Probably the most challenging problems faced in producing what ultimately amounted to 408 effects shots for the Aviator, were the budget and time constraints not normally associated with a film of this size. This coupled with the fact that these restraints are for a visionary director the stature of Martin Scorsese on a dramatic film of the artistic caliber of The Aviator, made the task all the more daunting. These restraints required and created a fresh approach to artistically match the level of the rest of the film. This approach required resurrecting older techniques and work flow methods used by the studio system in the heyday of film production ironically portrayed in the film itself.
Although most of us have budget limitations, The Aviator's visual effects budget was drastically cut from twelve million to less than six million dollars for a shot count that was at the same time raised up from 140 to 408 total shots. It must be the new math.
The filmmaking style of director Martin Scorsese, necessitated a less limiting approach to VFX that more closely matched his normal editorial workflow. A lower cost version of a core VFX "company" was created where the production paid the daily salaries and purchased low cost computers and software specifically for the production. This method is remarkably similar to the old studio system where every studio owned its own Special Effects department and equipment. Essentially any shot that the core team could produce would be at no additional cost to the production. The model is very similar to how a live action movie is produced; the crew and equipment are funded over a period of time, and the weekly budget remains the same whether you shoot ten shots a day or fifty. The same holds true for editorial post production as well, one version or ten versions of the cut can cost the same assuming you produce them in the same allotted schedule. This type of methodology created an organic and free flowing creative process identical to how Martin Scorsese normally and comfortably operates on his non-visual effects oriented films. Without having to worry about change orders or tracking who said what to whom, we could concentrate on only what was best for the film. Reality and math being as unforgiving as they are, an additional two million dollars (a bargain compared to the shot count) was ultimately needed to complete the work and extend the ever changing delivery schedule.
These financial limitations also created a workflow methodology creatively and ideally suited to this type of film. Since the film is a period piece, and the effects work needs to feel either seamless or appropriate to the time period, the solution was to film as many scenes as possible with techniques indigenous to old Hollywood films. In- camera illusions that involved hanging miniatures were used extensively. The now less expensive post time needed to complete these shots was either nonexistent or minimal depending on the cleanup.
All of the very specific creative work was done at the moment of photography even if shots required additional elements. The lighting, composition, grain structure, sky and cloud backgrounds were in fact pre-determined and required very little post enchancement which of course brought the expensive post time and costs down.
Radio controlled planes were employed for every major flying sequence and were shot in proper perspective to appear full size. The molds that created the flying models were the same molds created for the models used for motion control purposes. An unusual creative and cost saving procedure was adopted where the motion control photography was done outside under natural light. Besides looking more realistic, the sky background and its natural reflections in the plane were preserved. This also eliminated the various highlight, key, fill and background passes. Again postproduction was kept to a minimum further saving time and money without altering the quality of the image.