AI Anime Filter

[Subject] One young woman in a hyperreal flying selfie scene inspired by a whimsical witch-anime aesthetic. She appears early 20s, feminine presentation, slim build, light olive skin, large green-hazel eyes, long dark brown to black hair pulled back with loose strands blowing strongly in the wind, thin round glasses, medium gold hoop earrings, bright open smile showing teeth, rosy cheeks, and a joyful adventurous expression. She wears a dark navy dress or top. On her head is a very large bright red bow headband with white polka dots, tied dramatically above the crown. In her left arm she holds a small fluffy black kitten with yellow-gold eyes, white patch on the chest, and soft fur. Behind her left shoulder a straw broom is visible, angled backward in flight.
[Environment] High above a snow-covered mountain range under a vivid blue sky with soft white clouds. The ground far below is a textured expanse of icy peaks and ridges. The whole scene suggests fast airy motion through open sky, but remains bright and cheerful rather than dangerous. In the bottom-right corner of the image there is a small inset reference picture showing a more cartoon/anime-styled version of the same composition, accompanied by a curved red arrow pointing toward the main hyperreal image, indicating transformation from reference to realistic output.
[Composition/Camera] Vertical 3:4 composition with dynamic extreme selfie perspective, camera held high and close, subject face large and centered slightly right, arm extending toward the lens from the lower-right edge. The kitten sits in the lower-left foreground, close to the camera. The broom enters diagonally from the left-rear area. Hair and bow stream backward to emphasize movement. Bottom-right inset image occupies a small rectangular area and must remain clearly visible as a secondary element. Use a wide selfie lens feel around 20-24mm equivalent, but maintain attractive facial proportions.
[Lighting] Bright natural daylight from above and slightly front-left, with even illumination across the face, soft highlights on cheeks and glasses, and clear visibility of the kitten fur and bow texture. Sky and snow provide cool ambient bounce, while skin tones remain warm and lively. No harsh shadows; the mood should be crisp, optimistic, and airy.
[Style/Rendering] Photorealistic yet playful social-media comparison image, designed to show a cartoon-inspired concept translated into hyperreal photography. Clean, high-detail skin texture, realistic fabric, natural wind motion in hair, sharply rendered kitten fur, believable broom straw, saturated but controlled sky blues, and cheerful adventure energy. The inset should look noticeably more illustrated/anime-like, while the main image remains convincingly real.
[Detail constraints] Keep exactly one smiling flying subject, one black kitten, one straw broom, one oversized red polka-dot bow, and one small reference inset at bottom-right with a red arrow indicating transformation. Preserve the snowy mountain background and bright sky. Do not add extra characters, city elements, witches’ hats, magical sparkles, or multiple animals. This is a whimsical flying selfie with a realistic finish, not a fantasy battle scene.

Negative prompt: extra people, missing kitten, missing bow, missing broom, no inset reference image, no red arrow, witch hat, magical particles, dark storm sky, painterly main image, cartoon main image, distorted selfie face, warped cat anatomy, low-detail fur, generic clouds only with no mountains, text overlay, watermark.

Suggested parameters: aspect ratio 3:4, 20-24mm selfie lens feel, moderate depth of field, 28-38 steps, CFG/style strength 6.5-8, sampler DPM++ 2M Karras or equivalent, seed around 273644.

Delta prompt strategy:
1. If the cartoon-to-real comparison cue disappears: add "small anime-style reference inset at bottom-right with a curved red arrow pointing to the realistic main image".
2. If the bow becomes too small: add "oversized bright red bow with white polka dots dominating the top of the hairstyle".
3. If the kitten is missing or wrong: add "small fluffy black kitten with golden eyes and a tiny white chest patch held in one arm".
4. If the broom disappears: add "straw broom trailing diagonally behind the subject during flight".
5. If the scene loses motion: add "wind-swept hair and bow streaming backward, dynamic airborne selfie angle".
6. If the setting becomes generic sky: add "snow-covered mountain range far below, crisp icy ridges visible under the subject".
7. If the subject loses glasses: add "thin round eyeglasses clearly visible on the smiling face".
8. If the main image drifts cartoonish: add "main scene photorealistic, only the inset image remains anime-styled".
9. If facial proportions distort from wide angle: add "wide selfie lens with natural flattering facial proportions".
10. If lighting turns moody: add "bright cheerful daylight with clean sky and soft even facial illumination".

[Subject] A side-by-side winter portrait comparison featuring the same young woman shown twice in nearly matching styling and framing. She has fair skin, large blue-green eyes, long black hair tied into a high ponytail, oversized round wire-frame glasses, silver hoop earrings, soft pink lips, and a gentle friendly expression. She wears a fluffy white plush headband with a large bow on top, a pale pink puffer jacket with oversized white faux-fur collar, and a cream knit sweater underneath. On the left panel, her expression is slightly more neutral and direct, with one hand touching the collar near the lower right. On the right panel, she has a softer smile and slightly different hand placement near the coat opening. Keep the woman’s styling nearly identical in both panels while allowing minor natural variation.

[Environment] Snowy outdoor mountain setting in winter, blurred and pale in the background. The backdrop should show soft white snow, faint gray-blue mountain shapes, and floating snowflakes. The environment is simple and cold, but the subject remains warmly styled. This is not a single natural photograph: it is a split-screen comparison cover with two portrait panels placed side by side on a dark teal background. Each panel is framed as a rounded-rectangle card. White text overlays appear at the bottom of each panel: “NANO-BANANA PRO” on the left and “FLUX 2” on the right. Keep the full comparison layout because it is visibly part of the provided image.

[Composition/Camera] Vertical social-media comparison design, overall frame near 4:5. Two portrait cards fill most of the canvas, separated by a slim dark teal divider. Both portraits are medium close-ups from upper chest to slightly above the headband bow, centered and symmetrical enough to invite direct visual comparison. The left image is slightly tighter and cooler in facial expression, while the right image is a touch softer and more polished. Both subjects look directly at the camera. Preserve the clean side-by-side benchmarking layout and the card-like framing with rounded corners.

[Lighting] Soft overcast winter daylight with even frontal illumination on both faces. No harsh shadows. The light should feel diffuse, flattering, and cold-weather appropriate, keeping skin clear and smooth while preserving realistic facial depth. Snowflakes and pale mountain background remain softly lit. Overall color temperature is cool-neutral, but the pink jacket and cream knit add warmth. Maintain consistent lighting across both panels for fair comparison.

[Style/Rendering] Hyper-real winter selfie portrait with a social-media comparison aesthetic. The main emphasis is realism in skin, glasses, knit texture, faux-fur softness, and puffer-jacket material. The left panel should feel slightly sharper and more photographic, while the right panel can feel a little softer or more beautified, but both must remain plausible and high quality. The overall composition should read as a generator-versus-generator cover image, not a random collage and not a fashion magazine spread.

[Detail constraints] Do not remove the split layout. Keep exactly two vertical portrait cards of the same styled woman, with dark teal borders/divider and the white labels “NANO-BANANA PRO” and “FLUX 2” at the bottom of each respective panel. Preserve the pink puffer jacket, fluffy white collar, white bow headband, glasses, hoop earrings, ponytail, cream sweater, snowflakes, and snowy mountain backdrop. Do not convert the image into one single portrait or change the winter styling.

Negative prompt: single image only, missing split-screen, different people in each panel, blonde hair, no glasses, no bow headband, indoor background, Christmas room, ski goggles, heavy makeup, harsh sunlight, dark dramatic shadows, extra text, warped eyes, asymmetrical glasses, melted fur, deformed hands, anime illustration, painterly style, watermark clutter.

