Me Memoirs: Myron Holman

For some unknown reason, the prospect of another Holman on the premises didn't frighten them at all. They flew me out for a two-week contract on one of the Navajo Folk Tales projects (Coyote-Rabbit), and I met about everybody: Lee Harrison III, President, CEO, and Mainspring Of The Whole Works, Ed Tajchmann, VP Engineering, whose thesis on electronic animation shaped the machines, Bill Altemus, (never did find out his formal title --he was just the man who managed magic in designing, building and maintaining the machines. He made video do what video had never even thought of doing before), Jim Duca, Bill Richardson, Dave Harvey, Mike Webster, Joe Juliano, Domenic Iaia, Peg McKechnie, Angie Hersch, and a host of others. I wound up with Red Daniels and Marsh Parker. Red was to be my CAESAR driver for the project. They showed me some of the work that had already been done: Bill Perez' "Quartettus Caesar," Lee's "You'll NEVER animate a character with a computer!." "Noise!" and a stack of others.

The place was located in an industrial area in Denver, a few blocks East of Federal Boulevard, on riverbottom land that, fortunately, never flooded again while we were there. It was a big concrete building, with floors at truck-bed height, and loading doors down the sides. CI had the offices at one end, fronting West Second Avenue. The rest of the building held several other companies, like Arvey Paper, and an outfit that processed snapshots. It was one building of many. The one across the parking to the East held Manley Popcorn for years. Occasionally we'd chip in and buy a five-pound bag. Those are close to the size of a steel drum, only paper, and we'd eat popcorn for days. I'd use a film can for a platter in the CAESAR room.

The whole place was alive with excitement: the TV showed the New York office building occupied by Dolphin Productions, then the only other user of Scanimate. One wall of the tall building was lying in a heap in the street. "HAR!" roared Red, "If I'd'a'been there, I'd've been walking in the door there," -he pointed at the bottom of the collapsed wall -- "just about that time." The whole company had their corporate fingers crossed: had the Scanimate ridden the wall down from the umptieth story, or had enough rubble land on it to demolish it, then the insurance would've paid the company to build another --a luscious contract. Luck, as usual, went the other way: Dolphin's Scanimate was finally found dusty but unharmed when the salvage crew cautiously made their way upstairs a week or so later, and the whole engineering crew glumly admitted they'd built too well..

A week into the project, they asked me to sign on, and flew my wife out to get her approval, too. "You know what's gonna happen, don't you?" asked Red, "It'll snow; they'll divert her flight into Albuquerque, and she'll spend the whole weekend in the airport, and hafta fly back without even seeing the place." I laughed. Saturday came; Fran's flight was due in the morning: I flung open the motel room's curtains and stared into a mass of falling snowflakes. All I could do was laugh.

She was actually only delayed an hour or so, and I got to watch some fascinating antics at Stapleton as the ground crews removed snow. I'd never seen a road grader operated at speed before, let alone see one do a Bootlegger Reverse at the end of the runway.

Red told me that the company was fresh back from its Los Angeles foray --they'd spent a year taking Tinseltown by storm, and now were home again, with the bank signing the checks. Red had been their Scanimate Operator there, and happily described the whole debacle as he saw it.

Red was something of a character, and quickly became a friend. He'd come to CI in an odd fashion. He'd been a cab driver in Denver, when a freak accident cost him an eye --he was umpiring a Little League game and was hit by a thrown ball. His company offered him a desk job, but instead he used the disability money to go back to school and study electronics. It took: he could build you a Scanimate or a CAESAR from parts, and he wound up helping Marsh Parker write programs for CAESAR's Honeywell 316... don't laugh; that was a heavy-duty computer back then...AND the program had to be written in the 316's machine language --none of this sissy UNIX stuff. Between his roaring, "ARR!" and his swashbuckling attitude toward obstacles, I could only think of him as "piratical." The glass eye helped. He had several good yarns involving it. And his appetite for jalapeno peppers was legendary. He could've matched Blackbeard the Pirate bite-for-bite through a barrel of jalapenos, and come out of the contest a happy winner.

Red did warn me, "When a job comes into here, if your house burns down, it's the fire department's problem." I should've listened, but I'd been four years away from animation, and it had left a hole in my life.

