01Left role: the bow-tie character at the engine
Taller, in a dark flat cap, a white shirt with a black polka-dot bow tie, a navy waistcoat with a pocket-watch chain and a long black overcoat. For roughly the first second this figure is bent over the open hood with a hand on the plastic housing, face turned down. By about 2.5 seconds both roles are standing and the grey-suited figure is fully in view. From roughly eight seconds the left figure sits large at the left edge of the frame in profile, listening and gesturing with an open hand. At five seconds the face sits about a quarter of the way in from the left and a quarter of the way down.
02Right role: the grey-suit character in the yard
In a dark flat cap and a grey three-piece suit with a tie and watch chain, under a long black overcoat worn open. At one second this figure is half hidden behind the other, head tilted toward the engine. By about 2.5 seconds he stands beside the bow-tie figure with hands behind his back, and from about five seconds he faces the camera for the rest of the clip. The right role's mouth is clearly moving through the long excuse, with both arms out in a shrug near the end. The face sits a little over two-thirds of the way across the frame and about a third down at five seconds.
03Who speaks which line
Left says the opening diagnosis (0.0 to 1.8 s), the transmission line (2.5 to 3.8 s), "Yes, you did" (6.8 to 7.8 s) and the closing question (12.9 to 15.0 s). Right says "What?" (1.9 to 2.4 s), a short denial whose wording is disputed (4.5 to 5.5 s), "Nuh-uh" (8.0 to 8.5 s, heard in one transcript only) and the excuse (8.7 to 12.6 s). The transmission line is car slang for the car part. The wording comes from speech-to-text and a video model and may differ slightly; the dialogue page has the full script.
04Which photo goes where
The tool labels the left role Bow-tie character and the right role Grey-suit character, each with its own slot, so add each person's photo to the matching one. You can fill one role or both; an empty role keeps the character from the reference. As a rule of thumb, give the grey-suit slot your clearest front-facing photo, because that face speaks for most of the clip. For the bow-tie slot, a photo that shows the whole head and face clearly helps, since that figure is mostly bent over, in profile or at the frame edge. This comes from watching the reference and has not been tested.
05What the scene fixes and what is open
The yard, house, car, handheld camera and costume looks belong to the reference scene. How much of your photos' hair, skin tone and clothing carries over, especially the flat caps and long overcoats, is something to check in the result rather than assume. If you want the closest match, use photos of people already dressed in similar dark overcoats and flat caps.
06What this page does not claim
No generated example of this template is published here, and there is no measured success rate for any photo pairing. The finished video keeps the source soundtrack, strong language included. Upload only people who have agreed to appear, and label the result as AI-made when you share it.
