Interesting little surprises

I run into interesting people online – hell, my hubby is a person I met on the Megatokyo Story Discussions forum and Aff is my former clan leader now housemate. This is probably why most of my friends are online too.

Naturally, interesting stories come up and this is one of them. A discussion about Star Wars over at According to Hoyt lead to this delightful anecdote by Xenophon (who kindly allowed me to repost his story here.) Enjoy!

The first Star Wars film (episode IV) was a really interesting phenomenon in H’wood, not least because NOBODY (including the studio, director, producer, etc.) expected it to be a huge hit.

When SW was in production I was in High School living in Santa Monica, CA. Across the street from us lived a budding young producer who’d heard about SW through the grape vine. The buzz at the time was “It’s an old-school SciFi adventure film, with a twist. The big difference is that the milieu is lived in: beat-up, run down, dirty, and USED.”

Our neighbor **really** wanted to visit the special effects team working on the film. He figured out that the sfx team would really want to meet “the father of computer graphics” (a.k.a., my Dad). So he got my Dad’s permission to parlay Dad’s reputation into a visit for himself, my Dad, and me. Dad’s other “payment” for the use of his reputation was that we got our neighbor’s tickets to see SW at the theater belonging to the Academy of Motion Picture Arts and Sciences. No lines to wait in, and the best projection and sound systems in the industry at the time. At the time we made this deal, nobody expected SW to be a huge hit, so we had no idea what a big deal those tickets would be after the movie came out.

According to our neighbor, the reason he was so eager to visit the sfx team was that he’d heard that they were producing amazing results on a budget that he described as “half a shoe-string.” He really wanted to learn about how they were doing it.

We spent most of a school-day visiting the sfx guys. They too described the “big difference” from what had been done before as being the “run-down and dirty” look. Absolutely NO ONE claimed that anything about the *story* was special or different.

There was virtually NO CGI, as that was far too expensive for their budget. Instead, it was almost all old-school model building, etc. They did only one thing that they described as “high tech.” The fly-through of the surface of the Death Star (and the trench) was done by putting a camera on a computer-controlled arm, and moving it (extremely carefully!) past and through a BIG model of the DS surface and trench, doing stop-motion photograpy as they went. In this context, “BIG” means about 15 feet by 40 feet or so. The arm was an industrial robot arm they’d leased for the purpose, controlled by a PDP-11 (also leased). Apparently they were the first in H’wood to use this technique. The precision and flexibility of the computerized arm let them mount a video camera so they could preview the results and tweak the camera motions to the director’s satisfaction — all before shooting even a single frame of actual film. It also meant that late changes to the exact path of the camera required shooting *only* the specific frames that changed. Prior techniques required re-shooting the entire sequence.

They told us that the model-work for the DS surface was put together out of square modules about one foot on a side, fastened side-by-side on a grid. They had remarkably few different modules — four or five, IIRC (certainly a single-digit number) — and disguised this in part by rotating the modules so they’d be in different orientations to the camera. And yes, the sides of the trench were exactly the same modules, placed vertically. If you look closely, you can see that the trench is exactly one module deep and one module wide. This modular structure mattered because they needed nearly 1000 modules for the DS surface; about 75% in use at any time, the rest as replacement parts when they broke something.

We also saw some very amusing saved video camera footage in which mistakes in programming the arm caused the camera to run into gun towers. Or the wall of the trench. Or the floor of the trench. Including a few where they said things like (paraphrasing here) “and the point where we lose the video here is where the robot arm destroyed the video camera by shoving it through the 2×6 framing that holds up the bottom corner of the trench.” Oops! Apparently this represented another significant savings: destroying consumer-grade video cameras was WAAAY cheaper than destroying high-end film cameras and lenses.


Leave a Reply