I'm afraid that I fell in love with animation at a very early age. About ten months old, I think. After I got past the stage of wanting to be Snow White's Eighth Dwarf, and learned that those were actually drawings moving up there on the screen, I started wondering how they did all that, and, since I came from a family of artists, had some rich reading material to give me clues. We lived in the Midwest --Kansas City-- an animation desert, so I had to do a lot of library research. Just another lunatic kid with flip movies in all of his books. I finally got an animation job in Kansas City, at Calvin Productions, which had a small animation oasis amid its production studio facilities, and spent five very happy years there, working with Orlan Hill, Ed Faust, Rich Corben, Clyde LeMaster, Polly Kirkpatrick, Dave Chavez, Barbara Ganote, and a bunch of other good people whose names will doubtless come back to me as soon as this epic is out of my hands.. It was one of those jobs in which, every now and then, you'd look at work you'd done, and tell yourself, "This can't be right; I can't do work that good." It's an addictive feeling: once felt, you want MORE of it. LOTS more of it.

So CAESAR was a tonic.

CAESAR: Computer Animated Episodes using Single Axis Rotation.

...which is going all around the barn for an acronym, but, hey. It was catchy.

The basic idea behind all of the Computer Image machines had the simplicity of genius: video has always been constrained to make a rectangular array, identical for both the camera that generates the signal, and the monitor that receives it. Why not play with the voltages that direct the electron beam that paints that image? What could you get with that? How could you steer the beam for a nonrectangular effect?

CAESAR and Scanimate were both second-generation descendants of ANIMAC. ANIMAC was the prototype machine, and, while shockingly powerful, had more than a few operational drawbacks. The worst was the tube technology used; as the tubes warmed, the settings would drift like the Kon-Tiki in a hurricane, and yesterday's move was next to impossible to redo today. So Lee & Company went in two directions:

Scanimate was basically a video synthesizer, set up and adjusted "on the fly" while CAESAR was similar, but under computer control, and capable of keyframe operation. Both of them used multiple oscillators to drive the video controls to warp and twist the 945-line "high-res" video (hey, in those days, that WAS high-res video). The distorted video was then scanned by a 525-line camera to convert it to standard NTSC video. Since the two cameras ran at different rates, it was inevitable that the scanning camera would always be looking at some data that was older than other data. No problem there, except that there was always a seam --the "scan bar"-- between the two regions: a diagonal line that always appeared somewhere in the picture. You could adjust the phase to place the scan bar somewhere less offensive --and if you taped in more than one pass, you might be able to hide it among the invisible elements.

But we tried hard not to do multipass work: when I came on, we were using two-inch quad VTR's, and not only was "sync" a laugh to watch --the VTR Operator would stretch to full extension to reach both start buttons with both index fingers, and stab them simoultaneously-- but nobody had ever heard of H-Phase, so we did not know that there were four nonidentical fields of video for every adjacent pair of frames, and we had to hit the right one to make a good edit. I've seen VTR operators near tears: when an edited cut went down badly, we had to back up and re-lay the previous scene. If the re-done cut into THAT scene went bad, then back we went to the scene previous to that. One night we had to rerecord the entire spot, after several hours of laying it down the first time, backing up one scene at a time all the way back to the start, and then rerecording each scene again, with more unsuccessful cuts.

The quads were huge, ungainly machines, prone to endless troubles, and the luckless VTR operator would spend a good part of each day fudging with the adjustments, especially when starting up. So Red animated a set of color bars on CAESAR. He let them stand normally for a few moments, to give the VTR guy time to start adjusting and trying to clean up the crummy-looking signal. Then the red bar started changing color. On the vectorscope, you'd watch the glowing dot leave the red box, and journey all around the scope, coming back to red. Then the bar would wobble, and topple, landing at the bottom of the screen, where the adjacent blue bar would fall on it. And somehow it wound up on the head end of most of our two-inch reels. Every few mornings I'd hear the VTR guy swear from two rooms away, and I wouldn't have to ask what had happened...

But multipass was often necessary: neither CAESAR, Scanimate, nor ANIMAC knew anything about pictures. Just video. So the artwork went in as Kodalith images, broken up into "sections". Scanimate handled up to five sections, and CAESAR, eight. Each section could be handled independently of the others, except, of course, for a little "drag" on the electron beam... Color was handled with "grayscales" --artist's BenDay screens laid over the art on the Kodalith to give you four "colors" and the invisible background level. The different voltage levels could be set on your camera panel to differentiate between them, and the RGB sliders set to give them each a color. The color tended to "fringe" at the higher levels, since the electron beam painting them had to pass through all of the intermediate levels to get there, but overall, it worked reasonably well, as long as you put your bright colors, like red, on the topmost levels. However, as the components aged, differentiation got lots harder, and the fringing distinctly psychedelic.

So to get clean overlap, clean color, or more than eight sections worth' of animated art, you had to record a second pass keyed over the first.

