The Choice to Change Only One Part and Not the Whole Prompt

A modular approach to prompt engineering for consistent, evolving AI art.

We’ve all been there. You craft the perfect prompt. The composition is balanced, the lighting is magical, and the subject’s expression is exactly what you wanted. Then you think: "What if she was wearing a leather jacket instead of a cardigan?"

So you tweak a word. Suddenly, the lighting collapses, the background turns into a fractal mess, and your subject gains an extra finger. Frustrating, right?

The problem is often prompt entanglement – where keywords mix together unpredictably. The solution is modular prompting: changing a single, isolated element while leaving the rest untouched.

Here is a structured framework to help you do exactly that.

The 8-Pillar Prompt Structure

To change one thing without breaking everything, you need a clean architecture. Organize your prompt using these eight pillars:

1. Main subject

2. Face and body (if there are characters)

3. Clothing and styling

4. Environment and set design (location)

5. Light and atmosphere

6. Composition and framing

7. Photographic/artistic style

8. Technical keywords (for AI)

And here is the detailed breakdown of each pillar – the levers you can pull independently.

1. MAIN SUBJECT

• Who: Age, gender, ethnicity.

• Appearance: Hair type (color, length, texture), eye color, distinctive features (freckles, heterochromia, etc.).

• Expression: Serene, happy, intense, pensive, confident, etc.

• Posture: Body position (standing, sitting, in motion).

2. CLOTHING AND STYLING

• Outfit: Description of garments (e.g., white shirt, jeans, leather jacket).

• Style: Casual, elegant, sporty, bohemian, streetwear, etc.

• Materials & Textures: Linen, cotton, wool, silk, denim (adds realism).

• Accessories: Jewelry, bags, hats, glasses.

3. ENVIRONMENT AND SET DESIGN (LOCATION)

• Interior: Photo studio, café, library, bright apartment.

• Exterior: Urban park, city street, beach, forest, sunset.

• Environment Style: Minimalist, rustic, industrial, futuristic, vintage.

4. LIGHT AND ATMOSPHERE

• Light Type: Natural (morning, afternoon, golden hour), artificial (softbox, neon, window light).

• Light Quality: Soft and diffused, hard and direct, grazing.

• Atmosphere: Bright and airy, dramatic and dark, intimate and cozy, melancholic.

5. COMPOSITION AND FRAMING

• Shot Type: American shot, medium close-up, close-up, full figure.

• Angle: Front, high angle, low angle, three-quarter view.

• Depth of Field: Shallow (blurred background, bokeh) or deep (everything in focus).

6. PHOTOGRAPHIC / ARTISTIC STYLE

• Aesthetic: Photorealistic, pictorial, cinematic, vintage, film grain.

• Inspirations: Name of a famous photographer (e.g., Annie Leibovitz, Helmut Newton) or artist (e.g., Ilya Kuvshinov).

7. TECHNICAL KEYWORDS (FOR AI)

• Quality: Masterpiece, best quality, ultra detailed, 8K, UHD, sharp focus.

• Photographic Terms: Shot on (e.g., 85mm portrait lens), shallow depth of field, professional lighting, studio lighting.

Why This Works

When you structure your prompt this way, you create semantic boundaries. The AI learns to associate "leather jacket" with the clothing slot, not with the lighting slot. Changing the jacket won't reset your golden hour soft light because that keyword lives in a different pillar.

5 Practical Examples

Below are five prompts built using this modular system. Each example is followed by a variation – changing only one pillar.

Example 1: Changing ONLY the Outfit

Base Prompt (Street Style Portrait)

Main subject: young woman, 25, east asian, long black hair in a ponytail, freckles, confident expression, standing with hands in pockets. Face and body: oval face, brown eyes, slim build, natural makeup. Clothing: oversized cream wool sweater, black skinny jeans, white sneakers. Environment: urban park in autumn, fallen leaves, brick wall background. Light: overcast natural light, soft and diffused, no harsh shadows. Atmosphere: casual and relaxed, slightly melancholic. Composition: medium shot, eye-level angle, shallow depth of field. Style: candid street photography, Fujifilm Pro 400H aesthetic. Technical: masterpiece, 8K, sharp focus, shot on 50mm f/1.8, soft bokeh.

Variation (Changing ONLY Clothing → Grunge Style)

[...same subject, face, environment, light, composition...] Clothing: oversized black ripped denim jacket, faded band t-shirt, gray cargo pants, combat boots, silver chain necklace. [...]

✅ Result: Same face, same soft autumn light, same melancholic atmosphere – just a different outfit.

