OF MICE AND MEGABYTES
PART II · MICE FOLLIES
by Maurice Molyneaux
When not writing articles for ST-Log, Maurice Molyneaux designs game graphics, consults for software companies and creates animated cartoon productions using microcomputers. Despite a ridiculously French name, he claims to having been born in Vicenza, Italy, and denies vicious rumors that he eats escargots and calamari while computing. His DELPHI user-name is MAURICEM.
For those who missed last issue (shame on you!), I discussed how, in late 1986, I obtained copies of what was eventually to become Epyx's Art & Film Director. At the time, these Hungarian-designed programs were to have been published in the U.S. by Broderbund. With the software, I had produced a series of animations for a Star Trek game proposal, which were put together on videotape. A copy of the video was taken to Broderbund (by Stephen Friedman, American representative of the Hungarian company Novo-trade Software) in order to show them what kinds of things could be done with the software.
I received a call from Steve a few weeks after delivering the tape. He said, "Broderbund wants to know how much you would charge to make a video for them to use to promote Art and Film Director."
Naughty But Mice
Surprise! It was the last thing I'd expected to come out of that Star Trek demo! A challenge to be sure. They made no specifics; I had to submit my own concept. Fortunately, I got an idea almost right away. Months earlier I had been toying around with the idea of producing a computer-generated cartoon that would feature a rather smart-alecky mouse, and which would demonstrate various forms of animation (cel, stop motion, claymation) and poke fun at them (to show that Saturday-morning cartoon characters are flat and lifeless the mouse would slap a Smurf, causing it to spin, and we would see that indeed it does have the depth of cardboard!).
I adapted this concept. The video would compare Art and Film Director to other paint and animation programs, and, specifically, poke fun at Saturday-morning or Japanese cartoons. The mouse character would be pitted against the obligatory narrator and the programs themselves. In other words, whenever a feature was shown, it would be demonstrated by using it on the mouse.
The written outline I submitted was accepted, and I was asked to submit a storyboard. I had begun storyboarding (sometimes brainstorming with Vince Reynolds) parallel with writing the proposal, but it still took a long time to get the whole thing hammered out. I was determined to make a "cartoon," not just a series of unrelated animated vignettes, so the structure and pacing of the whole thing, the way gags built on gags, etc, was important. The following, from the original written proposal, makes clear what my intentions were:
"...The thrust...is such that the viewer waits anxiously to see just what can happen [to the poor character] next, and the improbability of the situations keeps it funny...By making the demo tape something of a film, complete with a continuing character, people will more likely stop and watch, to see the whole demo and how it ultimately ends, than if we just threw a bunch of pictures on the screen."
At this point I decided to create a cartoon human character. I doodled a short, chubby, bald fellow with a huge round nose. The reasons for creating him were twofold. First, the abrasive personality planned for the mouse was inappropriate for a few scenes I had in mind. Secondly, I was worried that I might not be able to do a good job creating the mouse. If the mouse proved impractical, this character would take his place.
A Star is Bored
The completed preliminary storyboard consisted of 112 "thumbnail" sketches. These were assembled, four to a page, with text describing the action, narration, etc. The basic gist of the video was this:
The human character was to be involved in three opening sequences: a "presented by" gag and two scenes demonstrating the limitations of traditional art forms. Next, we would poke fun at "crude" paint programs, then show off a flood of Art Director features. A bunch of graphic bits would then crash together to form the mouse, and Art Director tools would be demonstrated on him. Next, the mouse would be knocked around by the Film Director interface, then would run a movie projector and show some miscellaneous animations. Finally, the projector bulb would burn out, and the narrator, deciding there was nothing else to show, would erase the mouse animation. The mouse would then push the GEM Desktop screen off the monitor, jump out of the screen and throw the Art & Film Director disk at the "camera," cracking its lens.
I estimated the demo would run six to seven minutes. Broderbund wanted ten, but I thought it would be too much work to try for that much time.
The storyboards were approved, but changes were requested. The only serious blow was that I had to remove the sequence making fun of Japanese cartoons, because Broderbund was afraid that viewers passing by at a show would see only that part and think that scene was indicative of the programs' capabilities! (The storyboard called for choppy movement, obnoxious voices, and, of course, out-of synch dialog—just like a real Japanese cartoon.)
By this time the video had been christened "Notions in Motion." It was the beginning of April, 1987. Broderbund wanted the tape for the June CES.
