ai:atlas-prompts

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision		 Previous revision
ai:atlas-prompts [2026/04/02 23:32] – [🧙 Fantasy] mhai:atlas-prompts [2026/07/15 22:43] (current) – [Workflow] mh
Line 85: Line 85:
 ---- ----
  
-=== 🧩 Graphic / Design ===+==== 🧩 Graphic / Design ====
  
 <code txt> <code txt>
Line 93: Line 93:
 👉 Tests: flatness vs depth, shape hierarchy, contrast control, layout simplification 👉 Tests: flatness vs depth, shape hierarchy, contrast control, layout simplification
  
-=== 🧪 Technical / Scan-like ===+---- 
 + 
 +==== 🧪 Technical / Scan-like ====
  
 <code txt> <code txt>
Line 101: Line 103:
 👉 Tests: neutrality, absence of stylization, documentation behavior, strict material representation 👉 Tests: neutrality, absence of stylization, documentation behavior, strict material representation
  
-=== 🌤 Soft Natural (Baseline) ===+---- 
 + 
 +===== Lighting ===== 
 + 
 +==== 🌤 Soft Natural (Baseline) ====
  
 <code txt> <code txt>
Line 109: Line 115:
 👉 Tests: baseline lighting behavior, exposure balance, material readability under neutral conditions 👉 Tests: baseline lighting behavior, exposure balance, material readability under neutral conditions
  
-=== ⚡ High Contrast ===+---- 
 + 
 +==== ⚡ High Contrast ====
  
 <code txt> <code txt>
Line 117: Line 125:
 👉 Tests: shadow control, highlight clipping, dynamic range handling, lighting dominance 👉 Tests: shadow control, highlight clipping, dynamic range handling, lighting dominance
  
-=== 🌫 Volumetric Fog ===+---- 
 + 
 +==== 🌫 Volumetric Fog ====
  
 <code txt> <code txt>
Line 125: Line 135:
 👉 Tests: atmospheric depth, light scattering, edge clarity under haze, spatial layering 👉 Tests: atmospheric depth, light scattering, edge clarity under haze, spatial layering
  
-=== 🌈 Neon / Emissive ===+---- 
 + 
 +==== 🌈 Neon / Emissive ====
  
 <code txt> <code txt>
Line 133: Line 145:
 👉 Tests: color bleeding, emissive response, material-light interaction under colored lighting 👉 Tests: color bleeding, emissive response, material-light interaction under colored lighting
  
-=== 🌑 Dark Moody (Low Key) ===+---- 
 + 
 +==== 🌑 Dark Moody (Low Key) ====
  
 <code txt> <code txt>
Line 141: Line 155:
 👉 Tests: shadow preservation, readability in low light, noise handling, contrast discipline 👉 Tests: shadow preservation, readability in low light, noise handling, contrast discipline
  
-=== ☀️ Overexposed Bright ===+---- 
 + 
 +==== ☀️ Overexposed Bright ====
  
 <code txt> <code txt>
Line 149: Line 165:
 👉 Tests: highlight handling, detail retention under high exposure, washout resistance 👉 Tests: highlight handling, detail retention under high exposure, washout resistance
  
-=== 🔦 Directional Spotlight ===+---- 
 + 
 +==== 🔦 Directional Spotlight ====
  
 <code txt> <code txt>
Line 159: Line 177:
 ---- ----
  
-=== 🌇 Warm vs Cool Split Lighting ===+==== 🌇 Warm vs Cool Split Lighting ====
  
 <code txt> <code txt>
Line 169: Line 187:
 ---- ----
  
-=== 🧩 Structured ===+===== Scene Complexity ===== 
 + 
 +==== 🧩 Structured ====
  
 <code txt> <code txt>
Line 179: Line 199:
 ---- ----
  
-=== 🧱 Multi-object ===+==== 🧱 Multi-object ====
  
 <code txt> <code txt>
Line 189: Line 209:
 ---- ----
  
-=== 🌿 Dense Environment ===+==== 🌿 Dense Environment ====
  
 <code txt> <code txt>
Line 199: Line 219:
 ---- ----
  
-=== 🧹 Controlled Clutter ===+==== 🧹 Controlled Clutter ====
  
 <code txt> <code txt>
Line 209: Line 229:
 ---- ----
  
-=== 🌪 Chaos ===+==== 🌪 Chaos ====
  
 <code txt> <code txt>
Line 219: Line 239:
 ---- ----
  
-=== ⚔️ Chaos vs Control Conflict ===+==== ⚔️ Chaos vs Control Conflict ====
  
 <code txt> <code txt>
Line 226: Line 246:
  