Suggested parameters: aspect ratio 4:5 vertical overall; lens 50-70mm equivalent portrait feel; aperture look f/2.8 to f/4; steps 30-40; CFG/style guidance 6.5-8; sampler DPM++ 2M Karras or photoreal portrait sampler; seed suggestion 286411570.

Delta prompt strategy:
1. If the split-screen disappears: "two rounded-rectangle portrait panels side by side on a dark teal background with a slim divider"
2. If the styling changes between panels: "same young woman in both images with matching pink puffer jacket, white bow headband, glasses, and cream sweater"
3. If the winter mood weakens: "snowy mountain background with floating snowflakes and cool diffuse daylight"
4. If the bow headband is wrong: "large plush white bow headband centered over a high black ponytail"
5. If the fur collar loses softness: "oversized fluffy white faux-fur collar around the neck and shoulders"
6. If the panels look too identical and artificial: "same subject, similar styling, slight natural variation in expression and hand placement between panels"
7. If text labels disappear: "white lower text labels reading NANO-BANANA PRO on the left and FLUX 2 on the right"
8. If it becomes a fashion editorial instead of a comparison: "social-media generator comparison cover, clean benchmarking layout"
9. If skin becomes over-retouched: "realistic skin texture, subtle winter softness, no beauty-filter plastic skin"
10. If the background gets busy: "minimal pale snowy mountains softly blurred behind the subject"

[Subject] A side-by-side winter city selfie comparison showing the same young woman in both panels. She has fair skin, long dark brown to black hair tied in a high ponytail, large round wire-frame glasses, silver hoop earrings, blue-green eyes, and soft natural makeup. She wears a medium-gray wool coat over a cream knit sweater and a thick light beige scarf wrapped around the neck. In the left panel, her smile is brighter and more casual, giving a friendly spontaneous selfie feeling. In the right panel, her expression is calmer and more refined, with a softer half-smile and slightly more polished look. Keep identity, clothing, and overall framing nearly identical across both panels, with only small natural variation.

[Environment] Outdoor city street at night in winter. Snowflakes drift across the scene. The background is softly blurred with warm orange streetlights, car bokeh, and tall dark skyscraper shapes rising behind. The atmosphere is cold, urban, and softly cinematic. The overall image is a social-media comparison cover made of two rounded portrait cards side by side on a dark teal background with a narrow divider between them. White text labels appear near the bottom of each card: “NANO-BANANA PRO” on the left and “NANO-BANANA” on the right. Keep this full comparison layout because it is visible in the original image.

[Composition/Camera] Vertical 4:5 overall comparison layout. Each panel is a medium close-up selfie crop from upper chest to slightly above the head. The woman is centered in both cards, with shallow background blur and snowy city lights behind. Left panel feels a touch looser and more candid, right panel slightly more centered and polished. Preserve the rounded-rectangle card framing, the dark divider, and the symmetrical side-by-side benchmark structure.

[Lighting] Soft cold evening ambient light on the face balanced with warm city bokeh in the background. The left panel can feel slightly more natural and on-location, while the right panel is a bit cleaner and more evenly flattering. Snowflakes catch small highlights. The subject’s face remains bright and readable without flash harshness. Keep realistic winter-night exposure and gentle contrast.

[Style/Rendering] Hyper-real winter selfie portrait with a social comparison aesthetic. Detailed glasses rims, natural skin texture, visible scarf knit and wool coat texture, subtle snow accumulation on hair and coat, and believable urban bokeh. Both panels should look high quality and plausible, with the left leaning slightly more candid and the right slightly more beautified. The frame should read as a generator benchmark cover, not a single standalone portrait and not a magazine editorial.

[Detail constraints] Do not remove the split layout. Keep the same woman in both panels with matching gray coat, cream sweater, beige scarf, glasses, hoop earrings, and high ponytail. Preserve the snowy city-night background, blurred streetlights, skyscraper silhouettes, and lower labels “NANO-BANANA PRO” and “NANO-BANANA”. Do not change the season, turn it into daylight, or replace the city with a studio backdrop.

Negative prompt: single image only, different women in each panel, indoor setting, no snow, daylight, forest background, no glasses, blonde hair, puffer jacket, dramatic makeup, heavy retouching, plastic skin, warped eyes, asymmetrical glasses, cluttered text, anime style, oil painting, watermark clutter, merged scarf.

Suggested parameters: aspect ratio 4:5 overall comparison; lens 50mm equivalent selfie-portrait feel; aperture look f/2 to f/2.8; steps 30-40; CFG/style guidance 6.5-8; sampler DPM++ 2M Karras or realistic portrait sampler; seed suggestion 480237161.

Delta prompt strategy:
1. If the split-screen disappears: "two rounded portrait cards side by side on a dark teal background with a slim divider"
2. If identity changes between panels: "same young woman in both panels with high dark ponytail, round glasses, hoop earrings, gray coat, and beige scarf"
3. If the winter city mood weakens: "night city street in snowfall with warm streetlight bokeh and distant skyscraper silhouettes"
4. If the scarf changes: "thick light beige scarf wrapped around the neck over a cream knit sweater"
5. If the left-right difference is lost: "left panel brighter candid smile, right panel calmer polished expression"
6. If snow disappears: "visible falling snowflakes and light snow dusting on hair and coat"
7. If it becomes too editorial: "social-media generator comparison cover, realistic winter selfie benchmark"
8. If labels vanish: "bottom white labels reading NANO-BANANA PRO on the left and NANO-BANANA on the right"
9. If the background gets too sharp: "soft urban bokeh with glowing car lights and blurred towers"
10. If skin becomes over-smoothed: "natural skin texture, believable winter-light portrait realism"

[Subject] A side-by-side winter portrait comparison showing the same young woman in two closely matched variations. She appears in her 20s with fair skin, large blue-green eyes, soft pink cheeks, dark eyebrows, delicate facial features, and a calm friendly expression. She has long black hair gathered back with a fluffy white bow headband or plush winter hairband. She wears oversized round silver wire-frame glasses and medium hoop earrings. Her winter outfit consists of a light pink puffer jacket with a plush white faux-fur collar, a cream ribbed knit sweater, and beige knit fingerless gloves or mitten-style hand warmers. In the left panel, she faces the camera more directly with one hand lifted near the collar; in the right panel, she holds the jacket edges or collar with both hands and has a slightly more polished, symmetrical pose. Both panels should depict the same person and same cold-weather styling.

[Environment] Snowy outdoor mountain setting with softly blurred pale blue-white mountains and winter sky in the background. Light snowflakes drift across the frame. The composition is presented as a social-media comparison board: two tall rounded rectangular portrait panels side by side with a thin dark teal divider between them and a dark teal outer border or background. Keep the background clean, bright, and wintry, with no buildings or city elements. Do not include any bottom labels or text overlays.

[Composition/Camera] Vertical 4:5 diptych layout. Each panel is a tight medium portrait from chest level to above the head, centered on the subject. Eye-level camera, straightforward beauty composition, shallow depth of field, and balanced spacing around the headband. Subject fills most of each panel, with the fluffy white collar occupying the lower-middle area and the snowy mountains softly visible behind. The left panel feels slightly more candid and open; the right panel is more refined and centered. Maintain strong consistency so the image reads as a clear model comparison.