War and Pieces
One of the first things done was the design of the human character—actually making shape tables. I had to figure out a way to make a truly flexible character within the limits of the hardware and software.
(A quick technical discussion: Film Director creates animation using two basic building blocks: polygons and patterns. A polygon is a closed, unfilled, multi-sided outline defined with Film Director's own tools. Patterns are rectangular graphic blocks [with the background color transparent to other colors] "clipped" from pattern pages [full-screen, low-resolution pictures], which are usually imported from Art Director. These elements can be combined into more complex elements, such as Groups, Actors and Stages, which are then combined to make the frames of the animation.)
The idea of drawing a separate image for each pose was quickly discarded. Sixteen screens of pattern data could be loaded into the then-new versions of Film Director, but that wasn't enough for the complex movements the storyboards called for, especially with a character who was to be almost a half-screen tall.
The demo animation of a little boy that came with Film Director was interesting. The boy's shape table took only a single screen and contained all of his parts (hands, feet, etc.), each drawn in numerous positions. Great, except that in order to get such a flexible character in such a small space, the animator(s) had left out a few small details—specifically, limbs. The boy had no arms, legs or neck! He consisted of only a head, body, two hands and two feet. Simple to animate, easy to create shape tables for, yes. What I was after, no. I wanted to make a traditional cartoon character, not a bunch of disjointed parts.
I dug into my animation references and found an article about an obscure animated film called "Twice Upon a Time." That film used what is called the "Lumage" process, which refers to backlit animation using cutout parts. They had libraries of arms, legs and so on that they would put together on an animation board. I had seen the film and was not really impressed by the animation. Yet I was struck by the resemblance between the technique used in Film Director's boy demo and the Lumage process.
It was clear that the best way to design the characters was in pieces. However, it would be necessary to create shape tables of great complexity. I shuddered at the idea of having to draw scores of heads in every conceivable position. I needed to find a better way—and I did. The character would be composed of pieces, yes, but smaller pieces than the boy had. Rather than drawing complete heads, the head would itself be composed of separate elements, which I could then alter individually. Eyes and other facial parts were created independent of the actual head object, and thus I could have a character twitch an eyebrow a pixel or shift an eye or perform any number of small movements, without having to have complete parts for each and every pose.
A price was paid for this flexibility. In a given frame of animation, the human character ended up consisting of twice as many parts as the boy in the original Film Director demo, and his shape tables, when properly organized, filled two and a half screens.
Vince labeled the disk containing these shape tables "tiny little man." Seeing this, I dubbed the character "Tiny Mann."
The first shot to be animated was called "The Sculptor" wherein Tiny played artist, chiseling on a statue and causing it to crumble. To show how on a computer you can undo your last action, the statue reassembled itself. Of course, it's not possible to reassemble such traditional art forms—so smash! the statue collapsed again.
The storyboards called for a statue of "The Thinker." I got the bright idea to swipe the picture of Michelangelo's "Moses" from the Art Director title screen. Not only did it eliminate the need to draw a statue from scratch, but it tied in with the product much better. The picture was resized, recolored, and a base added to the statue. Copies were made of chunks of it to produce the rubble required. A hammer, chisel and Tiny were all that was needed to begin. I sat down and animated, trying to give Tiny the right kind of walk.
Problems were encountered when the statue shattered. I was animating 40 chunks of rubble, plus 12 pieces of Tiny; thus, things slowed to a crawl during playback. A little quick thinking solved this problem, and I had my first completed animation for the video.
I played it back, and, despite my knowledge of each and every frame, it surprised me. It was funny! It was funny in a way I hadn't expected it to be. Somehow, the plight of this little character got through to me.
I next tackled one of the more difficult sequences. It called for Tiny to play maestro and conduct color changes on a painting in time with the opening part of Bach's Tocatta and Fugue in D Minor (the music to be dubbed onto the final videotape).
Redesigning Tiny was no problem. Change his clothes to black, make his shirt a shirt-and-jacket combo, make his hands into white gloves (a five-second job using Art Director's X-Color option), add some tux tails, make a white "Leopold Stokowski" wig, and voila!
I had planned to time key points in the music, break these down into a chart based on frames per second and animate according to that. Afterwards, I would adjust frames to compensate for uneven playback speed on the computer. (It varies because the more screen updating that has to be done, the slower the machine is.) With a planned deadline looming before me, I skipped the timing chart and winged it. Picture a man endlessly replaying the opening strains of Bach's Tocatta and Fugue in D Minor, adding, deleting, changing frames, all to get the animation and music in synch. Five nights of this I went through! I thought I'd never want to hear that bloody piece of music again!