 👉 Tests: conflict resolution hierarchy, chaos vs readability prioritization 👉 Tests: conflict resolution hierarchy, chaos vs readability prioritization
 +
 +===== 🧪 Constraint Layer =====
 +
 +The constraint layer is a dedicated suffix to be added to the above prompts for more creative models.
 +
 +The goal is to force them to hold back on the creative aesthetic, storytelling and scene interpreting and focus on the subject.
 +
 +=== Methodology ===
 +
 +  * keep 100% of your original prompts intact
 +  * append the constraint suffix
 +  * not touch internal wording (for atlas integrity)
 +
 +=== Constraint ===
 +
 +
 +🔹 1. STRICT (for simple prompts)
 +
 +//For Styles and Light prompts//
 +
 +<code txt>
 +single subject focus, no additional objects, minimal environment, clean composition, subject centered, fully visible, no obstruction, controlled framing, simple background
 +</code>
 +
 +🔹 2. RELATIVE (for complexity prompts)
 +
 +//For Scene Complexity prompts//
 +
 +<code txt>
 +subject remains dominant, clear focal point, readable composition, controlled layout, balanced elements, no subject obstruction, structured scene
 +</code>
 +
 +This constraint layer was not used and the ''--style raw'' flag was used instead in **MidJourney**
 +
 +===== 👤 Human Anatomy & Expression =====
 +
 +==== 👁️ Female Portrait Close-Up ====
 +
 +<code txt>
 +ultra detailed close-up portrait of a female subject, natural skin texture, visible pores, subtle asymmetry, realistic eyes, soft natural lighting, shallow depth of field
 +</code>
 +
 +👉 Tests: facial realism, micro-detail rendering, eye accuracy, skin texture fidelity, subtle asymmetry handling
 +
 +----
 +
 +==== 🧍 Female Full Body Neutral Pose ====
 +
 +<code txt>
 +full body photo of a female subject standing in a neutral pose, centered composition, realistic proportions, studio lighting, full figure visible from head to toe
 +</code>
 +
 +👉 Tests: body proportions, symmetry, posture correctness, full-body coherence
 +
 +----
 +
 +==== 🏃 Female Dynamic Motion ====
 +
 +<code txt>
 +female subject in dynamic motion, running and twisting mid-action, natural motion blur, limbs extended, realistic anatomy under movement
 +</code>
 +
 +👉 Tests: motion coherence, limb placement, anatomical consistency during action
 +
 +----
 +
 +==== ✋ Female Hands Close-Up ====
 +
 +<code txt>
 +close-up of female hands holding a small object, detailed fingers, natural grip, visible joints and nails, shallow depth of field
 +</code>
 +
 +👉 Tests: finger count accuracy, articulation, contact realism, fine motor detail
 +
 +----
 +
 +===== 🧬 Mythic =====
 +
 +==== 🛡️ Fantasy Warrior ====
 +
 +<code txt>
 +female fantasy humanoid character wearing layered armor and fabric clothing, intricate design, detailed materials, standing in a neutral pose
 +</code>
 +
 +👉 Tests: layering logic, costume complexity, material differentiation, humanoid design coherence
 +
 +----
 +
 +==== 🧙 Mythic Caster / Mage ====
 +
 +<code txt>
 +a powerful female mage standing in a mystical environment, flowing garments, magical energy emanating from hands, subtle arcane effects, grounded realism
 +</code>
 +
 +👉 Tests: energy effects, pose + gesture coherence, integration of magic with realism
 +
 +----
 +
 +==== 🐉 Dragon Creature ====
 +
 +<code txt>
 +a dragon standing on rocky ground, wings partially open, detailed scales, visible musculature, four limbs and two wings clearly defined, realistic anatomy, natural lighting
 +</code>
 +
 +👉 Tests: non-human anatomy, limb structure consistency, wing integration, scale detail rendering, anatomical plausibility
 +
 +----
 +
 +==== 🦌 Human-Animal Hybrid ====
 +
 +<code txt>
 +a humanoid figure blending human and deer features, antlers, elongated limbs, partially human face, organic fusion between species, natural environment
 +</code>
 +
 +👉 Tests: anatomical blending, uncanny balance, hybrid morphology plausibility
 +
 +----
 +
 +
 +===== 🏙️ Environment & Scale =====
 +
 +==== 🏔️ Natural Landscape ====
 +
 +<code txt>
 +wide natural landscape with mountains and a valley, atmospheric perspective, distant haze, layered depth, natural lighting
 +</code>
 +
