows them to you (that's art execution). You provide criticism and ask the photographer to retake the shot (that's art direction).
Now that we have an idea of the scope of art direction, let's look at how it works in the game industry (obviously, this is a fictional story).
Kotz and Amy's Little Issue
Kotz is a wound-up monkey of an art director with a couple of solid hits under his belt. His enthusiasm is legendary -- with spittle flying and arms flapping, he spews out quirky artistic ideas nonstop and then leaps onto his drafting stool and slashes out his trademark designs, hooting "Ooo-ooo! " when he really gets in the groove.
Kotz was hired to design artwork for a game project called "DreamFly," a cartoonish, dreamy flight simulator that feels like a kid's book. There are little fairy creatures fluttering around a big, puffy airplane. The plane zooms over surreal blobs of landscape -- you know, concept stuff. Kotz loves this idea, and has already scrawled out a huge pile of sketches for the game artwork.
Now the project is ready to start, and Kotz's job is to find artists to build his design. He hires Amy, a skeptical veteran game artist, to create 2D sprites of the fairies dancing around the airplane. Kotz's initial direction to Amy consists of handing her a crude and childishly executed sketch of a fairy.
"So hey! Check it out! I bet you're just itching to build this little bugger!" Kotz chirps as Amy stares at the sketch in disbelief.
Let's freeze that frame and take a look at the problems so far. The sketch is completely inadequate for character definition -- there's not enough information. But there's another important problem developing here.
Problem 1: Staying Friends
Amy and Kotz are forging a relationship. This is an important interaction, and it's not going well. Neither of them is taking care to create a solid foundation for their working relationship, even though that's what they both want.
As Amy's boss, Kotz wants to communicate enthusiasm and good vibes, but he has set himself up for disaster by approaching the problem too lightly. Since this is the first project he and Amy will be collaborating on, he should have assessed the situation more carefully. He should get Amy's input on the art-direction process and make her feel comfortable giving feedback. Right now, Amy can't criticize anything without implying that Kotz screwed up -- but she needs to say something. That's a lose-lose situation.
What are Amy's options, given this situation? One option would be a response such as, "Umm... Kotz, I can't draw this." Her low, measured tones totally puncture his mood. "This sketch is, uh, really rough." Kotz's heart sinks. "What's wrong? It's a cool little fairy, right?" Amy's exasperation wells up as she dramatically sighs, "Oh, good grief... Kotz, this drawing is so far from a complete design! I can't work miracles."
Amy's reaction is justifiable since Kotz has obviously made some mistakes, but that kind of self-centered attitude won't get her far in her career. Even if she hates Kotz and is planning to quit, she'll do better if she doesn't burn bridges. How can she be honest without burning bridges? Amy can prevent conflict by buffering her negative reaction.
David Packard (of Hewlett-Packard fame) wrote a book called The HP Way (HarperBusiness, 1995), which describes Bill Hewlett's three-step process for critiquing ideas without dampening enthusiasm:
Upon first being approached by a creative inventor with unbridled enthusiasm for a new idea, Bill immediately put on a hat called "enthusiasm." He would listen, express excitement where appropriate and appreciation in general, while asking a few rather gentle and not too pointed questions. A few days later, he would get back to the inventor wearing a hat called "inquisition." This was the time for very pointed questions, a thorough probing of the idea, lots of give-and-take. Without a final decision the session was adjourned. Shortly there after, Bill would put on his "decision" hat and meet once again with the inventor. With appropriate logic and sensitivity, judgment was rendered and a decision made about the idea. This process provided the inventor with a sense of satisfaction, even when the decision went against the project -- a vitally important outcome for engendering continued enthusiasm and creativity.
Bill Hewlett was passing judgment, but Amy could use the spirit of this technique. Her first step is to recognize that Kotz's enthusiasm is very valuable to the project, even if it bugs her. As a veteran, she has on occasion worked with uninspired art directors who glumly create standard-looking artwork, which never makes for an exciting game. Good games have passion in them, and if she wants to make good games, she's got to understand and empathize with the passion, at least a little.
So her first reaction should be to the spirit of the sketch, not the problems with it. "I'm into drawing fairies! Back in art school, I drew these kind of wood-gnome-elf characters, kind of like this sketch, but more realistic-looking."
Kotz is curious. "Really?" he asks. "I'd love to see them! Were they photorealistic, or more like Norman Rockwell fake-real?"
"Well, sort of like... hard to say. I have some photos of the canvas I'll bring in tomorrow if you want. So anyway, what kind of outfit were you thinking?" and Amy steers the conversation around to gathering more design input.
Once she's gathered an overall sense for Kotz's vision of the character, Amy can start exploring the technical issues of creating game art that works in the engine. For example, she might think aloud, "Hmm, so if we've got four frames of animation, we could do a little wing-flap bobbing cycle. But maybe we couldn't get an eye-blink in there because it would happen too often."
