AI Tools

Generative AI models differ based on:

• Training distribution (dataset it was trained on)
• Internal aesthetic priors (different creative directors)
• Prompt parsing philosophy (composition keywords vs materials / lighting vs narrative metaphors)

an old worn book with a thick leather cover, slightly weathered edges and visible page layers, simple embossed details on the cover, resting on a flat surface, closed, low angle shot, realistic materials, 

soft ambient light,
clean composition,
balanced and readable structure,
neutral background,

Then build modular prompts (base structure): [subject] + [composition] + [lighting] + [style intent] + [constraints]

• Keep 70% identical across tests.
• Run cross-model batches.

Run on all models & Do NOT tweak per model

• Fantasy : stylized fantasy art
• Cinematographic : cinematic lighting, dramatic shadows, high contrast
• Hyper Realistic : ultra realistic, photorealistic, natural lighting, high detail textures
• Illustration : stylized illustration, hand-painted look, simplified shapes
• Bright / whimsical : vibrant colors, soft light, whimsical atmosphere
• Graphic / Design : minimalistic, flat lighting, clean shapes, product design aesthetic
• Painterly / Fine Art : oil painting style, visible brush strokes, artistic texture, classical painting
• Technical / Scan-Like : neutral lighting, flat illumination, orthographic feel, no shadows, documentation style
• Alternatives : Abstract / symbolic and Product-like / clean

• Soft : soft natural lighting
• High Contrast : strong directional light, deep shadows, high contrast lighting
• Volumetric / fog : volumetric lighting, light rays, subtle fog, atmospheric depth
• Neon / emissive : neon lighting, emissive glow, colored light accents, subtle bloom
• Low key / dark / moody : very low light, dark environment, minimal illumination, moody shadows
• Overexposed / bright : bright lighting, soft overexposure, high key, washed highlights

Keep the core part :

an old worn book with a thick leather cover, slightly weathered edges and visible page layers, simple embossed details on the cover, closed, low angle shot, realistic materials, cinematographic, soft natural light, 

• Isolated
• Structured composition : carefully arranged objects, balanced composition, clear spacing, readable layout
• Multi-object interaction : natural arrangement, slight overlap, coherent composition, readable subject
• Dense environment : rich environment, layered elements, moderate visual density, subject remains identifiable, depth
• Controlled clutter : cluttered desk, uneven arrangement, overlapping objects, semi-chaotic but still readable
• Chaos / fragmentation : chaotic scene, scattered objects, fragmented layout, unbalanced composition, visual disorder, reduced readability

Keep a single style (cinematographic, neutral) and probe through the different layers of complexity. Look for :

• Subject priority : Does the book stay the focus? Or does it get lost?
• Spatial intelligence : believable placement? gravity respected? overlaps make sense?
• Composition control : does it still “frame” the image? or just fills space randomly?
• Noise management : clean vs messy ? intentional vs accidental clutter ?

• Structural conflict : closed book, visible open pages
  • Tests whether the model protects object logic.
• Compositional conflict : chaotic scene, clean composition
  • Tests whether layout instructions override entropy.
• Visual hierarchy conflict : dense environment, subject remains dominant, reduced emphasis on subject

• Atlas Prompts

• nanoBanana 2

• Flux Pro 2

• Firefly 5

• Runway Gen4

• GPT Image 1.5

• Ideogram 3

• MidJourney

• MidJourney

• Runway : Video generation and Manipulation

• Latted : pay-as-you-go video models. Unified timeline and tools

• OpenAI FM : Text to speech+

• Scene: Describe the environment, main objects, and their relationships.

• Focus: Mention what should be most visible or important in the frame.

• Style: Specify realism, lighting, camera angle, color tone, etc.

• Purpose: Explain the context or use (e.g., documentation, tutorial, UI illustration).

• Reference Images: Mention if you have any uploaded files to guide style, layout, or color.

• Ultra wide Displays: 3440×1440: 43:18

• S23 Ultra: 1440×3088: 9:19.5 or 9:20

• Photographic / Photo realistic
• Anime
• Neon Punk
• Oil Paint
• Watercolor
• Fantasy Art
• Pencil Sketch
• Chalk Pastel
• Digital Art

Describe the subject in detail. Shape, size, how things are positioned.

Specify the desired dominant colors or mood board.

Describe materials.

• Extreme Long Shot
• Very Long Shot
• Long Shot/Wide Shot
• Full Shot
• Medium Long Shot
• Medium Shot
• Medium Close-up Shot
• Close-up Shot
• Extreme Close-up Shot

• Top-Down View (90°)
• High Angle View (45°)
• Eye-level View (0°)
• Low-angle View/Hero Shot (-15°)
• Extreme Low-angle View/Worm's eye View (-60°)

• 18mm : Super wide, capture a lot in the frame. Good for big landscapes or tight spaces. Adds depth and slight warping.
• 24mm : Versatile wide-angle view, great for portraits, street scenes, and landscapes. More background than a standard lens, less distortion and less warping.
• 35mm : Natural perspective closer to the human eye. Ideal for street photography or environmental portraits, where subjects interacts with their surroundings.
• 50mm : Standard and versatile. Very close to the human eye - natural and familiar feel.
• 85mm : Zooms in a bit, narrower view - flattering perspective for portraits. Minimizes distortion - popular choice for capturing people and food close-ups without warping the subject.
• 105mm : Telephoto - great for portrait photography, flattering look. Macro photography - shallow depth of field (blurry background) to isolate the subject.
• 135mm : Very narrow angle of view, nice magnification. last focal length for portrait photography - headshots with smooth blurry background.
• 200mm : Perfect for isolating subjects from a distance. Ideal for wildlife, sports, and astrophotography.

• Fisheye Lens : Super wide-angle, up to 180 degrees, bending the scene with a curved bubble effect. Trippy, distorted look.
• Tilt-Shift Lens : Blur parts of the image, creating a “miniature effect” that makes giant buildings look tiny.
• Lensbaby : Creative blur on most of the image and focus on a specific spot. Dreamy, magical effect.
• Infrared Lens : Reveal things invisible to the naked eye. Alien-looking landscapes. Mysterious and otherworldly vibes.
• Pinhole Lens : Soft photos with and a dreamlike quality. Experimental and back-to-basics approach to images.

• Short/Fast Shutter Speed : No blur, quick camera blink. Less light in, darker images. Good for catching splashes, jumps, or fast movements.
• Slow/Long Shutter Speed : Blurry waterfalls, streaky city lights. Shutter stays open longer, capturing movement and light over time. Good for showing motion or starry night skies.

• Wide Aperture : (low f-number between f/1.4 and f/5.6). Lots of light in with soft blur effect in the background, perfect for portraits or product pics.
• Narrow Aperture : (high f-number, like f/11 or f/32). Less light in but everything in focus, from front to back. Perfect for landscapes, buildings, or scenes with lots of details.

• Low ISO : (ISO 100 or 200) Perfect for bright scenes. Keep photo crisp and clear, no graininess.
• High ISO : (ISO 800, 1600, or beyond) Capture images in darker environments, sometimes adds noise.