Over the past decade the RUEbot evolved from design, construction and programming into a motion control system rivaling most. Initially, the camera system offered dolly, crane, tilt and pan movements for my puppet animation THETA. For SHEOL, I enhanced the RUEbot significantly. I added the axes of roll, focus, and what I call spin. Also, the lighting system is motorized to offer a shifting cloud effect as well as to guide the leafy growth of living ivy. To coordinate these complex movements required 7 years of trials and tribulations. Ultimately, the RUEbot performs as a gyroscope allowing the camera to move to any point within the circumference of 6 feet. The machine stands 9 feet high with a precision of 1/10000 of an inch on vertical and horizontal action. All rotational axes as well as the camera body rotate a full 360 degrees varying up to 1/1000 of a degree accuracy This equates to the ability of the camera to move throughout real space with an image plane as small as an inch wide.
I wanted a system that mirrors the sensation of animating in real life as opposed to computers. The process to integrate these two modes proved more difficult than expected. Each method approaches motion quite differently. As an example, things like gravity got in the way; puppets embrace gravity whereas computers ignore it. It is more difficult to make something appear light in weight with puppets and equally as difficult to make something appear heavy in the computer. Each actualizes in the opposite way, almost to the extent where I can predict the method to achieve something in the computer by imagining the reciprocal approach in real life. Eventually I found a connection however, in SHEOL, I animate with light.
The main objective was to develop a system to speed the process of computer animation and still produce a richly textured image. It requires less time to integrate the animation into a digitally captured background than to have every element of the scene created in computer space. This premise proves true, however, to comply with a motion control system is another matter.
Rendering time is greatly reduced, as only the characters need be generated. Shadow objects that play on the background are very basic shapes. More intricate, complex gestures and nuances are possible due to the low overhead on computer processing. Essentially, more time and energy is spent on animation rather than on calculation.
First, I shoot the tabletop animation digitizing each frame sequence directly into the computer. In the case of SHEOL, the shots of the plant-life photographed in time-lapse required several days and sometimes weeks to grow along with the camera and lighting direction.
Next, I place the computer-animated characters in reference to this background and make adjustments. Because the camera movement and character placement derive from the same point-of-view inside the 3D animation program, multiple passes combine perfectly in sync to one another.
Finally, the images are composited and enhanced to meld seamlessly together. With lighting and blending techniques available from most basic tools offered by photographic software this task also minimizes computer rendering time. There are countless anomalies associated with computer animation which consumed a great deal of time and effort to lock. I am proud to say this system is now in place and operational I call the RUEbot.