Red held the CI record on passes on a two-inch quad, somewhere around twenty-two or twenty-six, I think, on a job I did for IBM involving a bus-load of cartoon passengers: we had to see about all of them in one shot, and I couldn't do but two walking figures at a time, at most.

CAESAR characters had their parts laid out separately on the sections, and, yes, it is a trick to break down every character into no more than eight sections. And an even bigger trick to have enough video in those sections to let you bring the figure large in screen. Since you were painting with raster lines, the larger you made them, the further apart the lines were, and the dimmer they became. The smaller they got, the closer-together and thus brighter they became. And when you overlapped them, the voltages would add and make a VERY hot spot on the screen, to the immediate detriment of the high-res tube displaying the shot. Bill Altemus had designed and built a self-matting "overlap circuit" that worked reasonably well. There was some lag in the system that always caused some misregistration, but with a little jiggering with yet another phase control, you could usually minimize it on the most obvious elements.

You could also draw more than one position for a part --for instance, if you had an arm drawn with the elbow straight, you could bend the video with a sawtooth-wave oscillator (just make sure the zero phase point falls at the shoulder joint) to almost a right angle. To bend it further, you'd need to draw the arm bent. Both drawings could go on the same section of Kodalith: by blanking off the unwanted portions you could hide those until needed. Mouth positions would be built in a stack, and the blanking mode set to let you hold the blanking window steady while moving the video behind it for a sort of window-shade effect.

Fortunately, the audience never saw that part.

CAESAR let you animate between key frames. You had eight buttons, each representing one section, and three selector switches to choose the parameters you wanted to change (first time through, ALL parameters changed automatically), and three potentiometers beneath the switches. And a toggle switch for recentering the pots. I wore out several of those.

CAESAR had a display covering nine key frames, but there was a bug in the nine-key-frame program: the first eight were good, and any beyond that up to the ninth were corrupted so badly that if you tried to save any of that work by copying it to another frame, the corruption travelled with it... I quickly learned to limit myself to eight key frames.

You could step through the keyframes on the number pad to "flip" through your animation. (If you're using the Hanna-Barbera mouth system, running through the mouths A-B-C-D-E-F makes the character say "ABOUT" --if you've set them up right.). You could set accelerations. And you could transfer parameters from keyframe to keyframe.

When you used up your eight keyframes --or before, if you liked-- you'd update, and the Honeywell would flicker its display lights and jerk its memory tape reels back and forth until it had solved all the inbetweens. Then you could start in again. To go on, you only needed your last frame of finished animation as your first keyframe. You could also call back any finished frame and use it for another keyframe. And you could go back and animate ONLY one or a few parameters over a stretch you'd already animated...

Because CAESAR wrote alternately to each tape, you had a one-update UNDO capability. I'd tell clients confidentially, "Of course, I never need to..." and usually got a good laugh.

CAESAR allowed you to keep up to six "scenes" --which were basically just a way to let you play back sequences that used different numbers of sections in each without stopping and resetting the machine. There was no logical way to add or delete sections during animation.

As long as you kept the number of sections constant, and used adjacent keyframes to switch between them, you could use the same "scene" for both, even if the sectioning was on different spacing. And since CAESAR had two input cameras, you could switch to the opposite camera for new artwork.

And you could then change artwork on the "off" input camera. I've done jobs with one guy at the monitor to watch the action, and two of us with our heads in the camera cabinets, holding handfuls of Kodalith cels, swapping them out on the watcher's yelled cues, to run through an elaborate scene in one long shot.

We didn't get into building models in the computer until much later. Since everything worthwhile in those machines was based on waveforms, we wound up generating wireframes by waveform. To make a model, you'd build a point-list of xyz coordinates, and an edge-list of points, and merge them into what was actually three waveforms, which you'd add together to get the 3D shape. The prototype went onto CAESAR, and I did its first 3D spot, a drivethru bank booth. CAESAR ran out of multipliers early, so the design went onto System IV, which could handle that much more easily.

The beauty of CAESAR was that it was FAST! Animation setup was easy; not intuitive, unless you'd had some conventional animation experience, and fortunately, I'd had five years' worth. It was so fast that it became possible for the client to direct his own job on the fly, just as they did with Scanimate.

That was the good news. The bad news was that we DID have the clients direct their own jobs.

But the first job was the Coyote/Rabbit story --I was both animator and director (hence the Animating Director title), and could do a quality job, fast. The hardest part of that one was trying to keep my Coyote from looking like Chuck Jones' Wiley E. Coyote...

