My nerd crush on John Dykstra

If you were to watch 2001: A Space Oydessy (aka 2001, which came out in 1968) and Star Wars: A New Hope (aka ANH, which came out in 1977), you would notice a pretty significant difference in the visual effects despite both films winning the Oscar in that category (albeit 9 years apart). 2001 utilized cutting edge blue screen techniques that layered a variety of futuristic scenery on top of each other, but mostly with camera positions that didn't move. As anybody knows who has watched the Death Start trench scene of ANH can tell you that the shots not only move, but move quickly and achieve a high level of excitement in the process.

So, what's the main difference in the visual effects between the two films? John Dkystra.

Early in the production of ANH, George Lucas knew he was going to need a technological breakthrough in order to make the kind of movie he wanted to make. To accomplish this, he established his own special effects company called Industrial Light and Magic. One of the original employees of ILM was John Dykstra.

As chronicled in Empire Building, Dykstra and his team spent the months that Lucas was location filming in Tunisia back in California working on what would become the Dykstraflex motion control camera. Although Lucas was flabbergasted that no shots had been completed when he returned from Africa, the system that Dykstra's group created would change movie making for decades until the arrival of computer generated imagery.

To fully appreciate Dykstra's achievement, you have to understand a bit about how movies used to be assembled. While I'm sure some film student will poke holes in my explanation, my poor man's version follows here.

Most visual effects in movies are assembled from different shots and combined using a process called compositing. Pre-computers, compositing had to be done using special machines that would run different film reels together and record the result. Since film is inherently a transparent medium, this takes some careful planning.

Suppose you want a shot of a space ship moving across the screen with a star field in the background. Such a shot would be comprised of at least three separate pieces of film, or "plates". The background plate would be the star field. The foreground plate would be the space ship, typically a photographed model since space ships don't exist in real life the way they do in movies.

Intuition tells you that you simply combine the foreground with the background and you have your shot, but it doesn't exactly work that way. Think about what would happen if you had a single light source projecting the background and foreground plates together. What you'd get is the two images overlapping each other and the stars would appear to be on top of the space ship.

In order to remove the stars from the space ship, you need a third plate which blacks out the stars in the exact shape of the spaceship. There are a variety of techniques that can create that third plate, but before Dykstra's work, it was extremely difficult to do unless the camera position was fixed. If you could move the camera, you could make the model space ship look like it was flying in a much more realistic way. This was Lucas' requirement and Dykstra met it.

How? By taking an old camera and attaching electric motors to each basic function (pans, tilts, zooms, etc.) along with a series of integrated circuits that could remember the motions. The camera could be programmed to make a set of movements once and then exactly duplicate that set of movements as many times as needed. That meant Dykstra could program the camera, shoot the foreground shot of the model, change the lighting conditions, and then shoot the exact same movements for the blacking out plate.

The resulting composite was much more fluid and realistic than had been possible before. Although the system is largely obsoleted by computer generated images used in today's movies, it was a huge leap in capability at the time. Dykstra left ILM before work on The Empire Strikes Back began, but he continued to work on special effects having most recently been the Special Effects Supervisor for Spiderman 2.

 Was that interesting or helpful? Consider subscribing:  by reader  or   by Email

My nerd crush on Paul Allen

The second guy is always more interesting. At least, that's what it seems like when two people get together to form a company. Typically, there's a business guy and a tech guy and maybe it's just my inner nerd crying out, but the tech guy is a lot more fun to study. Bill Hewlett (HP), Steve Wozniak (Apple), Marc Andreessen (Netscape).

Maybe the argument breaks down when looking at Paul Allen, but I'm pretty sure he's never been on the cover of Time with Bono so at the very least, he's the lesser known of the two Microsoft founders. He's more interesting to me than Gates because of what he did at the beginning of the company he founded and what he did after he ended his involvement with it.

In the mid 1970s minicomputers, as they were then called to describe just about anything that didn't take up a whole room, were largely the stuff of hobbyists. There were lots of smaller players in the market, but Altair began to surface as a leader in the fledgling industry around 1975. At that time, Gates and Allen were students together at Harvard and decided to try to found a company by making a BASIC interpreter so that people could more easily program their minicomputers and avoid having to learn assembly language. It was a great opportunity.

However, if you are going to write software for a computer system, it's generally a good idea to have one of your own to see if it works. Gates and Allen didn't. Instead, they wrote their interpreter by hand and on punch cards not knowing if it would work. And they did it in 8 weeks.

So Paul Allen gets on a plane (according to legend, he was picked because he was old enough to rent a car and Gates wasn't) to Albuquerque, where Altair was based, to give a demo of their untested product. Imagine what that would be like for a moment. You carry with you a stack of punch cards that can instantly take you from being a bored college student to running your own business in a new industry. You have no idea if your little programming language interpreter will work because you didn't have a computer to run it on and even if it does work, should you drop the stack of cards and get them out of order, you'll never know. How exciting and nerve racking must that have been?

