Over the past week, the AI circle continued the fairy fight. GPT Images 1.5. Killed a gun
I'M JUST SIFTING FOR YOU -- THE MOST AMAZING, MOST ACCOMPLISHED GROUP OF 17 IN THE WORLD (polite) my daughterPrompt word.
It's all hard food。
For each set of hints, a map was obtained using Nano-banana Pro and GPT Images 1.5。
Whether you like spicy, designy, or real, this 17 must have your food。
It's not much to say. It's a good idea to start with。
Nano and GPT, I'm off the grid
17 Selected Cases
01. Hollywood scouts
I just shared it yesterday in The Five Minutes Hollywood Spectrum, and it was fun to make a series。
But this hint GPT can easily be copyrighted and refused; instead, Nano-banana is particularly smooth and can come up with anything。
- Prompt: The above image is the sole facial reference, 100% precisely re-constructing the person's facial bone structure, skin texture, hair and dementia. The subject of the upload remains the same, and the clothes remain the same. Leonardo Dicaprio: Shows his face, hair and dementia during the film "Screeching Dreamspace" (2010). Interact status: The two people are wearing their own costumes, and they're taking a photo of an improvising and joyous scene in the middle of a movie. Camera: Simulation of iPhone patterns, with natural depth, softness and true colour expression. Chart: 9:16 vertical. Using a living, unruly capture, the two subjects are located at the centre of the picture or at the centre of a slight deviation, as is the case in the background of a hand-shot by a friend. The bystander's perspective, as if you were standing on the set, taking this picture for them at close range. Main light source: fully following the environmental light logic of the selected film scene (e.g., if it's the "City Collapse" scene, it's the light of a distorted dream scene in Paris street buildings). Auxiliary light: Signs of spot-based refilling equipment, such as a little photocopy of a soft and light-filled or light-stand from a reflection panel, are visible, provided that they are fully integrated with natural light and do not show a sudden. Color style: Modern, high-tech cold, blue and gray. Uploaded person clothing: requests to keep the person's name on the list, and to keep the clothes on Leonardo Dicaprio's dress: based on the classic scene of the selected film, wearing a suit/business dress, which is fully compatible with the age and role of the play. The movie was briefly interrupted. The staff members shouted, "Come and take a photo!" The two men turned to the camera, and their faces were relaxed, their gestures were random, and the pleasures and consecrations of the break were felt. Classic scenes: Environment: Paris Street Building Collapse Distorted Dreams. Key elements: The scenes of the film need to be clearly visible in the picture: film cameras, light frames, radio microphone poles, orbital vehicles, staff for partial lenses, and auxiliary devices for the site layout to create an environment of “spots in progress”. Image style and detail: The combination of overwriting and iPhone's life flash. Details: Very high skin quality, natural texture and pores. There are real wear and use marks on clothing and scene prop. The scenery: warm, nostalgic, human. It's like a photo of the hidden behind the 2010 Dreamspace scene that was discovered by chance, and it's a light moment of happiness. An overwritten picture, which makes you feel like you broke into the scene of Dreamland 2010 and caught a natural, joyful and precious behind-the-scenes photo of Leonardo Dicaprio and your friends in a costume at some point。
Google Nano-banana Pro:
GPT images 1.5:
02. Advertisements
Sending a map that generates countless advertising parts of the characters。
Some people say to send pictures of the people they love, and the drawings that you want to make
- Prompt:“Based on Figure 1. It's a LMO Ic-a-style high-level photo. With a high angle view, the camera focuses on an amazing Japanese film actress. She's wearing a black windshirt, and her face and face are exactly the same as "Reference Photo Person." She's standing on numerous black and white fashion billboards on the ground, and because of her perspective, she can see that all the billboard models on the ground are all her own different styles, as if she were standing on her countless bilocations."
