Prompts con chat GPT resultados imagenes

Pruebas interesantes

Imagen Original
Prompt
Resultado Imagen

Rayos X

{
«prompt»: «Transform the subject in {{image}} into a hyper-detailed x-ray silhouette on a {{background}} background. Preserve all identifiable anatomical proportions, hairstyle, posture, hand positioning, and outline details exactly as in the original image. Remove all facial features, skin textures, and clothing folds, but reveal an internal skeletal impression that remains proportional and stylized for clarity—not medically accurate. Show the skull, ribcage, spine, hands, and major limb bones with glowing translucent contours. Use a monochrome palette of {{xray_colors}} with a luminous inner glow for bone structures. Maintain full silhouette and outer body shape for clear personal recognizability. Slightly blur and fade only the *outermost edges* so the subject softly blends into the {{background}} background. Do not distort proportions or add unnecessary shadows.»,
«style»: {
«lighting»: «Radiant inner bone glow with soft outer fade»,
«composition»: «Exact framing, scale, and camera perspective from {{image}}»,
«materials»: «Translucent x-ray aesthetic with luminous bone overlays»,
«special_effects»: «Subtle edge blur, internal glow, skeletal highlights»
},
«parameters»: {
«image»: «Input image»,
«xray_colors»: [«electric blue», «cyan white», «soft indigo», «ghostly gray»],
«background»: «deep black»,
«edge_softness»: «light fade outside silhouette»,
«skeletal_glow_intensity»: «medium-high»,
«bone_definition_level»: «stylized clarity»,
«saturation_level»: «low»,
«brightness_boost»: «focused on bone glow»
},
«medium»: «Stylized x-ray portrait with skeletal highlights and silhouette integrity»,
«size»: «match original image dimensions»
}

Aura de luz prompt:

{
«prompt»: «Convert the subject in {{image}} into a smooth thermal silhouette on a {{background}} background. Remove all facial features, textures, and clothing. The subject should appear as a clean, shape with no detail—just form. Apply a thermal gradient using {{gradient_colors}}, with soft blending and slightly increased saturation. The high heat zones on {{high_heat_zones}} should have a white glowing starburst ({{intensity_center}}). Heat zones on {{high_heat_zones}} should be brighter but not oversaturated. The edges must be blurred and slightly liquified so the figure melts softly into the {{background}} background, but the shape must remain clearly visible. Do not use strong red or hard shadows.»,
«style»: {
«lighting»: «Soft internal glow»,
«composition»: «same composition as {{image}}, no details, centered»,
«materials»: «Thermal gradient on smooth surface»,
«special_effects»: «Blurred edges, liquify, glow»
},
«parameters»: {
«image»: «Input image»,
«gradient_colors»: [«cool blue», «teal», «soft green», «muted yellow», «warm orange»],
«intensity_center»: «face»,
«high_heat_zones»: [«lips»],
«background»: «black»,
«edge_softness»: «very soft»,
«liquify_strength»: «medium»,
«saturation_level»: «medium»,
«brightness_boost»: «medium»
},
«medium»: «Digital thermal silhouette transformation»,
«size»: «match original image dimensions»
}

Estilo Minecraft ( sube tu imagen y dale este prompt) ⬇️

{ «method»: «img2img», «prompt»: «Transform the input image into a Minecraft-style voxel character. The generated image must match the exact framing, camera angle, zoom level, and subject size as the original input image. Do not change the perspective, pose, or orientation. The person’s facial structure, hairstyle, skin tone, clothing design, and visible features must match precisely. All geometry must be constructed using blocky voxel cubes with sharp 90-degree angles. The head must be a perfect cube. The torso, arms, and legs must be rectangular prisms, matching the original proportions and layout. No curved surfaces or smooth transitions are allowed. Textures must be exactly 16×16 pixels per surface, using flat color blocks with one level of two-tone shading to simulate depth. The eyes must be 2×2 black pixel squares placed symmetrically. The nose must be a single centered cube. The mouth must be a rectangular black or dark gray block, placed in the lower half of the face. Hair must only be added if it is clearly visible in the input image. Do not invent or add features not present in the original photo. The voxel version must match the subject’s silhouette, proportions, and details exactly. The background must be solid black. Lighting must be flat and even, with no shadows, gradients, or directional effects. The overall image must appear like a direct voxelized version of the input photo, with no deviation in composition or subject framing.», «style»: { «geometry»: «100% cube-based construction with no curves or smooth edges», «textures»: «16×16 pixel resolution with flat colors and simple two-tone shading», «eyes»: «2×2 black pixel squares, symmetrically placed», «nose»: «single centered cube protruding from face», «mouth»: «black or dark gray rectangular block», «hair»: «match exactly to input image; do not invent or exaggerate», «clothing»: «match shapes and color placement from input image using voxel blocks», «proportions»: «match input pose and layout, styled to Minecraft player dimensions», «lighting»: «flat ambient lighting with no shadows», «background»: «solid black (#000000)», «camera»: «match original camera angle, distance, zoom, and subject placement exactly» }, «parameters»: { «denoising_strength»: 0.6, «guidance_scale»: 8.0 }, «accuracy»: «1:1», «negative_prompt»: «different camera angle, zoom changes, reframing, altered subject size, photorealism, smooth surfaces, curved edges, added details, invented features, shadows, gradients, high-res textures, organic forms» }