Of course, it all turned out in the end. Allen didn't drop the cards, the BASIC interpreter did work when he ran the cards through a reader and into the Altair's memory, and Microsoft became the official maker of Altair BASIC. An empire was born.

If I were ever a multi-billionaire, the other reason I like Paul Allen is because I'd like to imagine I'd do the same kinds of things he's done. In 1986 (long before Gates did something similar), he established a charitable foundation that doles out $30 million a year. In 1988, he purchased the Portland Trailblazers and in 1997, he bought the Seattle Seahawks. Chairity, NBA franchise, NFL franchise. That would pretty much be my list of "to do's" too.

 Was that interesting or helpful? Consider subscribing:  by reader  or   by Email

My nerd crush on Ed Morgan and Karl Bacon

Almost nobody has heard of Ed Morgan and Karl Bacon but almost everybody has benefited from their engineering breakthrough that came in the late 1950s. All they did with their company Arrow Dynamics, was invent the steel tubular roller coaster track on which most modern designs are based. They revolutionized their industry and it all came about in an attempt to solve a problem that no roller coaster manufacturer had been asked to solve before.

You could argue that roller coasters have been around as long as trains. The earliest rides were forms of transportation that would use steam engines to go up a hill and then let gravity do its thing to get back down. People enjoyed them enough that companies sprouted to build stand alone rides that had their own lift hills. By the early 20th century, roller coasters were hugely popular with more rides than there even are today.

Before the 1950s, designs borrowed technology from the locomotive industry from which they spawned. Wooden or metal support systems gave elevations and turns to flat steel railings that the metal wheels of the car rolled upon. Because of the flatness of the rails, the angles they could be bent at were a limiting factor in ride design. For example, only a certain tightness was achievable in a curve because the flat metal rails could only be bent so far without becoming weak and unsafe.

Then, Walt Disney went to Switzerland.

While there, he was struck by the beauty of the Matterhorn peak in the Alps and wanted a replica of it for his new theme park in Anaheim, California. Not just a piece of eye candy, Walt's Matterhorn would have a roller coaster in it themed to feel like a bobsled ride that would take guests inside and out his miniature mountain. The finished attraction would be a 1/100th scale reproduction of the real thing, a size that was required in order to fit the perspectives of its surroundings within Disneyland.

Size restraints presented Morgan and Bacon with a difficult engineering challenge. The ride had a relatively small show building that was to be shaped to look like a real mountain and it had to house two different tracks to give its riders the sense they were racing with another sled. The confines of the building proved to be the limiting factor as it would require much tighter turns than were possible using standard roller coaster rails.

Thus the steel tubular roller coaster track was born to solve the problem.

As Morgan put it in Robert R. Reynolds' great book called Roller Coasters, Flumes and Flying Saucers:

"Steel tube rail lends itself much more to accurate bending because the forces are always the same no matter what axis you bend on. You can't do that with a flat bar, or angle iron, or anything like that."

Due to its cylindrical nature, a tube can be bent at a much tighter angle than a flat rail can. Combined with innovations in polyurethane wheels, this design allowed Arrow Dynamics much more freedom with the track layout while still meeting Disney's vast set of requirements. When it opened in 1959, Disneyland's Matterhorn Bobsleds became the worlds first tubular steel roller coaster. It underwent a major renovation in 1978 that validated the flexibility of its design and continues to operate today, almost 50 years after its initial construction.

 Was that interesting or helpful? Consider subscribing:  by reader  or   by Email

My nerd crush on Marc Andreessen

If you have a technology job, especially one related to web development, and you don't know who Marc Andreessen is, you should be shot.

OK, maybe not literally.

Tim Berners-Lee may have invented the web, but Marc and Eric Bina made it easy to use when they created the first widely used Internet browser, Mosaic. Oh how I miss that funky S with a globe in the middle of it running on the HPUX version of X-windows.

Marc happens to be a year younger than I am and I was in my first year at HP when he changed the world as we know it. Jealous? Absolutely. But I had to admire what he did and the way he did it. Before we used terms like "iterative programming models", those two guys lived it and made Mosaic better by the day as they aggressively solicited feedback from anybody who gave it to them. Because I started in the wave of web development that followed his early successes, I sort of celebrated his victories vicariously over the years. He did good work and got rich off it, good for him.

This month, Marc launched a blog:

http://blog.pmarca.com/

It's about a lot of different things, but the most fascinating posts are about his experiences with entrepreneurship and start ups. If you ever had a passing thought at having your own business, you should read his post from yesterday:

The Pmarca Guide to Startups, part 1: Why not do a startup

It discusses the ups and downs of starting a new endeavor in a unique way by a guy who has been around the block a few times over the last 14 years. Take a look at what he has to say and follow this current series of his. I'm sure everybody with a business can learn something from him on this topic.

Part 2 of Running Diary: Preparing for my performance evaluation will run later in the week.

 Was that interesting or helpful? Consider subscribing:  by reader  or   by Email