I used the original:
Google Nano-banana Pro:
GPT images 1.5:
03. Grand Tang and Paris
The Great Tang Mension crashed into modern Paris。
You have to upload your face and light
- Phrasing: The phrasing is out of order and is obtained as before. A 9:16 vertical version of the photo of Grand Tang and Paris: 100% returned to five officials and bones, using my uploaded FACE_REF as a reference; a heavy-work dress in the direction of the King of the Palace of Tang Dynasty, in the direction of the Great Princess of Tang Dynasty, in the form of a Trocardero Square under Eiffel Tower in Paris or on the side of the Seine River. The image is a photo, plastic and cover design at COSMO fashion magazine level, maintaining the consistent design style of the magazine (designed in English). FACE_REF: Five official reference maps (identity lock) of the person that I uploaded (ABSOLTE PRIORITY) 100% was restored to FACE_REF 's bone: face, face, eye length, nose, lip, jaw, bone structure fully consistent and non-floating. Real skin quality: The light texture and pores are visible and do not overwhelm the skin and net. Image of adult females. Half-cover: chest to waist, head to top, and top to top embroidery is the clearest, eye to eye to eye (push) B. Cover all over the body: from head to foot, a layer of dress and dress tattoos, steel structure of the iron tower is much stronger (optional) (whether A/B: must keep a “cover map” and have white space for layout, but do not tear paper/hand-painting effects. Paris Eiffel Tower: The classic steel tower. Weather is selected for one (random): 1) Paris has a dark rain (wet on the ground, light and high-sweet grays, high-fashion) 2) twilight (twilight after sunset, golden light, romantic atmosphere) has a clean background: no tourists, no modern signs, no cell phone UI, no watered subtitles, no fences. IGHTING & CAMERA PHOTOGRAPHY SENSION (JAMMAGE) Light: Natural light (emphasis on environmental light) + film-grade light; gold coats with real light but not explosive; silk light is clear. Colours: Advanced film sense, strong visual tension formed by a "red and green gold" color in French ash/blue gray and Dondee's clothing; whole body clean, permeable, and of high quality. The target beauty is: "Golden, bouquets, bouquets, tiaras, tiaras, tiaras, tiaras, big red, big green gold, soar." The inner layer: a tiara (brain-screeching), a neck-screeching and upper thorax (an open wind) Middle layer: a short shirt (a straight collar), transparent and light (ro/span) fabric outer layer: a large t-shirt (extremic sleeves, tows), open like wings Decoration: a long t-shirt (long ribbons), t-shirts, t-shirts, t-shirts, t-shirts with wind, t-shirts, large skirts, multi-coloured or t-shirts) Flour and process (a large hand-screas must “see” main material: silk, knitted, gold veils (lightly and thickly combined) main process (scracked/string), a golden wire embolling, and a large (tweaked) face (tweaked) or a big (tweed) bellowed bellowed) The golden comb (the golden comb on the hairbrush), the walk (the shaking) and the gold dung ear: the precious golden jade [COLOR MATRIX | colour matrix (must be implemented, and must be "heavy Tangwind")] (readjusted to the Tang Dynasty popular color: red and green, golden and brilliant) randomly select a set of one from the following three main types of "main colour" as the [bottom colour/large area master colour]. Scheme G (Big / Green ): Base colour: green green / tan (the usual tan) embroidery / laceration: Greek, gold, orange yellow, white Scheme A (Oyster Red): Bottom colour: police red / pomegranate (the most classic Tang Fung) embroidery / laceration: gold , green , curry , blue , purple / Scheme R (the purple bottom): Bottom colour : purple / grape (high) embroidery / laceration: gold , red , grass green , silver [Colour enforcement rules] [POSE Cover Posture (confidence, relief, Queenhood)] Half body: micro-eave head, eyelids, one hand caressing the hair, or a cape. With both arms open, the t-shirts and coats open, the clothes feel, the air is open. One 9:16 vertically written magazine cover-level photographs Random: Half body or All body (A/B2: 1) Random: Rain or Twilight (weather: 1) Random: Green / Police Red / Purple (Colour Matrix: 1) Shape: Direction of the Tang Dynasty Court dress + Omnibus/ Gold Shaft face + Full chest sleeve + Seal
Google Nano-banana Pro:
GPT images 1.5:
Women in the Forbidden City
Upload a person's face as a benchmark to generate a picture, with the following hint:
Prompt:A high-level audience, with a VOGUE team, is in a low-level company, with a high-level professional, with a high-level professional.