I did my own art, during the day at first, until we had some Kodaliths to animate with, then at night, so an assistant could do the Kodaliths the following morning and have one ready to roll after the morning preproduction meeting. That set the pattern for all my days at Computer Image. In at nine. Animate all morning Home early afternoon for lunch, drive my wife to St. Anthony's for her 3-11 shift in MICU, back to work, spend all afternoon and most of the evening recording the morning's animation, then draw artwork for tomorrow's job, then pick up the wife on the way home about midnight, and do it all again the following day. Carry artwork home to do on the weekends.

Six months in, with a string of finished jobs, I tallied up, and realized that I'd *thrown away* more animation than I'd done in five years of conventional cel animation!

And I'd kept far, far more.

I stayed there until 1986. Twelve years. I got to take four weeks of vacations...

The two-inch quads had been replaced with two-inch helicals, and those, in turn by one-inch helicals. Jim Duca and Bob Richardson left to form Duca-Richardson, the switcher division was sold, Image West was started, and a great many of the old gang went to L.A. with it, System IV was designed and built, the production division (including me) was sold and moved to different quarters.

When I left, it was:

KDS
8201 East Pacific Place, Suite 502
Denver, Colorado 80203
(303) 750-5000
Charles G. Schleichter, President

The best parts of the job were:

First, the chance to immerse myself in animation, especially character work. Not that it was hard to make objects and logos into characters (as long as you could get the client out of the room while you did it --he/she'd be charmed to death when they saw it finished. They were uniformly terrified if they saw you exaggerating or distorting anything on a keyframe, and never believed you when you tried to tell them it was good, right, and necessary).

Actually, the best part for me was the instant feedback... that made learning inevitable. I could study motion, find out what worked, and theorize why. I began working out my "mod system" of analyzing human motion --sorry, you'll have to wait for the book-- and put it to work.

I figure that if I could've collected everything I did there and put it all on one reel, that I'd've had between three and five lifetimes' worth of output for a traditional Disney cel animator... And if you are fascinated beyond reason with animation and movement, and I still am, God help me, then THAT is the biggest help you could get.

Granted, Sturgeon's Law applied: ninety percent of it was crud --flying logos, sleazoid auto dealers' ideas of showing their superiority over other sleazoid auto dealers, and suchlike, but Sturgeon's law also tells you that ten percent of it was "not-crud" --so with that volume of work, there had to be some real gems.

Particularly with the Navy. They ran their own TV studios onboard ships --usually one small compartment run by a low-ranking enlisted man, there to change tapes. Somebody, looking at the "INSERT COMMERCIAL HERE" cards, came up with the idea of doing their own commercials, and SITE --Shipboard Information, Training, and Entertainment-- got into animation. And, after we got past the stage of telling the troops which branch of the service to reenlist in, things got to be fun.

I particularly enjoyed "Luke and Duke" --originally radio spots, in which two old Western codgers --ex-Navy men, both-- mourned leaving the Navy. The scripts were imaginative, full of jokes, puns, and fun, and the voice talent was aces. I got to design the spots, and set up two "rules" for the visuals: Luke and Duke would never hold the same job two days running. And they didn't own ANYTHING that they hadn't stolen from the Navy --they slept under Navy blankets, they ate from Navy mess trays, their shack sported portholes, anchors, and a brass cannon...

I did the "Old Salt" and his parrot, "Salty" who demonstrated for the troops the evils of overindulgence in food and drink, failure to read contracts before signing, and so forth, and some of those were a hoot.

We even got fan mail. From ships at sea, no less.

We had one problem with the "General Custer" spot --that one had a "Radar O'Reilly" corporal inquiring of General Custer whether the men had all made a will ...DURING the Battle of the Little Big Horn. Not that the spot wasn't good-looking. Heck, no! I had a battle in silhouette behind the two lead characters, with Indians on horseback circling (and closing in on) the surrounded troops, who, also in silhouette, were firing and running about the battlefield, while dust and smoke blew through all of it. And I'd rigged the battle so the cross-cutting to closeups of "Radar" and Custer were seamless --the battle went on behind without gaining, losing, or miscutting a character. No, the trouble came with the last line. A third voice, a "Golden Throat" announcer, said, "Have you made your will? Don't wait. You never know when it will be too late."

The rest of the spot had almost drawn itself (I wish), but this one line didn't fit the action, and I sure couldn't use something as pallid as a camera card. Then inspiration came to me: the line went to outside the perimeter, to an Indian in warpaint, closeup, who delivered the line deadpan... and then smiled a big, evil "gotcha" grin. THAT was the trouble.

The spots went up the Ladder of Death --up through the successive layers of ranking officers, bottom to top, each of whom would look over the spots, decide if there was anything to fear in any of them, nitpick a few just to let us know who was boss, and pass the survivors upstairs where the process would be repeated. That grin was OUT! NOW! So I redid the shot, grinless.