[Lighting] Soft overcast winter light with cool ambient brightness and very gentle contrast. Even illumination across the face, no harsh shadows, and subtle pink warmth in the skin from the cold environment. Snowy daylight creates clean catchlights in the eyes and slight shine on the glasses rims. Overall the light should feel bright, diffuse, and flattering, with a crisp cold-weather atmosphere but no blue cast so strong that skin looks lifeless.

[Style/Rendering] Photorealistic AI beauty portrait with cozy winter fashion styling. Clean skin rendering, fine knit and puffer textures, believable faux-fur softness, and lightly stylized social-media polish. The goal is a premium generator-comparison image that still feels warm and aspirational. Preserve natural facial identity across both panels and avoid cartoonish perfection. No text, no watermark, no logos.

[Detail constraints] Show exactly one matching woman in two side-by-side winter portrait variants. Keep the fluffy white bow headband, round glasses, pink puffer jacket, white faux-fur collar, cream sweater, and beige knit gloves consistent in both panels. Maintain snowy mountain background, falling snow, rounded panel corners, and the dark divider/background frame. Exclude the source image’s bottom model-name labels. Do not add scarves, hats, extra jewelry, or extra people. The subject must remain soft, approachable, and cold-weather polished.

Negative prompt: text overlay, "NANO-BANANA PRO", "NANO-BANANA", watermark, logo, app UI, extra people, city street, indoor background, no snow, harsh sun shadows, different hair color, no glasses, wrong jacket color, no fur collar, bulky scarf, ski goggles, heavy makeup, anime look, plastic skin, duplicate person inside one panel, distorted hands, asymmetrical eyes, cropped headband, overly blue skin tone.

Suggested parameters for reproducibility: aspect ratio 4:5; portrait lens feel 70mm; aperture f/2.8 look; 28-36 steps; CFG/style strength 5.5-7; sampler DPM++ 2M Karras or equivalent; seed suggestion 592714308.

Delta prompt strategy:
1. If the image stops reading as a comparison: append "two tall rounded portrait panels side by side showing the same woman in matching winter styling".
2. If the headband changes: append "fluffy white bow headband framing the top of the dark hair in both panels".
3. If the outfit drifts: append "light pink puffer jacket with a thick white faux-fur collar over a cream ribbed sweater".
4. If the gloves disappear: append "beige knit fingerless gloves or hand warmers visible near the collar and jacket opening".
5. If the glasses are lost: append "large round silver wire-frame eyeglasses with soft daylight reflections".
6. If the mountain setting becomes vague: append "snowy alpine mountain background, softly blurred, with light falling snow".
7. If bottom labels reappear: append "image-only comparison board, no words, no captions, no model names".
8. If the lighting becomes too dramatic: append "soft overcast winter daylight, even facial illumination, no harsh shadows".
9. If the skin becomes too cold or lifeless: append "natural rosy cheeks and soft warm skin tones despite the snowy environment".
10. If both panels become identical in pose: append "left panel slightly more candid with one hand near the collar, right panel more centered with both hands holding the jacket".

[Subject] A side-by-side AI model comparison featuring two close winter city portraits of a young adult woman. In both panels, the subject has fair skin, a slim build, dark hair in a high ponytail, large round wire-frame eyeglasses, silver hoop earrings, and a soft approachable expression. She wears a light gray wool overcoat layered over a cream ribbed knit sweater and a thick pale-gray scarf wrapped around the neck. Left panel shows a slightly warmer expression with a broader smile and snow collecting on the hair and shoulders. Right panel shows a calmer softer smile and a cleaner, smoother rendering with similar styling.

[Environment] Both portraits are set outdoors on a snowy city street at night. Falling snow is visible in front of dark urban buildings and blurred traffic or street lights. The left panel has stronger colorful city bokeh and a warmer blue-gold nightlife atmosphere. The right panel has a more muted snowy street background with soft building outlines, parked cars, and a calmer wintry tone. Each panel is enclosed in a tall rounded rectangle with a dark divider between them.

[Composition/Camera] Two vertical comparison panels side by side, each showing a chest-up to upper-torso portrait. Subject is centered in both columns for direct visual evaluation. The winter styling remains matched across both images, and white text labels appear at the bottom of each panel: one reading “NANO-BANANA PRO” and the other “FLUX 2”. The overall design is optimized for a clean social-media model comparison layout.

[Lighting] Soft urban night lighting with cool ambient snow light and gentle warm reflections from streetlights. The faces remain clearly readable, with subtle frontal fill or flash-like clarity preserving eye detail. Snowflakes and coat textures remain visible, while each panel expresses slightly different tonal interpretation of the same winter-night setup.

[Style/Rendering] Realistic AI model comparison graphic focused on winter portrait realism, snow handling, skin fidelity, eyewear rendering, and urban low-light aesthetics. The image should feel like a polished benchmark post rather than a casual collage.

[Detail constraints] Keep exactly two portrait panels only, one labeled NANO-BANANA PRO and the other FLUX 2. Preserve the gray coat, gray scarf, cream sweater, glasses, ponytail, snowfall, and chest-up winter city composition in both. Do not merge them into one image or remove the comparison labels. Keep the layout clean and aligned for model evaluation.

Negative prompt: extra subjects, single-panel layout, no snow, daytime winter scene, no glasses, different outfits between panels, no labels, cartoon rendering, heavy fashion makeup, empty white background, no city context, no scarf, asymmetrical crop, indoor comparison, no rounded panel borders.

Suggested parameters: aspect ratio 4:5 overall with two vertical rounded panels, lens 60mm to 75mm portrait feel per panel, aperture f/2.8 to f/4 look, moderate depth of field, 24-32 steps, CFG 6.5-8, sampler DPM++ 2M Karras or equivalent, style strength low, seed around 604188.

Delta prompt strategy:
1. If the split comparison layout disappears: add “two tall rounded-rectangle portrait panels side by side with a dark central divider”.
2. If the winter styling drifts: add “light gray wool coat, thick pale-gray scarf, cream ribbed sweater, glasses, and high ponytail in both panels”.
3. If snowfall weakens: add “visible falling snowflakes collecting lightly on hair and coat in a night city setting”.
4. If the labels vanish: add “white bottom labels reading NANO-BANANA PRO and FLUX 2”.
5. If the city background gets too empty: add “blurred snowy urban street with headlights, building windows, and parked cars”.
6. If one panel becomes too different: add “same subject, same wardrobe, same chest-up framing for direct model comparison”.
7. If glasses disappear: add “large round wire-frame eyeglasses clearly visible in both images”.
8. If the color mood becomes identical: add “left side slightly warmer and more bokeh-rich, right side softer and more muted snowy realism”.
9. If the crop widens too much: add “tight winter portrait from chest-up optimized for side-by-side face comparison”.
10. If the image turns into a social collage instead of a benchmark: add “clean AI model comparison graphic with consistent framing and evaluative clarity”.

[Subject] A split-screen comparison showing the same young adult woman jogging outdoors in a park. She has a slim athletic build, light skin, dark hair tied in a high ponytail flying upward with motion, large round wire-frame glasses, hoop earrings, and a black fitted long-sleeve zip-front athletic top. Left panel: she faces forward with a serious concentrated expression, visible redness or flushed skin across the cheeks and forehead, and one arm pumping up in a running motion. Right panel: she records a close selfie while jogging, smiling gently with a more relaxed expression, still wearing the same glasses and black running top.

[Environment] Daytime urban park or tree-lined walkway with paved paths, green grass, leafed trees, and soft city-park background blur. The two panels show slightly different path textures and angles but the same general outdoor jogging environment. Light is natural and mild, likely late afternoon or soft daytime with no harsh shadows.

