
Enlarge a direct-gaze portrait in the upper half and place exactly two smaller moods below, connecting them through tape, torn paper, photocopy marks, newspaper and large Japanese lettering. Four saved variants preserve the one-large/two-small hierarchy without separately saved inputs, so they do not establish identity consistency or fixed issue details. Transit marks, phone, QR, dates and camera settings participate in the layout without factual validation. No generation or photography occurred.
Tool tags help you choose where to try a prompt. They do not verify which model or version created the example.