One thing I’ve found surprises a lot of people is that the planetarium produces the majority of the shows shown in the Hansen Digital Dome Theater. Mike mentioned earlier how producing our own shows has led to Clark Planetarium programs being featured in planetariums across the U.S. and around the world.
The process for making a show starts with the brainstorming of ideas for what would be a good topic to explore. Sometimes this can be hard because what we consider exciting and strange, the public doesn’t have an interest in. For example, I find Heliophysics to be fascinating. But when you try and pitch to a crowd the physics of the Sun and Earth interacting, that’s a hard sell. So we spend a lot of time surveying the public as well as other planetariums to see what topics they’re interested in. Based off the news headlines, I think good future topics might be dark matter, the greatest explosions throughout the universe, and of course the retirement of the shuttle.
After the topic is chosen, it’s written into a show script. The script will also include visual guidelines to make sure there’s a sense of continuity to the program. Each animator on staff helps contribute to the visual guidance to really give the show a lot of depth.
Now that we have a visual feel to the show, we start making the “assets”. Assets can be surface textures, models, special plug-ins and video. We also spend a lot of time pestering our education department for information. It may be cool to make a long and vibrant comet tail out by Jupiter, but it just wouldn’t be realistic. We try our best to correct for what Hollywood has been teaching for decades.
About the time we’ve gone as far as we can with assets, we’ll get the soundtrack from our sound engineer. He takes a lot of care selecting musical pieces that will fit the mood of the show and give it an over-all theme. The narration is also carefully weaved into the music to give tempo and clarity to the voice acting.
Here’s where the creative side of our minds get to kick in. Sometimes it may be that a 3D scene needs more detail because of the way the narration plays out. It could be that the fitting all the 3D objects into a scene means breaking them up into separate files because of a quirk in the rendering program. It’s like a puzzle. Every problem has a solution, it just might be a bit harder to find.
The process of making these images, which are 42 time the size of an average tv show, can eat up a lot of computer time. Luckily we just upgraded to a new render farm. This render farm was funded through the largess of the Larry Miller Family and made by the great team at Universal Systems. Each machine can process 8 instructions at the same moment, but this has practical limits. So if one part of the render doesn’t use all 8 instruction paths, the render may take much longer. If you added the total of the potential processing speed inside these machines, you get 1.08 trillion.
Once the parts of the image are computed, they write out into individual images. Storing all the component and final images can expand beyond several terabytes. Needless to say, losing any images hinder the production. We compensate by keeping all data in a RAID [redundant array of independent devices].
We are almost there. All the clips get edited together into a master image set. This final master image set goes through a “slicing” process to pull out the parts that will match to each projector in the theater. These slices are then made into 6 master video files.
Through the few years that full-dome video has been around, the process has been through some growing pains because of the rapid nature of technological development. We decided a long time ago not to ride the bleeding edge of technological developments but wait for the “pro-sumer” level for stability and cost savings.
Wow… so that’s a lot to read. I’d be glad to do a Q&A for my next post if anyone has questions, post them to the comments and I’ll clarify the overly lengthy post.
From the basement…