In June of 1975, I was contacted by George Lucas and Gary Kurtz with regard to my supervising the photographic special effects for Star Wars These first meetings with George and Gary outlined effects scenes that involved spacecraft engaged in acrobatics that any stunt pilot would be proud of: three or four ships performing rolls or loops while firing lasers at each other in the classic "dog fight" tradition. All this was to occur while being viewed from a camera platform that itself needed the fluidity and freedom of motion of a camera plane. This visual concept was a far cry from the locked-off camera approach to spacecraft miniature photography seen in the space classics of the past. This was a challenge, to say the least.

As the meetings and story breakdown continued, it became clear that this film would not showcase 20 or 30 special photographic effects shots, but would use spaceships, miniatures and all manner of photographic effects, as you would use automobiles in a film of contemporary time setting. In the entire film there are some 365 miniature and photographic effects shots. The challenge, therefore, became a task of mammoth proportions.

In order to produce the quantity and quality of special photographic effects shots called for in Star Wars, a complete in-house system would have to be developed. This system would include miniature design and construction facilities; the design and fabrication of a camera motion control system; electronics and mechanical facets, and a complete optical house and animation department. I felt that the in-house system would be the only way that consistency of quality and control over each of the separate operations could be maintained.

First I sought out the personnel that I felt were necessary to carry out the special requirements of this project. This included conceptualization, design, and execution of the visuals and of the equipment used to produce those visuals. The electronics portion of the system would be handled by Al Miller Electronics. Al Miller and I had worked together in the past and already had a basic design for the electronics/camera system worked out, and now we had an application for this system. For the mechanical aspects of this project, I contacted Don Trumbull, Richard Alexander and Bill Shourt. All three had talents in design and manufacture of one or two off-types of machines this project would require. In order to make certain that the people who would have to use the equipment had a hand in its design, I brought in Richard Edlund to be the Director of Photography. Grant McCune was given the responsibility of producing our miniatures. A production shop was set up by Bob Shepard, and all scheduling and co-ordination of special effects shots was done by George Mather.

In eight months we brought the facility in Van Nuys, California, from an empty warehouse to an incredibly versatile system. The talent and integrity of these people, plus our ability to communicate with each other, provided the key to the success of the system.

As the equipment began to evolve, the basic in-house departments began to form. Robbie Blalack (Praxis) began setting up an optical department. Adam Beckett began designing lasers and other animation which would be used in Star Wars. Mary Lind set up a film control room to handle each shot, as it traveled from camera on stage through rotoscope, and optical.

Because the photographic effects were to be done in the United States, and the live-action filming was to be done in England, some rather severe communication problems had to be overcome. George Lucas, Joe Johnston and myself described each shot in one or more storyboards, and its requirements were established, right to the frame count. Most of this information came from a cut Battle Sequence, made up of excerpts from war movies. This established the size and speed of the fighters and their positions in frame. With this first set of storyboards in hand, we set about finalizing our miniatures, and photographic systems.

The format of the system was first to be considered. Star Wars was to be released in the wide (2.35-to-1) format. The problems presented by trying to produce the incredibly complex photographic effects in this anamorphic format were staggering. There were problems of depth of field and image quality after numerous optical steps, hyperfocal distance of the anamorphic lenses, and complex articulate rotoscope work in a squeezed format. To eliminate the majority of these difficulties, it was decided that a spherical optic was necessary for our effects work. But we still had the problem of image degradation due to optical dupe steps. Once we had filmed the foreground elements and matched move backgrounds, these elements would then have to be combined in composite. The logical solution was to increase the negative size. Rather than use 65mm with its inherent processing and emulsion availability problems, we chose to use the horizontal 8-perf 35mm format similar to VistaVision. This not only gave us additional negative area, but also all of the advantages of high-quality spherical optics produced for conventional SLR still cameras with their incredible lens variety. One of the problems that the 8-perf format presented was optical combination. I decided that in-house opticals were a must in order to maintain our approach. We acquired an 8-perf printer (non-operative) and converted it to our electronics control approach. A second printer was made available to use. It was also converted to electronic control, and 4/8 perf operation. Due to the 8-perf horizontal format, we had to have lenses designed and constructed for the aerial position of our two-head printers, optimizing the image quality from the aerial head and making the printer considerably more flexible. With this choice, we had all the emulsions and processing available to conventional 4-perf production and a larger negative area.