Example 2: Changing ONLY the Lighting

Base Prompt (Cinematic Portrait)

Main subject: man, 40s, mediterranean, short curly grey hair, intense green eyes, thoughtful expression, sitting on a wooden chair leaning forward. Face and body: strong jawline, stubble, athletic build. Clothing: navy blue linen shirt, brown leather belt, beige chinos. Environment: vintage library, tall bookshelves, old globe in background. Light: single softbox from camera left, Rembrandt lighting, warm 3200K. Atmosphere: intimate, intellectual, quiet. Composition: medium close-up, slightly low angle, shallow depth of field. Style: fine art portrait, inspired by Peter Lindbergh. Technical: best quality, ultra detailed, shot on 85mm f/1.4, studio lighting.

Variation (Changing ONLY Light → Moody Window Light)

[...same subject, clothing, environment, composition...] Light: dim natural light from a single rain-streaked window on the right, long shadows, cold 4500K, low key, high contrast. Atmosphere: melancholic, lonely, introspective. [...]

✅ Result: Same man, same library, same pose – but the mood transforms completely.

Example 3: Changing ONLY the Environment

Base Prompt (Cozy Interior)

Main subject: young couple, late 20s, mixed ethnicity, laughing together, lying on a sofa. Face and body: relaxed smiles, casual affectionate poses. Clothing: matching plaid pajamas, wool socks. Environment: cozy living room, Christmas tree in background, fairy lights, fireplace glow. Light: warm ambient light from fireplace and string lights, soft wraps around subjects. Atmosphere: intimate, joyful, holiday warmth. Composition: eye-level angle, wide shot (full figures on sofa), soft focus edges. Style: nostalgic polaroid aesthetic, slightly desaturated. Technical: masterpiece, film grain, shot on 35mm film.

Variation (Changing ONLY Environment → Rooftop at Sunset)

[...same couple, same clothing, same lighting quality...] Environment: city rooftop at golden hour, skyline in background, small bistro table with wine glasses, string lights overhead, potted plants. Atmosphere: romantic, expansive, hopeful. [...]

✅ Result: Same plaid pajamas, same soft warm light, same joyful expressions – now on a rooftop.

Example 4: Changing ONLY the Depth of Field

Base Prompt (Product/Fashion – Deep Focus)

Main subject: female mannequin with avant-garde dress, no face (abstract). Clothing: architectural gown made of recycled plastic, metallic silver, origami folds. Environment: minimalist white photo studio, seamless paper background. Light: bright overhead softboxes, clean and even, no shadows. Atmosphere: clinical, futuristic, sterile. Composition: full figure, straight angle. Depth of field: deep focus (everything sharp from front to back). Style: fashion catalog, high key. Technical: 8K, ultra detailed, shot on 70mm medium format, f/11.

Variation (Changing ONLY Depth of Field → Shallow Bokeh)

[...same subject, same studio, same lighting...] Depth of field: extremely shallow, sharp only on the dress's central fold, rest dissolves into soft bokeh. Style: artistic fashion editorial, dreamy. Technical: shot on 85mm f/1.2, creamy circular bokeh.

✅ Result: Same metallic dress, same clean studio – but the eye is forced onto a single detail.

Example 5: Changing ONLY the Artistic Style

Base Prompt (Photorealistic)

Main subject: elderly fisherman, 70s, southeast asian, weathered skin, deep wrinkles, missing one tooth, laughing heartily. Face and body: leathery tan skin, muscular but aged hands. Clothing: worn yellow raincoat, wide bamboo hat, rubber boots. Environment: wooden fishing boat on calm morning water, nets and buoys. Light: early morning golden hour, warm directional light from low horizon. Atmosphere: joyful, resilient, honest. Composition: medium shot, eye-level, shallow depth of field. Style: photorealistic documentary portrait, National Geographic. Technical: masterpiece, 8K, sharp focus, shot on 50mm f/1.8, natural light.

Variation (Changing ONLY Style → Oil Painting)

[...same subject, clothing, environment, light, composition...] Style: oil painting, impressionist brushstrokes, Van Gogh inspired, thick impasto, swirling sky. Atmosphere: expressive, emotional, painterly. Technical: masterpiece, oil on canvas texture, visible brushwork.

✅ Result: Same fisherman, same boat, same golden light – but rendered as a painting instead of a photo.

Final Tip: Lock, Then Unlock

When experimenting, treat your prompt like a mixing desk. Lock the pillars you love (e.g., #light, #composition). Change only the one you want to explore. Most AI image generators respond well to this structured clarity – because you're speaking in categories, not chaos.

Next time you're tempted to rewrite everything, pause. Change one word. Change one pillar. Watch how only the intended element transforms.

Happy prompting.

Try my LoRA V67 prompts now, they were all created in Civitai, (except some Flux components).

https://civitai.com/user/angelomaiota/models?sort=Newest

https://civitai.red/user/angelomaiota/models?sort=Newest

Angelo Maiota