Theory vs. practice: more birds!

For the past few weeks I've been conceptualizing and prototyping the visual effects for the play. The projections are taking place on the ceiling. This is one of the most difficult aspects of creating the animations. Since the projection is on the ceiling, not on the wall, there are some unique problems involved. My thought was that I'd try to convert all of the animations into ones that would work specifically within this context. For example, the director wanted an animation of falling rain that then turned into falling bombs. I was afraid that an up-and-down orientation (drops falling from top of frame to bottom) would be confusing for both the character's orientation on stage and for the audience. So I changed the animation to a ceiling orientation. In my version, rain drops ripple outwards and then turn into bomb blasts. It's as if we're viewing everything from a top (or bottom) view. It doesn't matter where you're seated. An additional benefit of this in my thinking is that it made for better sound design possibilities. You can more easily sync a rain drop sound to ripples and an explosion sound to an explosion. Otherwise, you're left putting in washy rain and bomb sounds with no specific audio-video connection.

Well, the director didn't go for all that. He really wanted everything happening up-and-down, vertically. He also didn't like my take on the birds-into-planes sequence. He really wanted it as he specified... with the birds flying from the bottom of the frame up through the top of the frame instead of sort of stationary as I proposed.

So I spent most of yesterday and today trying to figure out how to make the animations work. The first thing was to understand the nature of the design problem which has to do with putting theory into practice. On Noise film, if there was anything that didn't work in prototype form, it didn't make it into the show; the shot got redesigned. That's what took so long to shoot that project.

But the animations I'm designing for this project are created theory-first. This approach approximates Hollywood feature-film practice. I was reading an interesting article about Robert Rodriguez. One of the reasons he serves as fx supervisor on his films is it enables him, as director, to change a shot if its fx cost is too expensive or unwieldy. in other words, there are often times when you can make a minor modification to a shot that makes it less costly to produce but doesn't affect the story. But in traditional Hollywood terms, an fx supervisor can't tell the director what to do. The hierarchy doesn't permit it.

This case is similar so I have to make the shot work as described. What the director's envisioning is something like the image at the top of this post. It's as if we're looking overhead seeing a bird flying. It then transitions into an airplane. The plane is about the same size as the bird because it's higher in the sky. It also has the same apparent speed because although it is traveling faster, it is so much farther away. The sequence needs to take place within 20 seconds.

The whole thing is much more complicated than it seems. First, there's the speed. If the bird flies too fast, it won't register and won't fill the 20 second timing. To make the bird fly more slowly, it has to be positioned higher in the sky. The problem now is that the bird is small since it is positioned farther away from the camera. The speed is right but the bird is hard to see! The other difficulty is one of convention. As some of my early tests indicate, it looks unusual to see a bird flying above us over a stationary camera. We see them fly and then our camera moves to follow them. They swerve or they swarm. But you don't often see birds flying in a relatively straight line overhead at a speed and size necessary for them to recognizably change to a different shape. This image is more of a conceptual, poetic image. It is an image that exists primarily in the mind.

One of the first things I did to make the shot work better is put some particle clouds in the shot moving against the bird. This helps because it creates a rationale for slowing down the bird: the bird is flying into the wind. Therefore I'm able to make the bird bigger. The other thing I did was slow down the clip. Rather than actually add frames to the cycle, I time stretched it. The idea was to create something that looks like a bird flying in slow motion. (I'm using the great flocking birds tutorial for AE at Creative Cow BTW.)

The other concern is one of approach. I did one version of the animation that was relatively photorealistic. But it just didn't fit in with the other animations which are more stylized. So now I'm trying a version with very stylized birds, almost like bird icons. This also helps in that it fits with the conceptual nature of the image since what we're seeing on screen is visual language instantiated within space.

What makes the whole thing really interesting to me is thinking about how certain approaches affect the visual appearance of things. It is almost as if what makes a Hollywood film look like a Hollywood film is the way its images stem from writing. From script to director and then to fx supervisor, the image follows a trajectory in which the idea precedes practice. This is what gives many fx films their certain look. The Star Wars prequels, for example, all have that theory-in-practice look although I'm not sure exactly what characterizes this look! The giant battle of the titans scene in Matrx 3 also has that appearance. It has something to do with the way with the physics of that scene works: two giant forces meeting in battle but propelled by flight. These are scenes in which ideas are visualized, theory turned into practice.