Concurrent with the assembly of the optical department and model facilities we continued the development of the photographic and motion control systems. Having completed the design in July of 1975, we began construction of the "Dykstraflex" described as follows:

A system using stepping motors for control of any motion in the camera/subject positional relationship. These motors drive a track boom system with seven axes of motion and very precise tolerances. By using the camera frame rate and count as a time base, we can record camera/subject positional changes at 24 frames-per-second. The positional change is viewed through the taking lens and is controlled by a joy stick for multi-axes moves or by an individual potentiometer for single-axis moves.

If one axis, or more, of the move requires modification, we then go back and rerecord that individual axis, while still playing back all the axes that are good. Once we have described the camera move, including follow-focus, and viewed it real time, we can then change the operational time base from frames-per-second. to seconds-per-frame, and repeat the move precisely at the much slower rate used for actual photography.

This allows complex matched-move, multi-element matte shots by shooting the foreground against the blue screen and then taking that program of motions to a twin camera system. A separate background element can then be photographed with matching motions. When the two elements are combined, the appearance is that of real time photography— allowing pans, tilts, rolls, and accelerations on shots having a multitude of elements which were shot at different times, on separate cameras.

Due to the complexity of the matte shots on Star Wars we chose to use blue screen for out matting system. Thanks to the help of several experienced blue screen people, particularly Bill Rineholt, we worked out an expedient and versatile method of shooting and compositing each of the proposed shots. This required tight control of both the original photography that Richard Edlund produced on stage, and the system that Robbie Blalack was developing in optical for our blue screen composites.

After some consideration of conventional blue screen systems, we decided to innovate a little. First, we determined that we wanted a transmission screen, Steward (T matte). We used daylight-corrected, fluorescent tubes to facilitate movement of the screen and to optimize the screen's efficiency. We converted their operation from AC to DC to eliminate possible flicker problems with real time or high speed photography. This provided us with a portable backlit blue screen, easily moved by two men to cover the format on extreme camera moves. This screen has approximately twice the output of a conventionally lit transmission blue screen of the same size, draws the same amperage and is virtually "hot spot" free.

In addition, a rotoscope department was developed to work in conjunction with the optical department. Again, we found it necessary to construct the equipment ourselves. The rotoscope department not only provided garbage mattes and articulate rotoscope blob mattes, but also generated original images to be used for explosion enhancement, laser and flack effects.

With the blue screen matting system, we still had to deal with the question of how to support the miniatures. Articulate rotoscope matting of the streaked images to remove a support proved incredibly difficult. To solve this problem, we constructed what we call a "blue pylon". It consists of a central support tube containing electrical wiring as well as cooling air—the umbilical for the model. It is then wrapped with mercury vapor neon which, in turn, is wrapped with acrylic plastic and coated with blue screen material. The miniature is supported by this blue pylon. When the blue record is pulled to create the matte, the pylon completely matches the blue backing and eliminates the need for articulate rotoscope.

The department called "Control" has the responsibility of vaulting and routing each piece of film generated for this project. Control catalogues each piece of negative, and print, and their relationship to the 2 to 12 other pieces with which they will be combined in composite. Control has the responsibility of keeping infinitely accurate records and making film elements and information available on a moment's notice, a tremendous task considering the some 3,838 elements required for the 365 projected individual shots.

While all of the mechanical and electronic portions of the facility were being designed and constructed, one of the most important portions of the film was getting under way— the miniatures. Designs provided by Colin Cantwell were being modified to suit our photographic system and new designs for additional ships were being created. Joe Johnston had a major hand in providing the configurations that the final miniatures would incorporate. Grant McCune was assembling the people and facility needed to provide 75 models. Each of these models appear in several shots from a variety of camera angles. To accommodate this requirement, each specific design incorporated the ability to mount the miniature from the front, rear, top, bottom, and both sides. Each of the models included complex practical lighting for engine effects, laser weapon effects, and cockpit lighting. Some of the models had articulated details: wings that moved, rotating antennas, etc. All of the electrical leads for their motors and lights also had to be available at each of the mounting points. Because special high-wattage quartz bulbs were used in much of the practical lighting, cooling air was routed through the armatures of the miniatures.

The great number of miniatures which we would have to construct ranged from 1600 square feet of highly detailed surface architecture of a mechanical planet called the "Death Star", to tiny one-inch-high duplicates of the robots in the film, R2-D2 and C3PO. The scales of the miniatures had to suit the photographic methods used to record them. Edge detail had to be sharp but not too spindly because of the blue screen matting system. The size of each of the specific miniatures was determined by analyzing each shot projected for that miniature to appear in, and finding a scale that would satisfy depth of field, detail and size relative to other miniatures that had to appear with it in any given shot.

The color and paint detailing was given the same analysis and laid out with regard to the planets or other ships that a particular model was to appear over or with in a specific projected shot.

[ continued on page 2 ]