Genur.art

The prompt you provided is already highly detailed and well-structured for a text-to-image model, with clear visual direction, constraints, and stylistic preferences. Let’s refine it slightly to enhance clarity, cohesion, and parsing efficiency without adding new elements or deviating from your original intent. Here’s the expanded prompt: A vast field of black umbrellas viewed from directly overhead during heavy rain, each canopy displaying a unique recursive fractal pattern formed entirely by water droplets, ripples, and surface tension — some droplets branch into miniature river systems, others form nested circular interference patterns, fern-like crystal geometries, and infinitely repeating cellular structures that appear naturally embedded in the wet fabric; between the umbrellas, narrow cobblestone streets converge to form an enormous hidden Mandelbrot-like pattern visible only from above; cool slate gray, ink black, muted silver, and soft steel-blue tones dominate the scene, punctuated by occasional warm reflections from distant street lamps shimmering across puddles; high-altitude top-down photography with graphic composition, realistic rain physics, crisp wet textures, subtle motion blur from falling droplets, dramatic negative space, sophisticated mathematical repetition, atmospheric realism — no characters, no glowing effects, no fantasy architecture, no neon, no CGI appearance, no text, no watermark. This version preserves all your original subjects, actions, colors, spatial relationships, and constraints while enhancing readability and model compatibility through tighter phrasing and logical grouping. No new objects, characters, or styles were introduced — strictly adhering to your rules., muted minimalist sketch style