IMAGE EDITING TASK — GPT Image 2 (image-to-image with a reference). Base image = IMAGE 1 "Персона 1" (cybernetic man in a dark checked three-piece suit at a desk; matte-black segmented helmet-mask with blue glowing edge-lines and blue glowing eyes; framed "eye / СБП / МИР / под наблюдением" poster and rainy Kremlin window behind him). Reference = IMAGE 2 "перс редакт 5" — the correct target. Fix EXACTLY TWO things in IMAGE 1; everything else stays 100% identical. 1. HEAD SIZE — make it smaller: The head/helmet in IMAGE 1 is too big. Shrink the whole head so its size and proportion relative to the shoulders and body EXACTLY match IMAGE 2 (a smaller, well-proportioned head sitting naturally on the shoulders). Keep the same mask design, blue lines, eyes and gaze — only reduce its overall SIZE to match IMAGE 2. Reconstruct the neck/collar area cleanly so the smaller head connects naturally to the shoulders. 2. MOUTH — copy the SLIGHTLY OPEN mouth from IMAGE 2: Look carefully at IMAGE 2: the lower part of the mask has a SLIGHTLY OPEN mouth — a dark horizontal opening / parted mask-mouth (a recessed dark gap, as if the mask's mouth is slightly ajar), NOT a smooth surface. Reproduce this EXACT slightly-open mouth on IMAGE 1: same shape, width, position and the same dark recessed open gap as in IMAGE 2. Keep the nose exactly as it is. HARD CONSTRAINTS — do NOT change anything else: Do NOT change the mask design, blue helmet lines, eyes, blue eye-glow, or nose (only the head SIZE and the mouth change). Do NOT change the suit, tie, pocket square, gloves, pose, hands, desk and all objects (whisky glass, notebook, pen, lamp, books), the framed poster, the window, rain, Kremlin tower or the background. Only two edits allowed: (1) shrink the head to match IMAGE 2, (2) reproduce IMAGE 2's slightly-open mouth. STYLE: photorealistic, cinematic, dark moody lighting — identical to IMAGE 1.
Free to start · Generate videos and images with AI in seconds