Chinese tip: A high-end fashion photo project with a VOGUE cover. The scene is set in the snow-covered ludicrous city halls, where the red poles are arranged. At the center of the picture is a beautiful young East Asian woman with bright and bright skin, fine as china, closed lips and a very restrained, implicit and elegant smile. She is well-versed, with deep eyes and strong and elegant. She wears a beautiful golden phoenix with pearls on her head; she wears a red, blue, embroidered robe with fine yunjian on her shoulder. She holds a traditional round silk fan. Snowflakes fall softly. The light is a soft, radiant natural winter light, and the overall atmosphere has a sense of film and luxury. Top is the beige text "VOGUE" with a smaller text "MING DYNASTY INSPIRED" on the left. Image symmetry, 4K resolution, overwritten style。
Google Nano-banana Pro:
GPT images 1.5:
Photos with star posters
UPLOAD A PICTURE OF A PERSON, THE STAR'S NAME IN THE HINT. FAMOUS ACTOR AI KNOWS EACH OTHER。
- Prompt:High end street, 3:4 vertical, all in. The dark clouds, the light, the 8K paints, the sense of substance. Foreground: Users (e.g., figure 1) stand half a metre from the wall, with their head to toe full exposed, their shoes visible, and their natural shadows fall behind the wall. Users (e.g., figure 1) carry their fingers behind their backs with big posters, laugh on their heads, interact with each other, move around and live around, and are in no way inflexible. Background: Architectural scale ads, Tom Holland giant shoulder image, glamorous smiles, plaster micro-films, reflections, loose tight marks are visible, real heavy, non-electronic screens. There are no clippings, no low clearance, no false stickers, no cartoons。
Google Nano-banana Pro:
GPT images 1.5:
The gestures and expressions in the hints can change。
Of course, we can switch. Put the star outside and put the pictures you uploaded in the poster。
Women Titans
You can turn anything into a giant, including a beautiful woman, into a small country
- Prompt:THE ONLY BLUE TEXT AVAILABLE (FIGURE 1) IS THE PHYSICAL ONE, WHICH IS MAGNIFIED INTO A GIANT CORE: SHAPE, COLOUR, MATERIAL, RATIO 1:1 AND SURFACE FIBERS IN MILLIMETRES. BETWEEN 10 AND 30 MICRO-WORKERS TRAVEL IN A BUSY AND ORDERLY MANNER — SCAFFOLDING, CLIMBING, POLISHING, PAINTING, CLEANING, MEASURING, COMMANDING CRANES. MAGNIFICENTS (CHART 1) ARE PLACED IN MICROREALISTIC STAGES SUCH AS DESKTOPS/STANDS/METALLIC DISCS, TOOL PARTS SCATTERED, SHALLOW, MOVING-AXIS, FILM LIGHT, 8K SUPER-CLEAR, INDUSTRIAL-GRADE. THE ATMOSPHERE: PRECISION, CRAFTSMANSHIP, SCALE SHOCK, ZERO CARTOONS, ZERO WORDS, ZERO ILLUSIONS。
Google Nano-banana Pro:
GPT images 1.5:
7. Bathing for beautiful women
- Prompt:WARM BATHROOMS, STONE WALLS, AN ANGLE IN THE TUB. 9:16 MIRROR SELF-PORTRAIT: YOUNG WOMEN (E.G., FIGURE 1), XXX SPECTRUM, LIGHT MAKEUP. WHITE TOWELS ARE COVERED, ONE HAND ON THE SIDE OF THE CHEST, ONE HAND ON THE WHITE SHELL PHONE COVERING HIS FACE, AND ONLY THE EYEBROWS ARE RELAXED. THE LITTLE NECKLACE, THE BRACELET, THE RING. THE TEXT OF THE CELL PHONE SCREEN IS READABLE, WITHOUT MALFORMATION. THE SKIN OF THE SHOULDER, ARM, UPPER CHEST AND LEG IS SOFTLY LIGHTED BY WHEAT, AND THE TOWEL IS CLEAR. IT'S A LIGHT-HOT SKIN, A SHALLOW VIEW, A PRIVATE AND QUIET ATMOSPHERE, LIKE A QUICK SHOWER. NO EXPOSURE, NO SPARE LIMBS, NO FLASHES, NO WORDS。
Google Nano-banana Pro:
GPT images 1.5:
08. NINE-GAUGE + NUDITY 3D
Prompt:
> CREATE A 3X3 GRID LAYOUT POSTER CONTAINING 9 DIFFERENT PARTS AND SUPERSING A STANDING NUDITY 3D IMAGE OF WOMEN:
The unity of character requires that women in all scenes must be the same person, with the same facial characteristics, hair and figure. Asian women with straight, dark hair, elegance, health-like body size。