It was the one with the grin that went onto the company demo reel. The grin got the laugh.

Regrettably, almost all of that work is lost. A few pieces are on other people's demo reels. A friend asked me if I'd kept copies of everything; I ruefully told him I was lucky to escape with my life.

One delightful bennie, though, was the occasional meeting with animators I admire.

Lee brought in a quiet guy who spoke English hesitantly. "Dave, this is Bruno Bozetto." My face lit up, "BRUNO BOZETTO??? --THANK YOU!" I said, sticking out my hand and shaking his. He was flustered. (I'm six foot three, bearded, and bearish; he is not nearly as tall, less than half my beam port-to-starboard, and I must have looked like a grinning maniac to him.) "Thank you?," he stammered, "For what?" "For Allegro Non Troppo," I said, and he blinked, and then beamed.

Chuck Jones visited CI twice. Lee showed him the job we were doing --I had a character shaped like those boxing dolls that stand back up when you hit them, making a bouncy entrance, blinking, and smiling as it delivered its lines. The VP Production, looking for a stick with which to beat the animators, asked Jones, "What's wrong with this animation? What can we do to make it better?" Jones shook his head. "Nothing. It has weight, it's three-dimensional. There's nothing you could do to improve it." and I went home that night on a pink cloud.

I had a relatively short time with Chuck Jones --Lee and the gang pretty well monopolized him, and I had the day's work to do, but he got to talk with us each visit for an hour or two for the best lessons on animation I ever got, and, sir, if you're reading this, anytime you want to talk, I'll be happy to keep your glass filled with whatever you're drinking, and sit there and listen.

I'll spare you the horror stories --a lot of things went wrong, including my quixotic decision to stay when my appeals for help resulted in a couple of overworked artists, a couple of shifts worth of VTR people, so I'd have some way to record what I'd done no matter what time it was, a host of Directors, AND a Production Manager, AND a VP Production. If you'd drawn an organizational chart, it would've been diamond-shaped, instead of pyramidal.

We had enough Scanimate drivers that we could run merely killer shifts for them, but, for CAESAR, there was only Red, and me. And Red was usually needed somewhere else. Usually, the first person I met coming off work was me coming back on...

So the Production Division turned into a Byzantine Court, but, being the Willing Horse, I missed that whole show --couldn't turn away from the job long enough to find out that Lucrezia and Lorenzo were conspiring against Niccolo, who, in turn, was plotting to overthrow The Prince, who schemed to undermine Cesare'... ...never mind. It was just a bad soap opera. Office Politics is a game you play, instead of Making Money.

And I should have fled when I was told that we didn't earn enough money in Production to pay for new equipment, or for repairs to the old. That left us with workarounds pyramiding on workarounds to make up for wornout equipment.

I buried two friends, Red Daniels and Glenn Bottje, and a VTR operator who was gone so quickly I never got his name --a nice guy who fell asleep at the wheel after one of our usual long nights, and died in the crash. .

Lee, Ed, Bill, and a few others did make one last college try; after selling the Production Division to KDS. They developed digital compositing equipment that made it a breeze to lay down clean colors and hundreds of key elements in a single frame, and how I wished that I'd had that when I needed it. But it was too late... the competitors had overtaken, and, finally, after all those years, Computer Image declared bankruptcy.

The end came for me during the massacre preceding the collapse of KDS, when Nick Smeloff and Ron Shaw called me in to let me go. They were expecting resentment and hostility, but I've never felt such enormous, welcome relief before or since. I never before dreamed that being fired could ever feel so good. Lucrezia, Lorenzo, et al., were triumphant, but were miffed to find that now the work I used to do was on their plate, and they certainly weren't about to work my old hours...

KDS hung on awhile longer, then folded as well, and everything was auctioned off. The Symbolics went to Mike Ginsburg (who has taken that ball and run with it --Ginsburg Productions is now one of the big guns in Denver), CAESAR and System IV went to D.C. (Roy Weinstock can tell you all about that), CAESAR to storage, and System IV to work. A little while back, Roy brought in the System IV keyboard; it's becoming an object d'art for him. CAESAR may or may not have gone into the dumpster; no matter. After driving it for twelve years, I can let it go.

I still have my CI stock. Lee, in a final act of generosity when KDS bought us, gave the old hands a good-sized piece of the company, always believing it would someday be worth a great deal... Framed, it's quite handsome, and it makes a fitting, if bittersweet memento. Somehow, it feels more like a diploma...

Back to the Scanimate Page


Back to Dave's Worlds