From these musings Kotz would learn what Amy's limitations are, giving him a chance to calibrate Amy's judgment to match his. "Hey Amy, wanna just skip the eye-blinks? It's not a big deal, right?" After a few rounds of this, Amy and Kotz will each become familiar with the other's style and where they disagree. If Kotz knows Amy's style, he'll trust her judgment and avoid micromanaging her in the areas where they agree -- and he can keep control of the areas where they differ.
Problem 2: Art Direction
The second problem is the obvious one: the sketch doesn't say enough to build art. Kotz needs to provide a lot more information before Amy can build any production artwork. Of course, Kotz doesn't want to spend a week describing art that could be built in a day, so there's going to be a tradeoff between thorough communication and time. Here are some common ways in which art directors communicate their ideas:
Torn-out magazine photos, scenes in movies, well-known artwork (for example, the Statue of Liberty) -- a good reference gets an idea across quickly, but rarely matches the desired style exactly. Most art directors use them in combination or with limitations: "like this photo, but no neon signs and dirtier." The artistic reference can be a completely different subject from the planned artwork -- for example, a mermaid drawing that has the right type of color saturation and detail for a fairy.
Character sketches, top-down scene layouts, storyboard cel-frames -- they're all powerful methods of description. Often it's quicker to label special features than draw to draw them -- for example, write "dirty leather belt" with an arrow pointing at the waist. Making quick, useful sketches is an art discipline in itself, so we'll leave that for a different article.
Stories and essays are also common ways to describe artwork -- they are easy to compose, but leave a lot open to interpretation. Often, text descriptions are best for giving an overall sense for the character or scene, and will work best when combined with a few detail sketches. Word descriptions are good for technical issues, such as an exact list of a character's poses or the number of frames for looping animations.
References to Artist's Earlier Work.
This is a really powerful way to define style since the artist knows exactly how the artwork was created. This is one of the most powerful benefits of working with the same artists on multiple projects. Without that shared experience, it's often time consuming to convey something subjective such as style. If the director can get the artist's earlier work and can take the time to review the new art in terms of the old, the definitions are often very convincing.
For describing animation, a quick and powerful method of communication is to actually act out the motion in person -- a goofy strut, a wild swing, a sulky slouch-stance -- all these motions can be acted out in seconds, whereas good sketches could take days.
A good art director can creatively communicate a number of technical requirements as well. For 2D sprite art, here are some specific examples of what the art director should share:
Expectations of Quality and Style.
Artistic references tell the artist how much detail to create, as well as what kind of artistic style is sought.
There's a host of inevitable limitations: dimensions of sprites in pixels, how your program recognizes transparent pixels, possibility of antialiased edges, number of colors (and if you are using a fixed palette, a copy of the palette), file format to deliver in, naming conventions, and so on.
Camera and Lighting.
What angle should these be drawn from: Overhead? Isometric? Side view? What background will they be viewed against? Is there perspective? Should there be bright, shiny highlights?
We'll need a description of the unique attributes of each character. That first sketch is a great start for character definition, but ultimately, we'll need more information: Man or Woman? Age? Culture (as indicated by clothes).
A list of required animations is a must. Art directors should also communicate the motion style that they expect. The technical issues usually rear their ugly heads here, too: number of frames available, type of transitions, and so on.
Ways to Direct Artwork
The methodology of art direction is pretty complicated in and of itself. Whole books have been written on the subject (see References). Still, there are some specific techniques that can help Kotz and Amy forge a more productive working relationship.
Framing is the concept of putting limits on the desired style, rather than naming the style itself. It's like drawing the space around an image, or "proof by negation" in math. For example, Kotz might tell Amy, "This fairy is fatter than Tinkerbell, but not as adult-looking. It's not a baby or a standard Christian angel, either." When used in addition to descriptions of the character itself, framing is a great way to rule out huge regions of style.
After Kotz has tried to describe the concept, he wants to know if Amy really gets it or is just nodding along. Amy can demonstrate her understanding (and reveal areas she doesn't get) by offering some new design ideas that she thinks fit with Kotz's concept. For example, she might suggest that there be a fat little fairy and a tall thin fairy, and the two argue and tease each other. Kotz then can see that she's thinking of the fairies as having very clear, strong, goofy personalities, rather than being ethereal, ghost-like creatures.
There are some dangers with this approach. First, Amy has to expect that her suggestions will be shot down -- if she has easily injured pride, she may not be willing to offer suggestions. Also, if Kotz thinks Amy is trying to take design control and doesn't understand that Amy's trying to prove her understanding, he's not going to be O.K. with her input. Obviously, Amy should preface the talk with a diplomatic comment, such as, "Let me see if I've got the idea. I'm not actually suggesting we build anything new, but..." It's also possible that Kotz will like Amy's ideas and give her more creative control than he was planning.
Our story ends happily. After Kotz and Amy have an hour-long discussion about fairies, Kotz leaves with the idea that Amy is into the concept but needs more details before she can actually build anything. Amy thinks that Kotz is kooky, but is open to the reality of making games and understands what she needs. Their future is all roses, and the world will soon be blessed with the fruits of their relationship: a game with great artwork.