The Good, the Bad and the Ugly (1966) — three-way cemetery standoff
See our character’s iconic-scene parody →In the middle of a circular cemetery, three gunmen stand in a triangle and hold out for minutes with nothing but their eyes and fingertips, without a single line of dialogue. Borrow it when you want to build tension by drawing out "who moves first", as in the moments before a final, a three-way match in a game, or a comparison ad pitting three products against each other.
Hundreds of Spanish soldiers built a cemetery like a Roman amphitheater, and the three men stand like dots in the middle of it (Wikipedia, clip at 4:27).
To recreate it: The first cut must clearly show where the three characters are, or in the later close-ups it gets confusing who is looking at whom.
The three stand at the vertices of a triangle on the circular ground, aiming at one another.
To recreate it: Decide the overhead layout first, and write each shot's gaze direction into the prompt to match that triangle.
From waist height, it catches fingers twitching over a holster to show "when he will draw" (clip around 5:32).
To recreate it: Generating the hand alone reduces finger distortion, so make a 1–2 second hand insert as a separate clip.
The face cuts tighten step by step until at the end only the brows and eyes fill the width of the screen (clip around 7:21).
To recreate it: Write the crop point, as in 'extreme close-up of narrowed eyes only, from brow to cheekbone', and have only the pupils move, with no blinking.
It widens the back-and-forth cross-cutting used for a two-person conversation to three people, showing the three faces alternately in the same order. According to the Cutting Concepts analysis, each time it goes around, the picture gets one step closer, and the villain gets the most cuts.
To recreate it: Make one set of the three characters' cuts at the same size, and make the next round again one step closer.
When one person looks to the side, the next cut shows the person he is looking at, so the gaze doesn't break even when the cut changes (Cutting Concepts).
To recreate it: Write the pupil direction in each cut, as in 'looking screen left', so the gazes line up when joined.
The louder the music gets, the faster the cuts, and the music cuts dead the moment the duel ends (Far Out).
To recreate it: Lay the music first and decide the cut points on the beat, then fit the generated clip lengths to that beat.
Cinematic multi-shot sequence, 12 seconds total. Shot 1 (3s): static extreme wide shot of three gunslingers standing in a triangle on a round stone plaza inside a vast cemetery of wooden crosses; Shot 2 (2s): static hip-level insert of fingers twitching above a holstered revolver; Shot 3 (2s): static extreme close-up of a squinting man's eyes under a hat brim, looking screen left; Shot 4 (2s): slow push-in on a scarred older man's narrowed eyes, looking screen right; Shot 5 (3s): static extreme close-up of a sweaty bandit's darting eyes. Consistent three dusty cowboys, harsh midday sun, 2.35:1 widescreen, 35mm film grain, no dialogue.
Leone is famous for alternating endlessly wide landscapes with extreme close-ups filled with just the eyes, and this duel stretches minutes with only that contrast and the editing, without dialogue (Wikipedia, No Film School). Leone worked out the main themes with Morricone before shooting and played them on set, matching the camera movement and editing rhythm to the music, and the duel follows the musical structure that runs from 'The Ecstasy of Gold' into 'The Trio' (Wikipedia). Because the cuts repeat regularly, the audience starts to anticipate the next cut, and that anticipation becomes the tension.
Extreme wide shot of three dusty gunslingers standing far apart in a triangle on a round stone plaza at the center of a vast cemetery of wooden crosses, rolling hills behind; static camera, harsh midday sun, 2.35:1 widescreen, 35mm film grain.
Hip-level insert of a weathered hand hovering above a holstered revolver on a worn leather gun belt, fingers twitching slowly, the edge of a poncho stirring in the wind; static camera, harsh midday sun, shallow depth of field, 35mm film grain.
Extreme close-up of a squinting man's eyes under a dusty hat brim, framed from brow to cheekbone across the whole widescreen frame, eyes sliding toward screen left; static camera, harsh sunlight, sweat and stubble, 35mm film grain.
Extreme close-up of a scarred older man's narrowed eyes framed from brow to cheekbone, staring toward screen right, a bead of sweat running down his temple; slow push-in, harsh midday sunlight, deep wrinkles, 35mm film grain.
Extreme close-up of a sweaty bandit's darting eyes flicking left and right between two unseen rivals, framed from brow to cheekbone, then freezing wide open; static camera, harsh midday sun, 2.35:1 widescreen, 35mm film grain.
After Shot 1 shows where the three are, run Shots 2–5 in the order of the three men. In the actual edit, repeat this set two or three times, making the cuts closer and shorter with each round (3 seconds, then 1 second, then 0.5 seconds), and set the cut points to the musical beat. After the last eye cut, hard-cut to the insert of the hand drawing the gun and cut the music dead.
So that the three characters' faces and clothes don't change from cut to cut, first make a reference image for each, and crop that image for the eye close-ups as the start frame. AI makes eyes blink often, so put in 'unblinking' and use the eye cuts for 2 seconds or less. Half of this scene is music, so lay the music in the editing program first and fit the cut lengths to it.
Wikipedia — The Good, the Bad and the Ugly · No Film School — The Mathematical Editing of Sergio Leone's 'The Good, the Bad, and the Ugly' · Far Out — Analysing the duel in 'The Good, The Bad, and The Ugly'