The stunt had to work in one frame
In silent comedy, a dangerous-looking stunt could lose much of its force if the camera cut away at the crucial moment. Buster Keaton’s films often let the audience see his whole body—and the hazard—at once. That made the physical feat legible, but it also meant the timing and placement had to be right before the camera rolled.
The famous falling-facade gag in Steamboat Bill Jr. (1928) shows the problem in its starkest form. Keaton stands still as a large building front crashes down around him; he survives because he is positioned precisely where an open window in the falling wall passes over him. The opening is not a lucky gap discovered at the last second. The stunt depends on a specially prepared facade, a carefully chosen mark for Keaton to stand on, and a drop timed to bring the two into alignment.
Real danger, planned movement
That preparation did not make the stunt harmless. A moving wall, a train, or a fall could still injure an actor if anything shifted, arrived early, or landed in the wrong place. Keaton’s performance supplied the apparent calm, while construction, blocking—the planned positions and movements of performers—and rehearsal supplied the narrow margin for success.
This is why the feats read differently from a modern action sequence built around quick cuts and digital effects. The camera’s steady, wide view lets viewers track the same body through the same physical space. They can see both the danger and the small detail that makes survival possible: in the falling-wall scene, a window opening just large enough, arriving exactly where Keaton stands.
