Archive for November 2008

Revised completion estimate and the release of the Book of DOG

Brooklyn, New York

Today is the day of reckoning, where I reckon just how much trouble I am really in. 2008 is coming to an end and Dan Kelly is committed to fully engage all control surfaces until the mission is complete – stopping only for food, sleep and exercise. So where are we, what does the status display reveal? How far are we from total completion?

Short version – we’re not there yet. What to do?

Aside from living like a fanatical monk, my primary strategy for finishing has been slowing down time, essentially a collaborative technique. I could expand the collaboration and get some actual help, either by kidnapping alternate versions of myself from adjacent universes or recruiting other artists in this one. Kidnapping myself requires a deeper fluency with quantum physics than I current enjoy, alas. Since I don’t know many artists who are as geeky as me, I can’t delegate the tricky roto and motion graphics. There is also plenty of straight up illustration and graphic design – signs, grafitti and other visual details – needed to flesh out the DOG universe. I have beaucoup friends who can do that sort of thing in their sleep.

To get artists started I’ve got to compile a comprehensive inventory of the required visual elements and, with or without qp fluency, open a gate between universes.

… but the apocalypse didn’t happen all at once. After the initial disasters and before national governments totally collapsed, supplies were air dropped to survivors in shrink wrapped cardboard boxes – lots and lots of boxes…

That’s a sample of backstory and it’s essential that collaborating artists have it.

If a visual element is going to work, it has to fit into DOG’s universe. Real people have histories and so do objects. Designing visual elements for props requires an artist to travel to the universe that the prop supposedly belongs to, with all the assumptions and circumstances of that universe. They must be an artist in that universe, do the work there. Who is the client, what is the job, what am i having for lunch, how much am i being paid, what’s my life like? Artist/actors, how’s that for polymathic?

To help the artists make the journey, I have to give them history and context.

Enter the Book of DOG, a Design Reference, (DR for short). The DR describes what’s possible in the DOG universe and provides a history of the characters and their environment (backstory). More controversially, the DR could reveal the project’s premise (what it’s supposed to accomplish) and it’s general feel (what categories of experience and emotion are intended to be conveyed).

The premise and feel are controversial in the sense that defining an objective for art might be oxymoronic and intending specific emotions can be crass, but… when controversy comes between an artist and her bliss, she should not retreat. So with awareness of the paradoxes in play, onward.

Here is an example of a artifact or prop – a label for the disaster relief boxes.
Melonie, Eiji and I didn’t have time to make labels for the boxes before traveling to Canada in 2006, we just crammed boxes into every nook and cranny on the set. “I’ll handle it in post!”
Thanks to Mocha, I’ve been able to track the movement of the boxes in this panning shot and map the labels to them. The lower boxes on the right stack show what they looked like ‘before’. Most of the other shots with boxes are static, so theoretically this is the toughest track. Notice too there is yet to be a reflection of the topmost box in the shiny wood of the closet door. Details, details.

So far I’ve asked 9 of my favorite artists to review a list of design challenges and choose whatever interests them. They’ll start with conceptual sketches and with a little luck execute their ideas before years end.

Everyone who has contributed significantly to DOG has been paid, I won’t ask talented people to work for free. All contributors are compensated – tho it might be deferred and somewhat below their usual rate. I won’t pretend that it’s not really all about excellent friends helping me wrap this sucker! Plus it’s always impressive to have lots of names on the ending credits.

Now, let’s have a look at the current completion estimate. My August 15 estimate projected 300 (297) hours of motion graphics or 25 days at 12 hours per day. It’s almost November 30th or 105 days since I posted the list and I’ve accomplished about 30% of the tasks. If 30% took 105 days then the remaining 70% will take an additional 245 days or 8… more… months.

Something about this doesn’t make sense to me, but numbers don’t lie. We knew the August numbers were highly suspect but to be off by a factor of 4? Certainly there’s was a learning curve to climb what with AE and Mocha, but that’s not over yet.