Layout structure:
– Bottom: 3 columns x 3 row grid
– Equal size of each cell
– clean white border between cells (2px)
– TOP LEVEL: A STANDING 3D RENDERING FEMALE IMAGES SUPERIMPOSE TO THE CENTER OF THE ENTIRE GRID
Line 1 (top line – three separate scenes, same woman):
[cells 1-1] Left: The same elegant Asian woman standing by a calm high-mountain lake, wearing a brown-coloured jacket with white-coloured broad-legged pants, gently touching the jacket collar and standing side by side. In the background, the magnificent snow mountains are resurfaced on the lake, the soft and blue light of the twilight, the light of the lake like a mirror, the light of the sky, the cold of the whole color, the blue and green, the quiet dream, the advanced fashion photography style, the film-scale image。
In [cells 1-2]: The same woman standing on the reef, wearing a tattered, pink dress (the dress follows the wind), looking back or looking far away, and displaying elegance. The background is a spectacular sun-down sea view, the sun slows down to sea level, the sky becomes orange pink, the waves lighten the rocks, the sea full of light, warm-colored golden hour light, romantic dreams, the cover of fashion magazines。
[CELLS 1-3] RIGHT: THE SAME WOMAN STANDING IN A PINE FOREST WITH SUNLIGHT PIERCING, WEARING A WHITE-WHITE V-COLOURED BROAD-LEGGED PANTS, ONE HAND TOUCHING HER HAIR, OR A NATURAL LEANING, AND A LIGHT AND ELEGANT GESTURE. THE BACKGROUND IS A LARGE PINE FOREST, WHERE THE GOLDEN SUN OF THE TWILIGHT SLIPS IN FROM BETWEEN THE TREES, CREATING A DREAM BEAM, WITH GREEN BEAKS ON THE GROUND, LUMBERS, WARM AND GOLDEN, AND QUIET。
—
Lines 2 and 3 (all three columns are horizontally cut in full photographs):
Key: Each column in rows 2 and 3 is a full picture of two parts, top and bottom, cut at the clothing boundary or horizontal line. All characters must be the same woman。
—
[ Cells 2-1, 3-1] Left column – Full picture of the same woman (horizontal cutting):
** Key: This is the same complete photo of the person, cutting up and down at the lower waist belt position (the cortex and the cortex of the pants), which must be perfectly aligned. The cutting line is located at the clothing boundary and is not on bare skin. The characters and background are continuous. ** **
** Full photo description:**
– Photo of the same woman's body from head to foot
– Background: Sunshine forest clearings, warm natural light from the side, defunct green plants and gold spots in the background, outdoor natural environment
– Perspectives: Half-side angles (3/4-side angles), displaying female elegance in the contours and body curves
– CLOTHING: UPPER HALF WEARING A SHORT, YELLOW LETTERED UMBILICAL UMBILICAL VEST (NOMBRE3), TIGHT DESIGN, ABDOMEN AND ABDOMEN; LOWER HALF WEARING A LIGHT BLUE WAIST TIGHT JEANS, FROM THE WAIST TO THE FOOT, AND CUT
– POSITIONS: HANDS WITH GRACE TO LIFT THEIR HAIR OR TO LIFT THEIR HEADS, ARMS WITH NATURAL BENDS, BODIES WITH A SLIGHT STAIRWELL, AND FINE CURVES FROM HEAD TO FOOT
– Overall: a healthy body of exercise, with a full display of the waist, a thin line
** Excision description: **
– Line 2 (first half): full picture from the head to the first half of the position of the lower waist belt. It consists of the head, hands, arms, neck, chest, a whole tan short vest (including under the vest), abdominal and abdominal skin. The background is the upper half of the vandalized green plants and golden light spots. ** The cutting line is at the borderline of the lower waist belts under the vest.** Zenium