Complicating matters, I couldn't just take Tiny's hands from point A to point B and stop. Real animation is the art of movement, and frozen poses are a major no-no! To keep Tiny from freezing up, I had to make his hands quiver as he held them aloft. I also had to make sure to vary the movements slightly so that the quivering wouldn't fall into a pattern and look mechanical.
The fun part of this otherwise painful exercise was animating all the secondary action: his tux tails flapping, his posture, body twist, etc Tiny "anticipates" thrusting his hand up by first pulling it back. At one point, he swings his arm so forcefully that one foot comes off the floor!
The final animation needed of Tiny was to be the opening shot. In it, one letter of the software company's name (originally the "o" in Broderbund, later the "Y" in EPYX) has fallen out of the logo, and Tiny goes to ridiculous extremes to put it in place. He was to throw it, catapult it with a seesaw contraption, and finally blast it in place using a cannon.
As props were drawn, I began to feel that the seesaw gag wasn't going to work. Tiny would have to leave the screen, push the thing in, set the letter on it, and then move around and jump on it. Then he'd have to get rid of it when it didn't work! That would take far too much time.
Vince suggested that maybe Tiny should pilot an airplane over the logo and drop the letter like a bomb. I almost rejected this outright, because there was no room above the logo. But then I decided to try it. To show the plane we'd pan up so the logo would be at screen bottom, and pan down again when the letter dropped. The only problem was to make it absolutely clear that the "camera," and not the logo, was moving. To facilitate this, clouds were added during the pan so the viewer's eye would have something to relate the logo to. A simple solution, and no one has ever mistaken the pan for anything else.
From Hand to Mouse
With the completion of Tiny and a number of miscellaneous animation sequences (including an example of a "crude" paint program called "NEO-Lith, Version 40,000 BC"), it was time to roll out the star of the show. Oh, and I'd given him a name too: "Megabit Mouse" (or just plain "Mega" for short).
I had lavished a lot of care in designing Megabit's parts. Tiny's design had forced me to keep his movements kind of stiff, because of thick limbs and other design flaws. These did not affect Tiny, as he was intended to be kind of "stuffy" but Megabit was supposed to be emotional, flamboyant and flexible. His design reflected the experience with Tiny. Mega's limbs were thinner, allowing for greater range of movement. Thus, more poses were created. His feet were bigger and more car-toony. Best of all, his face was far more expressive, this primarily due to wide, well-defined eyes. (Tiny's eyes were usually shut, and you never see his mouth.)
All along I'd feared that I wouldn't be able to do a decent job animating a mouse's tail. I find freehand drawing on a computer difficult, even with a good graphics tablet. Drawing them any other way resulted in their looking stiff. I solved this problem by drawing dozens of tail positions on paper, then tracing those positions onto a graphics tablet. A little cleanup, a lot of reorganizing, and I had my tails. Good thing too, because the tail really ended up accenting Mega's expressions. It straightens when he's startled, twitches when he's mad, and so forth.
Megabit's first appearance is in a scene where a bunch of graphic bits fly together and, bang!, form a mouse. This was simple. I posed Mega in his "formed" position, then duplicated the frame. I went back to the first one and pulled Mega apart, scattering his components to the screen edges. A quick three-frame Tween (to create in-between positions) and, voila!, instant mouse! After that, all he had to do was look around, then get turned into a paintbrush.
The following shots were not so simple. The next sequence has Mega carrying in a tool chest, tossing the Art Director function icons into the on-screen "toolbox," as well as getting smacked by an alert box, dragged across the menu bar by the pointer, dropped on his butt, magnified, and knocked down by the toolbox. Finally, he shatters the toolbox with a hammer!
Next shot: Mega's colors are altered, and antlers are added to his head to change him from "A MOUSE" to "A MOOSE." He grabs the eraser from the toolbox to erase the antlers, is poked by the cursor and erases his own face, then fumbles for the UNDO button. Once his face is restored, he is trapped in a window, filled around, scraped through so we can see his skeleton and finally spun. He dizzily collapses, stars spinning over his head.
Then he is flattened against the inside of the monitor glass, knocked around by the Film Director user interface, after which he walks into a Saturday morning cartoon and decks a Smurf-type character. Following that he runs a projector to show the miscellaneous animation.