 +👉 Tests: depth rendering, atmospheric perspective, large-scale composition
 +
 +----
 +
 +==== 🌆 Modern Cityscape ====
 +
 +<code txt>
 +large modern cityscape, dense buildings, streets and infrastructure, high level of detail, realistic perspective, daylight
 +</code>
 +
 +👉 Tests: structural repetition, urban density, perspective accuracy
 +
 +----
 +
 +==== 🪑 Interior Scene ====
 +
 +<code txt>
 +detailed interior scene of a room filled with objects, furniture, and decorations, natural lighting, realistic spatial arrangement
 +</code>
 +
 +👉 Tests: object relationships, spatial logic, scene composition
 +
 +----
 +
 +==== 🗿 Primordial Titan Scale ====
 +
 +<code txt>
 +a colossal primordial titan towering far beyond mountains, partially obscured by clouds, immense scale, environment dwarfed around it
 +</code>
 +
 +👉 Tests: extreme scale handling, integration with environment, sense of magnitude
 +
 +----
 +
 +===== 🎭 Scene Composition & Storytelling =====
 +
 +==== 📦 Single Subject Minimal Scene ====
 +
 +<code txt>
 +a single ancient object placed on a surface, minimal environment, soft lighting, strong focus on the subject
 +</code>
 +
 +👉 Tests: subject isolation, composition discipline, background suppression
 +
 +----
 +
 +==== 👩‍👩‍👧 Multi-Character Interaction ====
 +
 +<code txt>
 +two to four male and female characters interacting in a scene, natural poses, eye contact, subtle gestures, coherent group composition
 +</code>
 +
 +👉 Tests: multi-subject interaction, relational coherence, pose diversity
 +
 +----
 +
 +==== 🎭 Narrative Moment ====
 +
 +<code txt>
 +a narrative moment showing a female character discovering something unexpected, clear emotional storytelling, environment supporting the scene
 +</code>
 +
 +👉 Tests: storytelling ability, emotional expression, staging
 +
 +----
 +
 +==== ⚔️ Battlefield Chaos ====
 +
 +<code txt>
 +a chaotic medieval fantasy battlefield scene with male and female soldiers, action, movement, overlapping figures, weapons, dust and debris in the air
 +</code>
 +
 +👉 Tests: crowd handling, chaos vs readability, multi-character consistency
 +
 +----
 +
 +===== 💡 Lighting & Atmosphere =====
 +
 +==== 🌫️ Volumetric Fog ====
 +
 +<code txt>
 +scene filled with volumetric fog and visible light rays, atmospheric depth, soft diffusion, cinematic lighting
 +</code>
 +
 +👉 Tests: volumetric rendering, light scattering, atmospheric realism
 +
 +----
 +
 +==== 🌈 Neon Cyberpunk ====
 +
 +<code txt>
 +cyberpunk scene with strong neon lighting, emissive colors, reflections, night setting, glowing environment
 +</code>
 +
 +👉 Tests: color handling, emissive lighting, reflection interaction
 +
 +----
 +
 +==== 🌑 Low-Key Moody ====
 +
 +<code txt>
 +dark moody scene with low-key lighting, deep shadows, high contrast, dramatic atmosphere
 +</code>
 +
 +👉 Tests: shadow detail, contrast control, low-light behavior
 +
 +----
 +
 +==== ☀️ Overexposed High-Key ====
 +
 +<code txt>
 +bright overexposed scene with strong highlights, soft washed-out areas, high-key lighting
 +</code>
 +
 +👉 Tests: highlight control, exposure handling, detail retention
 +
 +----
 +
 +===== 🌊 Materials & Physics =====
 +
 +==== 💧 Water Splash ====
 +
 +<code txt>
 +water splashing dynamically, frozen moment, high detail liquid motion, droplets suspended in the air
 +</code>
 +
 +👉 Tests: fluid dynamics plausibility, motion freezing, particle detail
 +
 +----
 +
 +==== 🔥 Smoke & Fire ====
 +
 +<code txt>
 +smoke and fire rising in the air, complex swirling shapes, particles and embers, dynamic chaotic motion
 +</code>
 +
 +👉 Tests: chaotic systems rendering, particle behavior, flow realism
 +
 +----
 +
 +==== 🪞 Reflective Materials ====
 +
 +<code txt>
 +highly reflective objects made of glass and chrome, complex reflections of the environment, realistic lighting interaction
 +</code>
 +
 +👉 Tests: reflection accuracy, material understanding, light behavior
 +
 +----
 +
 +==== 💥 Material Impact ====
 +
 +<code txt>
 +a solid object breaking apart on impact, fragments and debris flying, realistic material fracture, frozen motion, high detail
 +</code>
 +
 +👉 Tests: rigid body physics, fragmentation logic, motion coherence, debris distribution
