According to the Los Angeles Times, this part of the Luxor trilogy was projected on a vertical movie screen 70 feet (21 m) high by 35 feet (10.5 m) wide, with 350 seats rising at a 45-degree angle, each row four feet (1.2 m) higher than the one below.
Douglas Trumbull stated that the movie combined live-action with a physical miniature model of the city, superimposed with digitally generated flying cars and other aircraft. It was projected onto a deeply curved vertical screen. A gantry crane camera was used to take tracking shots of the miniature. Trumbull shot the live footage for the film at 48 frames per second, twice the normal speed, thus creating an exceptionally bright, sharp image. The process also entailed producing twice as many computer-generated images, all in very high resolution.
Though the city of the future is a physical miniature (i.e. a scale model), illuminated from the inside by almost a thousand real lights, the buildings were designed on a computer, and often sliced or decorated precisely with a computer-controlled laser cutter. [Sources: How did they do it? Computer Illusion in Film & TV, by Christopher W. Baker, and The Making of Luxor Las Vegas documentary]
As reported in the October 1993 issue of Computer Graphics World magazine, and the 'Computer Illusion in Film & TV' book, the digital additions to the scale model of the city also included a group of nearly 1,000 motion-captured human figures. They had been duplicated from a few, stored in a digital library, with randomized movements, clothes and skin colors to create enough variation. In addition, their movements were also 'looped' out of sync to enhance the illusion. All were subsequently subjected to the right environmental reflections of light and shadow. Because of limited processing power, a single scene could only contain 200 of them. Douglas Trumbull would have liked a total of 2,000 residents, but this turned out too difficult to achieve as well.