[Composition/Camera] Vertical two-panel split layout separated by a narrow divider. Left side is a front-facing medium shot captured while running, with the subject centered on a park path and torso visible. Right side is a close arm-extended running selfie, slightly angled downward and to the side, with the subject filling most of the frame. Bottom labels identify the comparison: “FLUX 2 Klein” on the left and “NANO-BANANA PRO” on the right, with a small colorful icon above the right text.

[Lighting] Soft natural daylight, evenly illuminating the face and clothing. The outdoor light is diffused enough to keep skin details visible without harsh contrast. Motion in the hair and subtle environmental blur communicate running movement more than aggressive shutter blur.

[Style/Rendering] Photoreal social-media fitness selfie comparison, casual sports realism, natural skin texture, smartphone capture aesthetic, slight motion energy, clean outdoor portrait with everyday creator-content feel rather than polished campaign imagery.

[Detail constraints] Keep the split-screen comparison with one running portrait on the left and one jogging selfie on the right. Preserve the high ponytail, round glasses, black zip athletic top, park background, and the bottom model labels. Do not add extra people in the foreground, gym equipment, or dramatic fitness branding.

Negative prompt: indoor gym, sports bra, tank top, sunglasses, no glasses, static studio pose, marathon bib, bright midday harsh sunlight, extra runners in foreground, cartoon style, over-smoothed skin, heavy motion blur, missing split layout, text overlays beyond the comparison labels.

Suggested parameters: aspect ratio 4:5 vertical; lens 35mm to 50mm smartphone portrait feel; medium depth of field; 22-32 steps; CFG/style strength 5.5-7; sampler DPM++ 2M Karras or equivalent; seed suggestion around 384517269.

Delta prompt strategy:
1. Split-screen layout collapses -> append: side-by-side two-panel comparison with narrow vertical divider and bottom labels.
2. Ponytail loses motion -> append: high ponytail whipping upward from running movement.
3. Glasses disappear -> append: large round wire-frame glasses clearly visible in both panels.
4. Outfit changes -> append: fitted black long-sleeve zip-front running top with minimal branding.
5. Park setting turns generic city street -> append: tree-lined urban park walkway with grass and soft outdoor background.
6. Left panel loses exertion cues -> append: forward-facing run with flushed cheeks and focused expression.
7. Right panel stops reading as selfie -> append: close arm-extended jogging selfie with relaxed smile and smartphone perspective.
8. Lighting becomes too harsh -> append: soft diffused daytime natural light with even facial exposure.
9. Image becomes polished commercial fitness ad -> append: casual creator-style running content, realistic and social-media native.
10. Labels disappear -> append: comparison text at the bottom, FLUX 2 Klein on left and NANO-BANANA PRO on right.

[Subject] A split-screen in-car burger-eating comparison featuring the same young woman shown in two near-matching daylight selfies inside a parked or slow-moving car. She appears to be in her 20s with fair skin, long dark hair styled in a high ponytail, thin round silver eyeglasses, large silver hoop earrings, and a plain white crew-neck T-shirt. In both panels she wears a seatbelt across her chest and holds a burger with both hands. In the left panel she looks slightly awkward or uncertain with subtly tense eyebrows, while in the right panel she appears more relaxed and natural with a small food bite visible in her mouth or a casual chewing expression. The right side should feel slightly more realistic and human than the left while preserving the same identity.

[Environment] Daytime car interior with visible headrests, windows, pale roof liner, side pillar, and in the right panel part of the steering wheel at the lower edge. Natural daylight enters through the side windows, creating a clean everyday realism test. The image is presented as a two-panel comparison card with rounded corners and a thin dark divider. Bottom labels should read “FLUX 2 Klein” on the left and “NANO-BANANA PRO” on the right, with a small colorful sparkle icon near the right label.

[Composition/Camera] Vertical split-screen layout with both panels framed as chest-up car selfies. The left panel is slightly tighter and more off-center, showing more burger detail and a mildly uncomfortable expression. The right panel is centered more cleanly, with the steering wheel visible and a more balanced portrait feel. Keep both crops comparable enough for direct model evaluation and preserve strong visibility of the seatbelt and burger in both images.

[Lighting] Natural daylight only, soft and realistic, with gentle highlights on skin, glasses, and burger bun. The lighting should feel like a real phone selfie inside a car during the day, with no flash and no dramatic contrast. The right panel may appear slightly cleaner, brighter, and more natural in skin rendering.

[Style/Rendering] Hyper-real everyday comparison graphic focused on food handling, hand realism, car interior fidelity, and candid facial expression. The image should feel like a practical benchmark for AI realism in ordinary situations rather than a stylized ad. Maintain social-media comparison card aesthetics and clear readability.

[Detail constraints] Keep exactly one female identity duplicated across two car-interior panels. Preserve the white T-shirt, seatbelt, glasses, hoop earrings, burger, daylight car cabin, and bottom model labels. Do not add other passengers, fast-food branding, dramatic makeup, fantasy lighting, or outdoor scenic distraction. The core point is ordinary realism under highly familiar conditions.

Negative prompt: extra people, no seatbelt, no burger, luxury car glamour, restaurant interior, overly polished food ad lighting, mismatched identity between panels, distorted hands, melted burger, text errors, watermark, cartoon car interior, night scene, heavy beauty retouch.

Suggested parameters: aspect ratio 4:5, smartphone selfie lens feel, daylight car interior, steps 28-40, CFG/style strength 6.5-8, sampler DPM++ 2M Karras or equivalent, seed suggestion 268155704.

Delta prompt strategy:
1. If the two identities drift: “same exact young woman in both panels with matching glasses, earrings, ponytail, and facial structure”.
2. If the car context weakens: “clear daytime car interior with seatbelt, headrest, window light, and steering wheel visible in the right panel”.
3. If the burger handling looks wrong: “natural two-handed grip on a realistic burger with proper bun, lettuce, and patty structure”.
4. If the right panel is not better: “right panel slightly more natural and photoreal in expression, skin texture, and hand realism than the left”.
5. If the labels fail: “bottom left text FLUX 2 Klein, bottom right text NANO-BANANA PRO”.
6. If the scene becomes too commercial: “casual everyday car selfie, not a food advertisement”.
7. If the seatbelt disappears: “seatbelt crossing diagonally over the white shirt in both panels”.
8. If the lighting becomes artificial: “soft natural daylight through car windows, no flash, no dramatic shadows”.
9. If the burger gets too clean: “ordinary handheld burger with small bite marks and realistic food mess”.
10. If the left expression becomes too polished: “slightly awkward uncertain facial expression on the left, more relaxed casual expression on the right”.

[Subject]
A two-panel comparison image featuring the same young adult woman in a casual indoor party setting. She has light skin, large round wire-frame glasses, dark brown or black hair in a high ponytail, hoop earrings, and a candid low-key expression. She wears a simple black sleeveless tank top. In the left panel she leans forward with crossed arms or folded posture on a couch or soft surface, glancing sideways with a thoughtful expression. In the right panel she reclines back against a sofa arm, shoulders angled toward camera, giving a detached or tired look.

[Environment]
Dim apartment or house-party living room with direct flash photography, beige or neutral couch fabric, blurred people in the background, cups or casual party clutter, and a dark ambient room tone. The scene should feel like a real late-night candid captured in a social setting, not a staged studio shoot.