Animating Mega wasn't much different than animating Tiny, but it was more difficult in some ways. Mega's design was far more complex, averaging 25 to 26 patterns per frame as opposed to Tiny's 12 (which also necessitated running Film Director's playback at full speed just to keep everything running smoothly). Mega's parts filled four entire screens of shape data and consisted of approximately 600 patterns! Imagine trying to pick just the right hand out of 220 possible choices!
(A comparison: Aegis Animator's eel function saves to disk each image clipped from a page of artwork. To animate Megabit using Aegis would require saving over 600 cel files to disk! And you would have to try to remember which of those 600 files was the one you wanted on a given frame, unless you wished to load by trial and error. This is why I consider Film Director to be the only true cel system available on the ST.)
Mega's tail was a mixed blessing. While it did add a lot to his appeal, it was a pain to work with. If I had it swaying from side to side and needed to add or delete a few frames during that movement, I usually ended up having to adjust the tail over the next 20 to 30 frames just to get everything smoothed out, otherwise the movement would be uneven or jerky.
To further complicate matters, these shots required me to simulate GEM drop-down menus, alert boxes, mouse pointers, the Art and Film Director toolboxes, and on and on!
How long did it take? Even with fully completed shape tables and all patterns and groups defined, it still took an average of two 16-hour days to animate each of the longer sequences (averaging 1.25 minutes each). Mind you, this was two days of just putting cels on the screen and manipulating them. No drawing, just animating.
Fortunately, Broderbund had decided to skip having the demo for June CES and wanted it for later in the year. It was agreed that I would present a "Director's rough cut videotape" to Steve Friedman by September 20, 1987, that he would deliver to Broderbund for review.
The challenge of getting the animation on tape again reared its ugly head. I couldn't just borrow any RF/composite-equipped 520ST, for most of the animation I'd created required one megabyte of RAM to run. My old mid-1985 edition 520ST, upgraded to one meg, had no RF output and thus no composite video line to take to a VCR. No RGB-to-composite converters could be found either. The situation looked desperate, but my faithful ST dealer came to my rescue. His tech swapped out the old Revision B motherboard in my 520ST for a new Revision H board complete with RF modulator, moved my RAM expansion board onto it, and only charged me $20!
In spite of all my last-minute rushing, it became clear that the final scene wasn't going to be ready by the deadline. I had hoped to get access to some Genlock equipment to composite Megabit onto some video of a real computer desk. That hadn't happened. The rough cut had to be received before Steve left for Europe on September 21st, so I couldn't delay. On the 16th I dumped all the animations (except the ending) to VHS videotape, rough-dubbed the Bach music and shipped it off. I expected comments in a few weeks, followed by a round of final corrections (and the completion of the ending), and then the okay to go to the final tape I was in for a big surprise.
Shortly after Steve's return from Europe, Broderbund cancelled the programs. They would not ship after all, meaning they no longer needed the video.
Broderbund and I came to terms where they relinquished all rights to the video to me, and I did not demand any further payment. My hope was that some other publisher would pick up the programs and want the video.
Months later, in early 1988, Epyx decided to pick up the programs and the video. A contract was signed, and I had to get back to work and complete it. The final shot was still needed. Trouble was, I had soured to the original ending. While visually interesting (Megabit jumping out of the monitor and shattering the camera lens with a disk), it ended on a negative note (Mega very angry), not a funny one. I felt this was not right.
Finally, I decided that when the projector bulb burned out, the narrator would ask Mega what they should do next. Mega would think of things which the narrator would reject. Finally, Mega would pull out a disk and toss it to the narrator, whose hand would appear on screen to catch it. The disk contained the end credits, and that's what they would show. But, just before those ran, Mega would again be clobbered by an alert box (which would state "The End") and have the antlers placed on his head one last time. But, rather than getting mad, he'd just shrug as the "iris" closed to black.
Bingo! The shot allowed for some funny interaction between the narrator and Megabit, and it ended on an upbeat note. The only potential problem was how to have a near-life size digitized human hand come into view without the animation slowing to a crawl. A little eel trickery was all that was needed, and the video was done.
Well, almost. The Smurf-type character was too Smurfy for Epyx, so I had to change it. Also, as the programs had been ported to the Apple IIGS, I had to remove all ST-specific references. The title was also changed from "Notions in Motion" to "Art & Film Director" to make it clear what the tape was.