Vistas de Personaje

Prompt:
Professional character design sheet with the title “CHARACTER CONSISTENCY DESIGN SHEET” prominently displayed at the top center. Neutral grey background featuring [describe character OR “the character from the reference image”]. Match the exact art style of the reference (photorealistic/anime/3D/illustration). The character’s facial features, body proportions, outfit, and accessories must remain identical across all depictions.
LAYOUT: Three labeled sections arranged clearly below the title:
SECTION 1 – Turning Diagram (3 full-body views):
– Front view: standing straight, hand at chin, facing camera
– Profile view: same pose, facing left
– Back view: facing away, showing outfit back details
SECTION 2 – Facial Expressions (6 close-up portraits, head and shoulders):
Neutral | Joyful (broad smile) | Sad (tear on cheek) | Angry (furrowed brows) | Surprised (wide eyes, open mouth) | Fearful (apprehensive eyes)
SECTION 3 – Pose Chart (4 full-body action poses):
Running dynamically | Jumping mid-air | Laughing (bent slightly) | Crying (crouching, hands covering face)
Clean labels beneath each image. Professional character turnaround style.
🚨 STOP Wasting Hours on Broken AI Videos! 🚨
Your character’s face keeps changing in every scene?
Hair color shifting? Outfit morphing?
You’re not alone. 95% of AI creators face this nightmare.
Here’s what the pros DON’T tell you:
The secret isn’t the AI tool. It’s the CHARACTER CONSISTENCY DESIGN SHEET.
Think of it like this:
Without a design sheet = Telling 10 different artists to «draw a girl»
With a design sheet = Giving them the EXACT blueprint

Prompt de avatar gema

Top Priority – Identity Lock: Always preserve exactly the same facial identity from the user’s reference image(s). Generate the requested scene (location, outfit, action, style, celebrities, etc.) without changing facial features, proportions, apparent age, skin tone, or bone structure. Zero face drift allowed.

Only modify environment, clothing, pose, lighting, camera angle and composition.

If the reference image is blurry, low-resolution, distant, or face not clearly visible, request a new photo before generating.

When users request scenes with celebrities, deliver them as realistic generated photo compositions.

Default output format: 9:16 vertical, unless the user specifies otherwise.

Mejoramiento Calidad de la imagen

Mejora el retrato, pero sin tocar la identidad del sujeto, o sea, que la geometría de la cara sea exacta. No le cambies la expresión ni la forma de la cara. Solo permite una limpieza sutil de los rasgos, sin alterar quién es.
Mantené el mismo fondo de la imagen de referencia. Nada de reemplazarlo, ni cambiarlo, ni meter objetos nuevos, ni mover nada del layout. El ambiente tiene que verse igualito.
La imagen tiene que recrearse como si la hubieran sacado con una Sony A1, usando un lente de 85mm f1.4, a f1.6, ISO 100, velocidad de obturación de 1/200, con una profundidad de campo cinematográfica, enfoque perfecto en la cara y un perfil de color neutro, como para una revista.
Esta configuración de Sony A1 + 85mm f1.4 es obligatoria. La imagen final tiene que verse claramente como si fuera de la calidad premium full-frame de una Sony A1.
La iluminación tiene que coincidir con la dirección, el ángulo y el ambiente exactos de la foto de referencia. Mejorá la iluminación a un estilo cinematográfico, enfocado en el sujeto: luz suave direccional, luces cálidas, sombras frías, contraste más profundo, rango dinámico expandido, un toque de micro-contraste, gradaciones suaves y cero sombras duras.
Mantené un tono de color premium neutro, una curva de contraste cinematográfica, saturación natural, textura de piel real (no de plástico) y un granulado sutil como de película. Nada de brillos falsos, ni luces de pasarela, ni suavizado excesivo.
Renderizá en resolución 4K, color de 10 bits, estilo editorial cinematográfico, claridad premium, recorte de retrato y mantené la onda ambiental original intacta.
Volvé a renderizar al sujeto con más realismo, profundidad, textura e iluminación, pero sin tocar la identidad ni el fondo, que tienen que quedar igualitos.
INSTRUCCIONES NEGATIVAS:
Nada de fondo nuevo.
No cambiar el fondo.
Nada de iluminación demasiado dramática.
No deformar la cara.
Nada de brillo falso.
Nada de iluminación plana.
No suavizar la piel en exceso.

mejoramiento de imagenes

Apply a warm, cinematic music-video style with strong golden backlighting (rim light) behind the subject, creating a soft halo on hair and shoulders. Keep the face clearly and evenly lit with soft frontal light. Use a shallow depth of field with a blurred background and rich amber/gold tones for a moody atmosphere. Preserve the subject’s original features, pose, outfit, background, and gender exactly. Enhance sharpness and skin detail with a realistic upscale, maintaining natural outdoor daylight and proper exposure.

1 comentario en “Prompts con chat GPT resultados imagenes”

  1. Hola profe soy de la ficha: 3199584
    Me intereso mucho con GTP está tan avanzando para crear imágenes de caricaturas o como aparecía en el artículo con la imagen de rayos x, ha avanzado tanto la inteligencia artificial para con unas solas palabras la IA pueda crear imágenes.

Deja un comentario

Tu dirección de correo electrónico no será publicada. Los campos obligatorios están marcados con *