|
|
Throughout history the goal of realism acts as a landmark achievement for new visual mediums. Without a doubt, computer animation opens new possibilities to this aspiration. The ambition holds an ultimate prowess until the achievement surpasses the motive of the image itself. Once this occurs the attitude of the image returns to the vanguard. Although the aesthetics of authenticity always persists, this push towards reality interests me the most. Not so much the success of it in so much as the desire to accomplish it.
In the process of creating the RUEbot the challenge to have the machine align with computer precision evolved into my goal, further, my obsession. Concurrently, computer software evolved to dismiss the necessity for such a device to meet this perfection. So why pursue this endeavor? My only answer is to achieve it.
Camera movement in the computer is limitless. Reality however provides a few concrete obstacles. It is these deterrents that inspire progress, I believe. To understand the flawed and clumsy movement of reality becomes a motivating force for computer animation to meet according to many. Alternately, to correct these errors grew into the reality of the RUEbot. Software includes many formulas to persist the illusion of realism; however, the code does not offer the ability to command motion control. This connection between the two I developed myself.
|
|
In order to link the computer animation software to the motion control system, I programmed the RUEbot through the Visual C++ language. A Graphic User Interface controlling velocity, ramping, doping, repetition, and duration of each camera move and also the networking of several motors eventually developed into RUENET. The intelligence allows further axes to be easily added for future projects. As well, any 3D software can work with my system so long as it offers the ability to output and input coordinate data interactively, which is quite common now. The data can easily be transformed into my interface for the many available animation platforms.
To generate a camera move using a tabletop method I move to key-frame positions using a joystick. The camera position and the target position for each frame are measured and converted into coordinates placing the virtual camera in the computer. The data itself can be converted into many formats, using any coordinate system. Conversely, a camera move generated from the animation software controls the RUEbot directly. Inverse formulas calculate positioning from virtual space into real space. The formula for this perfect balance took a solid year of programming and calibration to meet the accuracy required.
One of the more interesting though complicated aspects involved the formula for focus and field-of-view. Designing a graphic controlled equation to convert the camera-to-subject distance against the rotation of the lens took several months. The program inversely compensates for the shift in field-of-view affected by focus changes in the computer animation. As well, the system allows different lenses to be calibrated.
Another more daunting aspect was the concept of level.
As the machine does not sit perfectly flat, a formula devised to generate, collect,
and analyze a series of control points corrects the changing skew, pitch and
yaw of the stand. So many anomalies surfaced in the pursuit of correcting the RUEbot to meet with the exactness of the computer animation. The means now correct distinguishes a certain possibility of virtual understanding of organic motion. My efforts demonstrate in the fantasy created within SHEOL and
RUEbot™.