A final VHS rough cut tape was made. Then, as I tried to get the narrator's voice and the music recorded, I was informed that the tape had to be ready for the Applefest in Boston (May '88)—in less than two weeks! Furthermore, the rough cut had come out to 14 minutes—fully twice my originally planned length! Epyx wanted no more than 12.
A mad scramble was undertaken. Final changes were made, the soundtrack was recorded, paperwork was prepared for the video editor. I packed up a 1040ST, my hard drive, and my (then just arrived) Video Key and headed out to dump the animation onto three-quarter-inch videotape. All was taped, and everything was packed up and shipped off to the editor (Jim Yocom, at the Indiana Vocational Technical College) via overnight air freight.
After much straining from all involved parties, the editor's rough cut tape made an appearance at the Applefest in Boston. (The funny thing is, not a scrap of the video had been created on a IIGS!) A few weeks later, the release version of the tape made its debut at June CES in Chicago—a full year after the thing was originally planned to be finished.
Touch and Go
How is the final product? Personally, I was a bit disappointed. A last-minute order forced cutting two entire minutes of material, taking it from 12 to 10 minutes, throwing the pacing off in a few places. Also, due to the last-minute rush, there was no time for fine-tuning. Some lines of the narration needed a little more emphasis, and some of the sound on the animation was deadened in a few spots. Also, the original music which appears over the credits does its creator (Randy McClanahan) no credit. The music degraded over the course of rerecording, dubbing, etc., so it sounds uneven and a little off-tempo in places, although the original recording sounded just fine.
Not to say it came out badly. Indeed, considering the touch-and-go status of the project and the crash effort it took to get it finished, it came out very well. I particularly want to take a moment to mention some of the people who helped make it possible: Stephen Friedman of Software & Video arts for negotiating the deals and seeing the thing through; Randy McClanahan for music and the caveman's "Ugh!" (yes, that's him!); Mrs. Andy Morrison for narrating the prerelease cut; Vince Reynolds for his help in producing parts of the animation; Digital Vision for providing a color Computereyes digitizer; Jim Yocom for doing such a good job editing a less-than-perfect master and getting it to look so good in so little time; and, finally, Madeline Canepa and Joe Miller at Epyx for their help in getting it done at last!
Believe It or Else
And thus ends the story of how the video came to be. However, there is a subject I've avoided until now, and it is the real core of this whole matter.
At the Worcester Atari Show (in Worcester, Massachusetts, October, 1987), just before Broderbund dropped Art and Film Director, I gained some interesting and important insights. At the show, a lot of CAD-3D and Cyber animations were displayed, including an eight-minute, four-megabyte animation by Tom Hudson called "Spider Patrol." The odd thing was that people tended to look at those flashy demos, say "neat," and just go about their business.
Meanwhile, just for laughs, STLog's editor, Clayton Walnum, and I set up a 1040ST at the ANALOG booth, and I popped in the animations of Megabit (not the video tape; the actual computer animations). I left one of those running as I went around the show, and whenever I went by the ANALOG booth I was flabbergasted to see people crowding around that little monitor, watching the animation not just once but several times, and bringing other people over to see it. I am not attempting to toot my own horn here. What happened was that I hit on something that is fairly obvious, but most people don't realize.
To wit: Here we are, with a wealth of powerful graphics tools on a system within the price range of almost anyone And yet, the stigma of "cold" computers nicely fits the majority of animations and graphics created on them. They are technically intriguing, but the designers are so caught up in doing the neat technical trick that they don't make the work interesting—they don't make their graphics involving to those who view it. Megabit Mouse drew an awful lot of attention, and he really isn't doing anything phenomenally complex, nor is he even that impressive graphically. However, he does radiate a definite personality: he reacts to his situations, and, most importantly, he communicates these attitudes to the viewer. To paraphrase Joe Adamson: Megabit Mouse does not exist! Yet he lives!
Think about that a moment. We are often so wrapped up in making computers "productive," with word processors, databases, et al., that we miss many of the other tasks they are beautifully suited for. What I learned is that it is possible to make what is really nothing more than a bunch of colored dots on a screen be more than just technically impressive, but actually elicit reactions from an audience!
That, my friends, is one small example of the real power computers give us.
Author's Note: In honor of all those old Warner Brothers cartoons that inspired me during the making of the Art & Film Director video, I used the names of actual Warner cartoons as subheadings in this article, and also for the title—so don't blame me for the bad puns!