[Composition/Camera]
Vertical split-screen comparison layout with rounded corners and a narrow teal divider between the left and right panels. Both crops are intimate medium close-ups from chest-up or torso-up. Maintain bottom labels: "Higgsfield SOUL 2" with neon-green icon on the left and "NANO-BANANA PRO" with multicolor icon on the right. The two frames should feel like alternate outputs of the same candid scenario.

[Lighting]
Direct on-camera flash with hard frontal light, visible skin highlights, flat but realistic party-photo illumination, mild falloff into the dark room, and bright catchlights on the glasses. Preserve the imperfect casual flash aesthetic rather than smoothing it into polished editorial light.

[Style/Rendering]
Photoreal social-media comparison graphic with documentary nightlife energy, realistic skin texture, believable flash harshness, slight background noise, and authentic candid-photo mood. Keep it grounded and imperfect in a convincing way.

[Detail constraints]
do not remove the two-panel layout, keep the same woman in both panels, preserve the glasses, black tank top, indoor couch setting, background partygoers, and the bottom model labels. The image must feel like a realism comparison for candid flash photography.

[Negative prompt]
studio backdrop, clean white room, no background people, glamour beauty lighting, outdoor night street, missing glasses, different outfits, heavy makeup editorial styling, anime, cartoon, deformed arms, blurred face, no couch, no flash look, empty background, missing text labels

[Suggested parameters]
- aspect ratio: 4:5 vertical overall
- lens/focal length: 35mm to 50mm equivalent with point-and-shoot flash feel
- depth of field: shallow-medium
- steps: 30-38
- CFG/style strength: 5.5-6.8
- sampler: DPM++ 2M Karras or Euler a
- seed suggestion: 56427318

[Delta prompt strategy]
1. If the candid party vibe disappears, append: "late-night indoor house party with blurred people in the background"
2. If the flash look softens too much, append: "direct on-camera flash, harsh frontal light, realistic casual snapshot"
3. If the split comparison vanishes, append: "two side-by-side benchmark panels with rounded corners"
4. If the woman changes between panels, append: "the same woman appears in both frames with matching glasses and ponytail"
5. If the couch setting is lost, append: "leaning on a beige fabric sofa in a dim living room"
6. If wardrobe drifts, append: "simple black sleeveless tank top, hoop earrings, round glasses"
7. If the background becomes too clean, append: "softly blurred partygoers and casual room clutter behind her"
8. If skin becomes too airbrushed, append: "realistic flash skin texture, slight shine, natural imperfections"
9. If labels disappear, append: "bottom labels for Higgsfield SOUL 2 and NANO-BANANA PRO"
10. If the mood becomes too posed, append: "unpolished candid snapshot energy, intimate and slightly awkward"

[Subject]
A split-screen comparison image featuring the same young adult woman in a swimming pool at night. She has fair skin, large round metal glasses, hoop earrings, dark hair pulled back tightly or in high tied sections, and a bright expressive face. In the left panel, only the upper face and glasses rise above the waterline, creating a dramatic half-submerged close-up. In the right panel, more of the body is visible above the water, showing a dark one-piece swimsuit and a friendly smile.

[Environment]
Nighttime pool scene with a nearly black background, strong flash-lit subject, reflective water surface, visible ripples, underwater light scatter, and floating particles or droplets. The pool water should feel deep and glassy with realistic reflections and refraction. Keep the setting minimal so the viewer focuses on water realism and facial detail.

[Composition/Camera]
Vertical two-panel comparison layout with rounded corners and a thin divider between panels. Left panel is a tighter close-up portrait dominated by face and waterline. Right panel is a medium portrait showing upper torso in water. Both panels should align visually as a model-comparison graphic. Bottom labels identify the models: "Higgsfield SOUL 2" on the left with a small neon-green icon, and "NANO-BANANA PRO" on the right with a multicolor diamond icon.

[Lighting]
Direct flash or strong frontal night lighting that creates crisp highlights on skin, bright reflections on the glasses, luminous specular highlights on the waterline, and a glossy wet-skin look. Preserve strong contrast against the dark background while keeping the face readable.

[Style/Rendering]
Photoreal AI comparison graphic focused on water realism, skin texture, flash portrait aesthetics, and high-contrast nocturnal mood. The result should feel like a creator benchmark post comparing image generators on a difficult refraction-heavy scene.

[Detail constraints]
do not remove the split-screen comparison, preserve the same woman in both panels, keep the glasses, waterline crossing the face or torso, dark night background, one-piece swimsuit on the right, and the model-name text labels at the bottom. Water realism and reflective distortion must remain central.

[Negative prompt]
daylight pool, beach scene, multiple people, no glasses, dry skin, no waterline, missing reflections, underwater camera only, cartoon water, anime, broken facial symmetry, distorted eyes behind glasses, random pool toys, text missing, over-smooth skin, different woman in each panel

[Suggested parameters]
- aspect ratio: 4:5 vertical overall
- lens/focal length: 50mm portrait with flash feel
- depth of field: shallow-medium
- steps: 32-40
- CFG/style strength: 5.5-7.0
- sampler: DPM++ 2M Karras or Euler a
- seed suggestion: 90734126

[Delta prompt strategy]
1. If the waterline looks fake, append: "realistic water refraction crossing the face, crisp reflective surface tension"
2. If the split comparison disappears, append: "two vertical benchmark panels side by side with rounded corners"
3. If the night mood weakens, append: "dark nighttime pool background with flash-lit subject"
4. If the glasses deform, append: "large round metal glasses with realistic reflections and correct lens shape"
5. If the left panel becomes too open, append: "tight close-up with only eyes, glasses, and upper face emerging above water"
6. If the right panel loses body context, append: "upper torso visible in water, dark one-piece swimsuit, smiling toward camera"
7. If the water becomes too clean, append: "tiny suspended particles, ripples, and specular highlights in the pool"
8. If labels vanish, append: "bottom comparison labels for Higgsfield SOUL 2 and NANO-BANANA PRO"
9. If the image becomes too glamorous, append: "benchmark-style realism test, flash portrait honesty, not luxury resort imagery"
10. If likeness drifts between panels, append: "the same woman appears in both panels with matching face and glasses"

[Subject] A side-by-side comparison image showing the same young woman holding a large stemmed wine glass directly toward the camera in a dim restaurant or bar setting. She appears in her early 20s with fair skin, green-hazel eyes, large round silver wire-frame glasses, hoop earrings, dark brown to black hair tied into a high ponytail, and a black sleeveless top. The left panel shows a softer, subtler smile and a more subdued mood, while the right panel shows a brighter open smile and more vivid expression. In both panels, the wine glass dominates the foreground and partially distorts or overlaps the face due to glass curvature and reflections.

[Environment] Two equal vertical portrait panels in a dark indoor nightlife environment. The background is mostly black or very dim, with a few warm amber bokeh lights and hints of restaurant or bar interior on the right panel. The left panel feels more flash-heavy and synthetic, with the glass flattening the scene. The right panel feels more real and dimensional, with better depth, cleaner hand rendering, and more natural low-light atmosphere. Bottom labels identify the sides as “Higgsfield SOUL 2” on the left and “NANO-BANANA PRO” on the right, each with its respective small icon.

[Composition/Camera] Tight vertical close-up portrait in split-screen A/B format. In both panels the woman is centered and holds the wine glass close to the lens so the bowl of the glass fills a large portion of the lower-middle foreground. The top rim of the glass crosses the face around nose or eye level, producing reflections, magnification, and distortion. Keep the framing intimate and social-media-like, as if taken during a night out with direct flash or strong phone light.