– Line 3 (second part): Full picture from the position of the lower belt to the lower part of the foot. It consists of a lower waist belt (starting from the top of the belt), hip, leg, foot. The background is the lower half of the defunct green plants and golden light spots. ** The cutting line is on the top of the lower waist belt**. Zenium
** Alignment requirement: 2-1 and 3-1 must be cut perfectly in the lower waist belt position (under the vest and at the end of the belt), like a knife to cut the full photo horizontally. The cutting line is at the line of clothing (the top edge of the light blue belt) and must be seamlessly connected in the upper and lower pairs, with a perfect alignment of the pants on the waist, a consistent body, a consistent background, a consistent light, and no error or colour. ** **
—
[ Cells 2-2, 3-2] Middle row – Full beach sunset photographs (horizontal cutting):
** Key: This is the same complete picture of the sunset of the seaside, cutting up and down in two parts of the horizontal line at sea level or in the middle of the sea water, which must be perfectly aligned. **
** Full photo description:**
A full picture of the sunset of the seaside, from top to bottom, contains: the evening sky (the orange pink changes gradually); the dark sea (dark blue/dark blue, near the dark ocean at night, the sea surface is calm or has a slight ripple); and the light sand beach (light rice or gold). Same beach and sunset atmosphere as line 1 1-2 scene. It's a beautiful place to live in。
** Cut instructions:**
– Line 2 (first half): The first half of the full picture, from top sky to near sea level. It contains the evening sky (an orange pink gradient) and the upper half of the dark sea (dark blue/dark blue)。
– Line 3 (second half): The second half of the full picture, from sea level to the bottom of the beach. It contains the lower half of the dark seawater and light sand beaches。
** Alignment requirement: 2-2 and 3-2 must be cut perfectly at the level of the sea level or in the middle of the seawater, as if the whole view of the scene had been cut open with a knife, the upper and lower two should be seamlessly connected, and the sky-water-sand beach transition was smooth, with no error or colour difference. The colour of the seawater on the right edge has gradually become darker, with a natural dark transition to the right green plant background. **
NOTE: THIS AREA IS LARGELY BLOCKED BY STANDING 3D IMAGES OF WOMEN
—
[ Cells 2-3, 3-3] Right column – Full picture of the same woman (horizontal cutting):
** Key: This is the same complete photo of the person, cutting up and down in two parts at the same level of the belt position (under the vest and the belt). Cutting lines are at the clothing boundary (white or light grey belt position). The characters and background are continuous. ** **
** Full photo description:**
– Photo of the same woman's body from head to foot
– Background: The outdoor natural environment, soft and daylight, a defunct green plant and tree background (like a park or forest path), natural light irradiated from the side, a warm and bright atmosphere
– Perspectives: right-hand angles, showing women's right profile
– Clothing: upper half with a dark gray or black sleeveless sports vest; lower half with black sports tights with light gray or white broad belts on the waist
– Position: Stand side by side, to the right, to the right side of the picture, to the right hand, to the right hand, to the right hand, to the right hand, to the right hand, to the left hand, to the left hand, to the left. Head as arms move naturally back or back, face to the right (possibly closed eyes or looking up), legs to stand together, and body to stand straight
