WHEN WILL IT BE DONE? PT. 7

Read Part 6 here.

Building and designing a plane in midair

I’ve had to figure out how to make the film while making it, with only one chance to get it right. It’s like building and designing a plane while flying it. If you don’t get it right the first time, you crash. That’s the heart of the adventure: there’s only one “take” for every major section of the film, because each takes years to conceive and create. To get a “take two,” we’d have to redo all that effort, which is beyond my, and likely anyone’s, human capacity.

Here’s the concrete version of that statement:

  1. Create the methods, techniques, workflow, and technologies to make the film, while making the film, as none of them already exist.
  2. For each multi-plane animation section of the film, I create a process that allows me to figure out if the photographs can be animated to giant-screen resolution. This process is different for each set of data and each section of the film. Each is only done once and is not reusable.
  3. Only one “take” for each shot in the film. Once the initial resolution check is cleared, work on the source photographs is roughly 90% of the labor, animation is only 10%.
  4. For each of the film’s five sections, work averages about three years for the volunteer teams working on those. We only get one chance to do it. Minor fixes and adjustments can be made, but the what the shot will be is locked before we start the major work.
  5. Because it’s all photographs — changes to angle or speed or especially content are usually not possible. The shot is the shot. One take that will take months or years of work.

It’s worth pointing out again this is terrifying way to create a film.

The big delay

The film is currently more than a couple of yearsbehind where we thought it would be. This is entirely due to the opening section — which includes a three-minute prologue (completed), a fly-through from the Big Bang (not pictured) to the center of the Milky Way (the delayed part), and continuation fly-through from the center of the Milky Way to the surface of the Earth, mostly completed in 2014, except for the background of the universe — the delayed section.

The delayed section is the classic “flying through the universe” shot we have seen in nearly every deep-space movie and TV show, from Star Trek to Cosmos. These are always created from CGI or other visual effects. By 2012, I had an idea for a solution for doing it using only real photographs.

Here’s a typical CGI example:

Astronomers on the Sloan Digital Sky Survey (SDSS) have imaged over one third of the 360-degree sky that surrounds Earth in all directions. One third does not sound like much, but in a 360-degree spherical view, it’s actually a massive area.

More importantly, they had performed spectral analyses on most of the galaxies in this area and assigned a position and distance, the critical elements for scientifically accurate multi-plane animation.

A scientist. Miguel Aragon of Johns Hopkins University, did a simulation using images of most common galaxy types, SDSS position data image, 3D CGI – around 400,000 using images from around 40 galaxies types vs. our 5,000,000+, each one unique image. But it proved it was theoretically possible and potentially stunning as this is a very impressive clip. He also kept it to HD instead our 6K and 8K to make renders manageable.

In January 2013, our first volunteer team, having just finished the code to process all of Cassini and Solar Dynamic Observatory’s imagery, was ready to start on SDSS. They estimated three to six months to take photographs from the SDSS and turn them into elements that could be animated. I doubled that to six to 12 months. Knowing what else remained, I announced the July 2014 date.

However, by fall 2013 that team had run into problems unrelated to the film. I was in the middle of a Kickstarter campaign for a symphonic recording of “Adagio for Strings.” The team leader assured me he hoped all would be resolved and everything would be complete by late spring 2014. I pushed the public release date to August 2014.

As soon as the recording was complete in January 2014, I realized that the team leader was going to have to leave the film. And that no usable work really existed. So February and March 2014, we put out a global call for help and got a new volunteer team up and running within weeks. By April 2014 the team had a plan and work was well underway. The new team leader, Bill Eberly, along with the leader of the high-resolution deep-space image team, Jason Harwell, have turned out to be the most important volunteers on the film.

The decision

In retrospect, we should have not been surprised this has proved to be so hard. After all, the understatement of all time is “It’s a big universe.” The universe is the very definition of huge, massive, unending, and uncountable — and whatever other terms you can find in a thesaurus.

Very quickly the new teams and I realized that the three- to six-month timeline was a fantasy, and that years of work lay ahead. We had a frank and open discussion with all the key volunteers and backers of the film. The choices were:

  1. Replace the opening section with something else. Given the process of the film, a fly-through of space would have to be CGI. Other types of footage — interviews, stock footage, or anything else — would not fit. This options easy to reject.
  2. Drop the opening section — the first act — of the film. We would have a 25-minute film that started without a beginning. It would also not be a standard length for release in giant-screen theaters, massively reducing its potential audience.
  3. Finish the film as planned, but taking much longer. How much longer, none of us really could say.

We all voted unanimously and enthusiastically (well, I was probably the least enthusiastic) for #3. And we’ve never looked back. I’m sure from the outside it has not looked like a wise or prudent decision. After all, it gives the appearance that the film is trouble, that we are trouble as filmmakers, and/or that the project is at risk.

Plus the distributor and film buyers want firm dates, predictable schedules for easy-to-understand content they can book and screen. Completely understandable, but at the end of the day, it’s left to filmmakers to give up years of their lives, so the decisions of what’s best for a film are ultimately and rightly ours.

Part 8 is here!

No Comments

Post A Comment