 +
 +===== 🧩 Structure & Symbols =====
 +
 +==== 🏛️ Architectural Symmetry ====
 +
 +<code txt>
 +perfectly symmetrical architectural structure, centered composition, precise geometry, repeating patterns
 +</code>
 +
 +👉 Tests: symmetry precision, alignment, geometric consistency
 +
 +----
 +
 +==== 🔁 Complex Patterns ====
 +
 +<code txt>
 +object with complex engraved patterns and repeating textures, highly detailed surface, consistent repetition
 +</code>
 +
 +👉 Tests: pattern consistency, detail repetition, drift resistance
 +
 +----
 +
 +==== 🔤 Embedded Readable Text ====
 +
 +<code txt>
 +a sign or surface with clearly visible English text reading "The Old Library", integrated naturally into a scene, realistic materials, readable typography, natural lighting
 +</code>
 +
 +👉 Tests: spelling accuracy, text legibility, typography coherence, integration of text into environment
 +
 +----
 +
 +==== 🔮 Fantasy Runes & Glyphs ====
 +
 +<code txt>
 +an ancient surface covered in glowing magical runes and glyphs, intricate symbols, non-human writing system, carved or engraved, mystical atmosphere, soft ambient light
 +</code>
 +
 +👉 Tests: symbolic consistency, pseudo-language generation, pattern logic, stylization of non-human writing systems
 +
 +----
 +
 +===== 🔍 Extended Probes =====
 +
 +==== 🐜 Macro Photography ====
 +
 +<code txt>
 +extreme macro shot of a small object or insect, very high magnification, shallow depth of field, fine surface detail
 +</code>
 +
 +👉 Tests: macro realism, depth of field accuracy, micro-detail generation
 +
 +----
 +
 +==== 🛰️ Aerial Top-Down ====
 +
 +<code txt>
 +aerial top-down view of a large environment, terrain or city, clear layout, visible structure from above
 +</code>
 +
 +👉 Tests: top-down coherence, layout logic, large-scale organization
 +
 +----
 +
 +==== 👗 Fashion Editorial ====
 +
 +<code txt>
 +fashion editorial photoshoot of a female subject, stylized pose, detailed fabric, controlled lighting, studio or location setting
 +</code>
 +
 +👉 Tests: pose direction, fabric behavior, stylistic control
 +
 +----
 +
 +==== 🎨 Abstract Emotion ====
 +
 +<code txt>
 +abstract representation of an emotion using shapes, colors, and forms, non-literal visual language
 +</code>
 +
 +👉 Tests: abstract reasoning, non-representational creativity, symbolic interpretation
 +
 +----
 +
 +
 +===== Local AI add-ons =====
 +
 +==== Positive Baseline (SDXL) ====
 +
 +<code txt>
 +high quality image, detailed, coherent composition, <REST OF PROMPT>
 +</code>
 +
 +==== Negative Baseline ====
 +
 +<code txt>
 +low quality, worst quality, low resolution, blurry, jpeg artifacts, watermark, signature, logo, unwanted text, text artifacts, bad anatomy, distorted anatomy, deformed body, extra limbs, missing limbs, extra fingers, missing fingers, fused fingers, malformed hands, malformed feet, duplicate subject, cropped subject, out of frame
 +</code>
 +
 +==== Workflow ====
 +
 +Lock the whole ritual circle:
 +
 +<code txt>
 +same prompt
 +same negative prompt
 +same checkpoint type family
 +same resolution
 +same seed policy
 +same sampler
 +same scheduler
 +same steps
 +same CFG
 +same VAE behavior
 +no LoRA
 +no ControlNet
 +no IPAdapter
 +no ref image
 +no upscale
 +no inpainting
 +</code>
 +
 +At first: checkpoints only.
 +
 +Then later:
 +
 +<code txt>
 +checkpoint + LoRA
 +checkpoint + sampler variation
 +checkpoint + CFG variation
 +checkpoint + style LoRA
 +checkpoint + ControlNet
 +</code>
 +
 +Don’t mix those into the baseline.
 +
 +Baseline settings suggestion for SDXL checkpoints:
 +
 +<code txt>
 +Resolution: 1024 x 1024 for square prompts
 +Steps: 30
 +CFG: 6
 +Sampler: dpmpp_2m
 +Scheduler: karras
 +Batch size: 1
 +Seed: fixed per prompt across all checkpoints
 +LoRA: none
 +Refiner: none, unless every checkpoint test uses the same refiner strategy
 +</code>
 +
 +For non-square prompts, use SDXL-friendly sizes close to 1 megapixel:
 +
 +<code txt>
 +Portrait: 832 x 1216
 +Landscape: 1216 x 832
 +Wide: 1344 x 768
 +Square: 1024 x 1024
 +</code>
  • ai/atlas-prompts.1775165545.txt.gz
  • Last modified: 2026/04/02 23:32
  • by mh