The Dykstraflex made a camera move repeatable, not just steady

The stage

American Cinematographer's July 1977 account of Star Wars describes a camera system built at the newly formed Industrial Light & Magic: the Dykstraflex, a track-and-boom rig with stepping motors on seven axes, supervised by John Dykstra with electronics from Al Miller Electronics and mechanical design by Don Trumbull, Richard Alexander and Bill Shourt, with Richard Edlund as director of photography on the effects unit. ILM's own account of its founding, as retrieved on 16 September 2026, confirms the studio was formed in 1975 for the production and that the Dykstraflex filmed bluescreen passes of miniature ships on ILM's Van Nuys stage.

What the documents establish

The 1977 account establishes a specific capability, not a marketing claim: the system recorded camera and subject position at 24 frames per second and could play that move back later, converted to a slower exposure rate, so a starship and a camera move photographed on one pass could be repeated for a second element, a matte pass, or a different lighting setup. That repeatability, not raw speed, is the documented achievement; the article gives no top speed, load capacity or cost figure, and this entry states none either. ILM's living history page corroborates the production context, the location and the team without adding new technical specification, consistent with a company page rather than an engineering record.

The department view

The Dykstraflex belonged to the effects camera department, distinct from principal photography, and its operators worked from a programmed move, previewing it on a joystick or potentiometers before committing film. That let the department shoot one pass, change only the model or the lighting, and rerun the identical move, trading a live operator's spontaneity for a composite that would register as one continuous shot once multiple elements were combined. It is a direct ancestor of a virtual-production brain bar cueing a repeatable camera path through a render engine rather than a physical rig.

What to check before you build

This is an editorial checklist. A claim tracing motion control to virtual-camera work should verify what is actually shared, programmability and repeatability, not any specific hardware or software lineage between the Dykstraflex and a modern system. Neither source states that later systems reused Dykstraflex components or code.

• Does the source describe a repeatable, programmable move, or only a stabilized or motorized one?
• Which department owned the rig, and did it operate separately from principal photography?
• Is a technical figure, axes, frame rate, speed, actually stated in the cited document, or assumed from later retellings?

The Dykstraflex's real precedent for a virtual-production stage is procedural: a camera move became data that could be previewed, stored and repeated, the same principle a modern brain bar applies to a render-engine camera decades later.