What’s the plan then Dan? Well, that’s for the next update. I’ll probably evaluate whether there’s help to be had, if there is a faster approach than I’ve taken so far, what changes are the most important, and how much longer I can manipulate credit cards before I’ve got to go back to work.

Backatcha in December,

dk

Are we there yet? Roto upgrade, tracking, abandoned highway and the Obamas

Brooklyn, New York

Hey there! Three weeks just flew by, phew.

Since we last communed, I spent a week trying to accelerate my rotoscoping execution with the help of Pete O’Connell. His method is to track the elements and then build splines around the track. In lay terms, rather than trying to create an animated outline around Christina’s shoulder at critical points of movement such as change of direction and acceleration/deceleration, O’Connell’s method uses AE’s point tracker to roughly map the position of her shoulder automatically. Automation both speeds up the process and generates a more consistent result. Sounds great right? The long list of caveats starts with AE’s point tracker, it’s persnickety with grain or low contrast. Anyway, I spent a week getting up to speed on the method.

Being able to eliminate or replace elements of a scene has certainly facilitated DOG repair. It’s also expanded my imagination, what I’m willing to dream about in the future. I can jump and make nets appear, but it’s a bit cheeky to do it three times every second. So, in order not to blow my cover, I spend some time actually weaving nets and making sure they are securely fastened to bulletproof anchors. Translation – knowing how to do something often leads to new ideas based on that knowledge. This contrasts with ideas that require completely unprecedented categories of knowledge for their realization. Going ahead with a project not knowing exactly how to do it is called making the net appear. That’s the definition of DOG.

Speeding up the roto work is super crucial as there is still plenty left. How much is left, how soon will you be finished Dan Kelly? Can you believe some folks still ask me that question? Not everyone gets to read these updates. Caught off guard in meat space, I find myself resisting the urge to bitch slap the uninitiated. My 8/15 estimate for total, absolute, dang near completion was 11/15, 4 days from now. By next week’s update y’all be privy to the RCE or revised completion estimate.

I shifted gears in November and returned to the opening montage, which will include the cruise ship and car models developed by Jonathan. The post production team is pretty skeletal at present, just him and me. As I concentrate on the shots that feature his models he continues to develop and refine his modeling and rendering. During my September visit to Michigan, Patrick, Jonathan and I stood in my driveway and talked about how the reflections on the Honda Odyssey defined it’s geometry. The car models might be more convincing if the reflections from the existing environment were amped up, so Jonathan built tree models and moved them much closer to the cars than they actually appear in the scene.

The highway scene has evolved a lot. Originally it was just a empty highway with cloud shadows racing down it. As mentioned in previous posts, a call was made to flesh out the back story, to give it more umph. I imagined 4 lanes of dead cars stretching into the distance – perhaps an automated highway after it’s controlling infrastructure crashed or car brains scrambled by the devastating effect of an EMP type weapon. After the cars were added however it read more like a traffic jam than a post apoc enigma. It needed something to show time passage, to illustrate abandonment. Snow! Snow also suggests climate change and unlikely weather. Several birds with one stone.

Here’s a rough composition of the highway shot. I tossed the Pasadena sign in as a placeholder for someplace warm.
Miami might be better and more recognizable as a warm place. Plus there are lots of pine trees in Florida, so that works.

The highway shot is slightly shaky because I was shooting on top of the Honda with a tripod and my weight on the roof made the sheet metal unstable. I actually like the shake, but it has to be tracked so that the cars and snow have the same movement. It turns out that this is a very difficult shot to track. AE can’t handle it at all, and even the uber tracking software Mocha has to be carefully prepped.

I’ve been pondering how to farm out more of the finishing work. I know lots of artists and there is a variety of supporting imagery that I could use help on – posters, graffiti, logos. I’ve been thinking about extracting these tasks from the master list and assigning them to available collaborators. Deferred or no pay in most cases but credit certainly, lots and lots of credit!

On November 3 at 9:00 pm I tossed the cats in the van, drove 13 hours to Michigan and added my queer shoulder to the wheel.
Because of my valiant dedication to democracy, these people will be moving into the WH. Cute aren’t they?