[Lighting] Flash or strong frontal phone light creates bright specular highlights on the wine glass, liquid surface, eyeglasses, and skin. The left side should feel slightly harsher and more artificial, with the glass flare flattening the face. The right side should maintain stronger realism, with more believable skin tones, glass thickness, liquid refraction, and background depth while still keeping the bright nightlife flash feel. Preserve white reflections and small point highlights on both the glass and eyeglass lenses.

[Style/Rendering] Hyperreal dual-panel nightlife benchmark image focused on glass realism, facial consistency, and flash-light behavior. Left panel: more synthetic and flatter. Right panel: more believable social-nightlife portrait with stronger physical material response. The tone should remain intimate, casual, and creator-like rather than editorially polished.

[Detail constraints] Keep exactly one woman in each panel and make them clearly the same person with the same glasses, ponytail, earrings, and black top. Preserve the dominant wine glass in the foreground, the dim bar ambiance, and the split-screen comparison layout. Do not add extra people, extra drinks, table clutter, or heavy color effects.

Negative prompt: single panel, no glass in foreground, cocktail garnish overload, champagne flute instead of rounded wine glass, outdoor scene, bright daylight, extra people, nightclub lasers, no glasses, no earrings, smiling differently beyond recognition, text removed, watermark, distorted fingers, missing stem, fantasy sparkles, luxury editorial studio shot.

Suggested parameters: aspect ratio 4:5, two-panel split, close portrait lens or smartphone flash portrait feel, shallow to moderate depth of field, 28-36 steps, CFG/style strength 6-7, sampler DPM++ 2M Karras or equivalent, seed suggestion 269145883.

Delta prompt strategy:
1. If the wine glass loses dominance: append "large rounded stemmed wine glass held very close to the lens, occupying much of the foreground".
2. If identity drifts between panels: append "the exact same woman appears in both panels with matching glasses, ponytail, earrings, and face".
3. If the left side is not flatter: append "left panel feels harsher, flashier, and more synthetic with flatter glass response".
4. If the right side is not more realistic: append "right panel shows cleaner skin, more believable glass refraction, and stronger low-light depth".
5. If the bar atmosphere disappears: append "dark restaurant or bar background with a few warm amber lights and minimal visible interior detail".
6. If the glass rim is misplaced: append "the wine glass rim cuts across the face, overlapping the nose and lower eye area".
7. If reflections vanish: append "bright specular reflections on the wine glass, liquid surface, and eyeglass lenses".
8. If the black top changes: append "simple black sleeveless top visible beneath the neck and shoulders".
9. If labels vanish: append "bottom labels read Higgsfield SOUL 2 on the left and NANO-BANANA PRO on the right".
10. If the mood becomes luxury-editorial: append "casual flash-lit nightlife snapshot, intimate and spontaneous, not a polished campaign shoot".

A pose-reference style image showing a young woman squatting casually in front of a vintage green van on sunlit pavement. She wears a black sleeveless tank top, fitted black pants, and classic black high-top Converse sneakers, creating a clean minimalist outfit that makes the body posture easy to read. Her dark hair is tied into a high ponytail, and she wears round glasses and hoop earrings, giving the image a recognizable creator-style face while keeping the clothing simple. The main frame is centered on her symmetrical squat, with elbows relaxed near the knees and hands hanging naturally, making the pose feel approachable, casual, and useful for reference. In the upper-right corner, a small inset image shows a stylized illustrated version of a similar character in a related outfit and seated stance, with a red arrow pointing from the inset toward the live-action pose. This turns the composition into a transformation or inspiration graphic rather than a standard portrait. The vintage van behind her provides a strong color block and lifestyle backdrop, while the inset makes it clear the image is about translating a stylized reference into a real-world pose. Lighting is warm late-afternoon daylight, giving the skin and vehicle paint a soft golden tone. Emphasize realistic posture anatomy, canvas sneaker texture, black clothing simplicity, faded van paint, inset-image framing, and the tutorial-like feel of a real-photo adaptation from illustration. The final image should feel practical, stylish, and social-media ready, like a pose study or visual reference guide for recreating an illustrated character stance.

[Subject] A young woman and a golden retriever posed on top of an old car at sunset. The woman sits on the roof of the car in the center-right foreground, smiling upward toward the left with a joyful relaxed expression. She has fair skin, large round wire-frame glasses, long black hair pulled into a high ponytail, and silver hoop earrings. She wears a plain white short-sleeve T-shirt under blue denim overalls with rolled-up pant legs and white sneakers. One knee is bent up, the other leg folds across the roof naturally, with hands resting loosely near her knees. To her left sits a golden retriever with warm golden fur, upright posture, relaxed open mouth, and friendly expression. Keep exactly one woman and one dog.

[Environment] Outdoor open-air scene during golden-hour sunset. The sky fills most of the background with soft warm orange, peach, and pale gold clouds, no buildings or trees visible. The pair are seated on the roof of an older red-brown vintage car with worn paint, dark windshield frame, and rounded roofline visible in the lower half of the frame. This specific social-media cover image also includes an instructional overlay in the upper right: a small circular profile photo of the same woman, a red plus sign beneath it, a small rectangular anime-style reference image showing a girl and dog on a car, and a red curved arrow pointing down toward the realistic main scene. Large promotional text overlays appear across the lower middle: big white “COPIA”, smaller green “IMAGENES”, and yellow “NANO-BANANA PRO”. Keep these overlay elements because they are visibly part of the provided image.

[Composition/Camera] Vertical 4:5 or near-vertical social-media cover composition. The woman and dog are centered on the car roof, framed from full seated body and dog body upward with lots of sky around them. Camera angle is slightly low, looking up enough to make the subjects silhouette cleanly against the sky. The car hood and windshield occupy the lower foreground, while the sky dominates the upper two-thirds. The instructional inset cluster lives in the upper right corner and should remain separate from the main subjects. Preserve the poster-like social post layout with subject stack in the center and bold text overlay low in the frame.

[Lighting] Soft golden-hour sunlight from the right-front or low side direction, bathing skin, denim, dog fur, and the car roof in warm amber light. Shadows remain soft and gentle, with no harsh midday contrast. The sky should glow warmly behind the subjects, and the dog’s fur should catch rich honey highlights. Keep a dreamy sunset realism without losing facial clarity.

[Style/Rendering] Hyper-real lifestyle photography with social-media promotional layout. The main central scene should feel like a realistic cinematic sunset portrait, while the upper-right inset intentionally looks like a stylized anime or illustration reference card. Clean skin texture, realistic denim folds, detailed dog fur, warm nostalgic car texture, and smooth cloud gradients. The full frame should read as a creator tutorial cover image showing “reference image plus realistic result,” not as a plain untouched photograph and not as a magazine editorial.

[Detail constraints] Do not remove the tutorial overlay elements. Keep the circular profile image, plus sign, anime-style inset card, and red arrow in the upper right. Keep the large word “COPIA”, the green “IMAGENES” box, and “NANO-BANANA PRO” near the lower center. Maintain exactly one woman sitting on the car roof and one golden retriever at her left. Preserve the sunset sky, vintage red car, denim overalls, white T-shirt, glasses, ponytail, and warm golden-hour mood. Do not convert the scene into a clean landscape photo without graphic overlays.

Negative prompt: extra people, extra dogs, cat, city skyline, trees, grass field, indoor studio, cloudy gray sky, midday sun, dark night scene, modern SUV, convertible, leather fashion outfit, missing glasses, blonde hair, anime-only main scene, painterly sunset, warped dog anatomy, deformed hands, missing overlay card, missing text, random logos, watermark clutter.

Suggested parameters: aspect ratio 4:5 vertical; lens 35-50mm equivalent lifestyle framing; aperture look f/4 to f/5.6 for broad subject sharpness; steps 30-40; CFG/style guidance 6.5-8; sampler DPM++ 2M Karras or realistic lifestyle sampler; seed suggestion 145829367.

Delta prompt strategy:
1. If the tutorial overlay disappears: "add upper-right social-media tutorial overlay with profile icon, red plus sign, anime reference card, and curved red arrow"
2. If the main scene loses realism: "hyper-real sunset lifestyle photo of a woman and golden retriever on a vintage car roof"
3. If the dog drifts: "golden retriever sitting upright on the left side of the car roof, friendly open-mouth expression"
4. If the outfit changes: "white T-shirt under blue denim overalls, rolled cuffs, white sneakers"
5. If the sunset gets too intense: "soft warm golden-hour sky with peach and orange clouds, gentle contrast"
6. If the car becomes too modern: "old red vintage car with rounded roofline and worn paint"
7. If the pose changes too much: "young woman seated on the roof with one knee up, smiling upward with relaxed posture"
8. If the text layout is lost: "large white COPIA text, green IMAGENES label, yellow NANO-BANANA PRO text near the bottom center"
9. If the image becomes too editorial: "creator tutorial cover image, social-media promotional layout, not a luxury campaign"
10. If the inset looks realistic instead of illustrated: "small anime-style reference image in the upper right, clearly different from the main realistic scene"

[Subject] A side-by-side close-up comparison image showing the same young woman behind wet glass in two different render styles. She appears in her early 20s with fair skin, soft natural features, light green-hazel eyes, large round silver wire-frame glasses, hoop earrings, and dark brown to black hair pulled into a high ponytail. Her face is bare or nearly bare, with minimal makeup and a calm neutral expression. She wears a black high-neck top. In both panels she is framed very close to the camera, looking straight ahead through water-speckled glass.

[Environment] Two equal vertical portrait panels with a dark teal-black background and many water droplets or condensation spots across the foreground glass plane. The left panel is labeled “Higgsfield SOUL 2” with a small neon snake-like icon near the bottom. The right panel is labeled “NANO-BANANA PRO” with a colorful four-point star icon. The scene should feel like a moody rainy-window selfie or low-light glass-booth portrait, with the left side flatter and more flash-like, and the right side deeper and more realistic with cleaner depth and better lighting control.

[Composition/Camera] Tight face-dominant portrait in split-screen comparison format. The left panel crops the face larger and slightly off-center, emphasizing magnified glasses and a more compressed selfie feel. The right panel gives slightly more breathing room around the head and shoulders, keeping the face centered and the black top more visible. In both panels, water droplets on the glass remain a major foreground element and partially overlap the face and lenses. Maintain the two-panel A/B structure exactly.

[Lighting] Low-light cool-toned portrait lighting with strong reflections on wet glass and eyeglass lenses. The left panel should feel more frontal-flash or flatter direct illumination, making pores, reflections, and droplets feel harsher and more synthetic. The right panel should feel more balanced and dimensional, with the face emerging from a moody teal background and the droplet reflections integrating more naturally. Preserve bright highlights in the droplets and subtle shadows along the cheeks and neck.

[Style/Rendering] Hyperreal split-screen realism benchmark. Left side: slightly harsher, more artificial flash-selfie interpretation with less subtle modeling. Right side: more refined photoreal portrait with better skin realism, smoother glass interaction, and more coherent low-light atmosphere. The overall aesthetic should stay intimate, damp, moody, and minimal, not glamorous or heavily stylized.

[Detail constraints] Keep exactly one comparison image with two vertical panels, the same woman in both, the same glasses, ponytail, hoop earrings, black top, and rainy-glass effect. Preserve the heavy water droplet foreground and the dark background. Keep the bottom model labels and icons. Do not add extra props, city backgrounds, umbrellas, or colorful neon scenes.

Negative prompt: single panel, extra people, smiling selfie vibe, lipstick glamour look, dry glass, no droplets, outdoor street scene, bright daylight, colorful club lights, no glasses, no earrings, text removed, watermark, distorted eyes, blurred face, beauty-filter skin, fantasy sparkles, cyberpunk city, open mouth laugh, exaggerated makeup.

Suggested parameters: aspect ratio 4:5, dual-panel split, very close portrait lens or smartphone close-up feel, shallow to moderate depth of field with foreground droplet clarity, 28-36 steps, CFG/style strength 6-7, sampler DPM++ 2M Karras or equivalent, seed suggestion 180522644.

Delta prompt strategy:
1. If the rain-on-glass effect weakens: append "heavy foreground water droplets and condensation spots on the glass directly in front of the face".
2. If identity drifts between panels: append "the exact same woman appears in both panels with matching glasses, hoop earrings, ponytail, and eye color".
3. If the left side is not flatter: append "left panel feels more direct-flash, harsher, and slightly more synthetic in lighting and skin response".
4. If the right side is not more realistic: append "right panel shows more dimensional low-light realism, cleaner skin, and more natural droplet integration".
5. If the black top disappears: append "black high-neck top visible at the base of the portrait".
6. If the face crops wrong: append "tight face-dominant close-up portrait with glasses large in frame".
7. If the background gets busy: append "dark teal-black minimal background with no additional scenery".
8. If glasses reflections vanish: append "retain bright wet reflections and screen-like highlights on the round wire-frame lenses".
9. If the labels vanish: append "bottom labels read Higgsfield SOUL 2 on the left and NANO-BANANA PRO on the right".
10. If the mood turns glamorous: append "intimate rainy-window realism, subdued and moody, not polished beauty editorial".

GLOBAL LOCK: Subject is Natalia Dyer, an American actress with an oval face, high cheekbones, large expressive brown eyes, and fair skin with natural warmth. Her hair is dark brown, long, and wavy, styled into two thick, loose braids falling over her shoulders. She wears a dark, high-collared cloak/coat. Her expression is neutral, serene, and slightly melancholic, looking directly at the camera. The camera is a static Medium Close-Up (MCU) with a cinematic 35mm lens feel. High-fidelity skin textures and realistic lighting are mandatory.

[00:00–00:01]
Subject is centered in a grand, atmospheric gothic cathedral. Background features intricate stone arches and stained glass windows. Lighting: Misty, volumetric light beams (God rays) filter through the windows, creating a teal and orange contrast. Subject's face is softly lit by the ambient glow. Motion: Subtle dust motes dancing in the light beams.

[00:01–00:02]
Subject is centered in a vast golden hour meadow. Background features tall, dry grass and a distant horizon under a setting sun. Lighting: Warm, intense amber backlighting creating a soft rim light on her hair and cloak. A subtle lens flare peeks from the corner. Motion: Very slight swaying of the grass in the background.

[00:02–00:03]
Subject is centered in a dense autumn forest. Background is filled with vibrant orange and red maple leaves. Lighting: Dappled sunlight filtering through the canopy, creating soft patches of light on her face. Shallow depth of field with a creamy bokeh effect on the leaves. Motion: A few leaves slowly falling in the background.

NEGATIVE PROMPT: 
Facial distortion, changing eye color, changing hair style, inconsistent facial features, cartoonish look, plastic skin, extra limbs, blurry face, text, watermark, logo, flickering lighting, sudden jumps in subject position, robotic movement, oversaturated colors, low resolution.

[Subject]
A vertical two-panel comparison image showing the same young adult woman in an ethereal underwater ballet-inspired scene. She has light skin, long dark hair, hoop earrings, and delicate feminine features. She wears a pale pink fitted crop top and a voluminous translucent pink tulle skirt like a ballet tutu. In the left panel she is closer to the water surface with hair falling naturally and eyes lowered, seated or kneeling softly in the pool. In the right panel she is more fully submerged, hair floating outward, facing camera with a calm expression, the skirt billowing underwater.

[Environment]
Fantasy pool environment with clear turquoise water, pink lotus flowers floating around the subject, elegant crystal chandeliers partially hanging into the scene, bright caustic water patterns, and dreamy pastel reflections. The water surface is visible near the top of the frame, with shimmering distortion and reflective ripples. Keep the space surreal but refined, like a luxury underwater editorial set.

[Composition/Camera]
Two equal vertical comparison panels with rounded corners and a thin divider. Left panel emphasizes surface-level immersion and a softer pose. Right panel emphasizes underwater depth and floating movement. Bottom labels identify the models: "Higgsfield SOUL 2" on the left with neon-green icon and "NANO-BANANA PRO" on the right with multicolor diamond icon. Maintain a mobile-friendly comparison-card structure.

[Lighting]
Bright, clean underwater lighting with soft pastel highlights, visible light caustics across skin and fabric, subtle chandelier sparkle, and even high-key illumination. The mood should be luminous and airy rather than dark or murky.

[Style/Rendering]
Photoreal fantasy editorial comparison graphic with premium underwater styling, soft romantic palette, convincing water refraction, realistic floating hair, delicate tulle texture, and polished dreamlike beauty. The scene should feel like a creator comparison post testing how well different models handle complex underwater elegance.

[Detail constraints]
do not remove the split-screen layout, preserve the same woman in both panels, keep the pink outfit, lotus flowers, chandelier elements, underwater caustics, and the model labels at the bottom. Water behavior, hair float, and fabric diffusion must remain central.

[Negative prompt]
dark murky water, beach scene, fish, scuba gear, missing split layout, no flowers, no chandeliers, wrong outfit color, heavy gothic mood, anime, cartoon, deformed legs, tangled hair blobs, broken hands, text missing, different person between panels, dry fabric

[Suggested parameters]
- aspect ratio: 4:5 vertical overall
- lens/focal length: 35mm to 50mm equivalent underwater editorial feel
- depth of field: medium
- steps: 32-42
- CFG/style strength: 5.5-7.0
- sampler: DPM++ 2M Karras or Euler a
- seed suggestion: 22851479

[Delta prompt strategy]
1. If the underwater feeling weakens, append: "clear submerged scene with visible water surface and light caustics"
2. If the hair does not float naturally, append: "long dark hair drifting weightlessly through the water"
3. If the skirt loses volume, append: "soft translucent pink tulle skirt blooming outward underwater"
4. If the lotuses disappear, append: "pink lotus flowers floating around the subject in the pool"
5. If the chandeliers vanish, append: "crystal chandelier elements hanging elegantly into the scene"
6. If the split comparison layout disappears, append: "two side-by-side comparison panels with rounded corners"
7. If the panels stop matching the same character, append: "the same woman appears in both panels with matching face and earrings"
8. If the scene becomes too dark, append: "bright pastel turquoise water with clean high-key underwater light"
9. If the labels are missing, append: "bottom labels for Higgsfield SOUL 2 and NANO-BANANA PRO"
10. If fantasy takes over realism, append: "photoreal underwater editorial, premium creator test graphic, not cartoon fantasy"

[Subject] Side-by-side comparison image with two vertical portrait panels of the same tattooed young woman interpreted in two different model styles. Left panel: realistic version, early 20s, feminine presentation, light olive skin, long straight black hair in a high ponytail with small orange flowers placed through the hair, thin round metal glasses, small septum ring, large yellow drop earrings, layered silver necklaces, black tank top, floral chest and shoulder tattoos, and puckered kiss-face lips. Right panel: stylized or beautified version of the same woman, still wearing thin round glasses, yellow earrings, layered necklaces, and visible tattoos, but with turquoise-blue hair styled into two high buns with long front sections falling down. She has a softer smile, smoother face, slightly more illustrative or beautified finish, gray sleeveless top, and the same general identity translated into a more stylized rendering. Bottom labels identify the left as "NANO-BANANA PRO" and the right as "NANO-BANANA".
[Environment] Minimal studio-style portrait background with soft beige or warm neutral backdrop, no environmental props. This image is a direct comparison poster illustrating the difference between a more realistic output and a more stylized output while preserving the same character identity markers.
[Composition/Camera] Vertical 3:4 canvas divided into two equal rounded-corner columns with a narrow divider. Both portraits are chest-up, centered, front-facing, and tightly cropped. Subject fills most of each panel. The left panel emphasizes realism and sharper photographic fidelity. The right panel emphasizes a cleaner, more beautified, somewhat illustrative aesthetic. Composition must remain highly symmetrical and consistent so viewers can compare style drift, identity retention, and detail fidelity.
[Lighting] Soft frontal portrait lighting with balanced illumination on both faces, gentle catchlights in the eyes, subtle reflections on glasses, and minimal shadow. Light should be neutral and flattering, allowing differences in texture realism and rendering style to show naturally. Skin and tattoos must remain readable in both panels.
[Style/Rendering] Comparison poster for AI portrait generation quality. Left panel should feel photorealistic, detailed, and grounded. Right panel should feel smoother, more stylized, and slightly digital or illustrated while retaining realism-adjacent portrait structure. Both images should remain polished and attractive, but the contrast between realism and stylization should be obvious. No extra poster graphics beyond the bottom labels.
[Detail constraints] Keep exactly two portrait panels, preserve key identity markers across both sides: glasses, yellow earrings, layered necklaces, tattoos, youthful female face, and centered framing. Maintain bottom labels "NANO-BANANA PRO" on the left and "NANO-BANANA" on the right. Do not add side props, cluttered backgrounds, extra people, or text beyond the labels. The comparison is about fidelity versus stylization, not about different scenes.

Negative prompt: single panel only, missing tattoos, missing glasses, no yellow earrings, no labels, cluttered background, wildly different identities between panels, cartoon exaggeration, anime eyes, painterly texture, distorted face, low-detail jewelry, no septum ring, extra objects, watermark.

Suggested parameters: aspect ratio 3:4, 70-85mm portrait feel, shallow depth of field, 28-36 steps, CFG/style strength 6.5-8, sampler DPM++ 2M Karras or equivalent, seed around 465328.

Delta prompt strategy:
1. If the split-screen disappears: add "two equal vertical comparison panels with a narrow divider and matched crop".
2. If identity diverges too much: add "same woman, same facial structure, same jewelry, same tattoos across both panels".
3. If the left panel is not realistic enough: add "left panel photorealistic, crisp skin detail, natural hair texture, grounded realism".
4. If the right panel is not stylized enough: add "right panel smoother, more beautified, slightly illustrative finish with turquoise twin buns".
5. If the tattoos fade: add "clear floral tattoos across chest, neck, and shoulders visible in both panels".
6. If the earrings disappear: add "large yellow drop earrings clearly visible on both sides".
7. If the orange flowers on the left vanish: add "small orange flowers tucked into the black ponytail on the left panel".
8. If the labels disappear: add "bottom left label NANO-BANANA PRO and bottom right label NANO-BANANA in bold white text".
9. If the background gets busy: add "plain warm neutral portrait backdrop with no props or decor".
10. If the glasses distort: add "thin round metal eyeglasses with natural reflections and correct lens proportions".