– Overall: a healthy body of motion, with a beautiful body curve from head to foot on the right side
** Excision description: **
– Line 2 (first half): Full picture from the head to the first half of the belt position. Includes head, right arm stretch, shoulder neck, chest, whole gray/black sports vest, waist. The context is the upper half of the vandalized green plants and natural light. ** The cutting line is located at the intersection of the exercise belt (slight gray or white broad belt) under the exercise vest.** Zenium
– Line 3 (second part): Full photographs range from the position of the belt to the lower part of the foot. It consists of a belt (slight gray or white, starting from the top of the belt), hip, leg, foot. The background is a defunct green plant and the lower half of natural light (possibly ground roads or grasslands). ** The cutting line is on the top edge of the belt of exercise pants**. Zenium
** Alignment requirement: 2-3 and 3-3 must be perfectly cut in the position of the belt of the sports pants (under the sports vest at the border with the belt), like a knife to cut open the entire picture. The cutting line is located on the line of clothing (the top edge of the light gray or white broad belt) and must be seamlessly connected with the upper and lower two, with a perfect alignment of the waist belt, the person ' s body, the background, the light, and the light. ** **
—
STAND 3D LAYER (MOST IMPORTANT):
HIGH-QUALITY 3D IMAGE OF THE SAME WOMAN
Position: Stand at the center of the 9th Palace, with two feet on the bottom grid line (margin at the bottom of line 3), and the body stretches up from the bottom to cover the 2nd and 3rd column areas and parts of the perimeter
Style: Superrealism 3D digital rendering, C4D or Blender advanced rendering, sexy fashion
Clothing:
– Clothes: a grey-coloured scheme with no cuffbands, cut-up, low-back design
– Lower: light blue super-short jeans, white for old effects, clear in the back pocket sutures, with a very short length showing most of the long legs
– Shoes: White sneakers with white socks
Position:
– Stand on your side, with a 3/4 side angle, with a back to the camera
– A slight look back on the head, with a side or half face
– THE BODY IS SLIGHTLY SKEWED TO THE SIDE, CREATING A SEXY S CURVE
– Slightly backward in the buttocks, with emphasis on the body curve
– A hand is naturally placed on the side of the body or on the waist of the buttocks
– The two legs are standing on their feet, and their feet are standing on the ground
– Long hair falls naturally on the shoulder and back
– sexy, undoing eye-killing pose
3D EFFECT FEATURE:
– Strong 3D shadows are projected on the background grid, with clear shadows on the ground from both feet
– Visible depth effect, clear character, a slightly less nuanced background
– The edge of a person has a soft luminous or glowing effect
– SKIN AND CLOTHING MATERIAL WITH 3D RETROFITTED HIGH-QUALITY SENSES (HIGH LIGHT, REFLECTION, SUBSURFACE DISPERSION)
– There are floating geometric elements, particles or light around
– SHADOW: 3D CHARACTER ENTITY IS CLEAR, PROJECTING A NATURAL SOFT SHADOW, ESPECIALLY WITH A CLEAR SENSE OF PROXIMITY BETWEEN FEET AND GROUND, AND THE PERSON IS LIKE A REAL PERSON STANDING ON A GRID
—
Technical requirements:
– All cells must be perfect squares and equal in size
– All the scenes must be the same woman, with facial features, hair, and a perfect fit
– Three scenes in line 1: super-high-level graphics, homogeneity, film-class photo effect
– All three columns (2-1/3-1, 2-2/3-2, 2-3/3-3): complete photographs cut at the level of the clothing boundary or horizontal line, with the cutting line at the edge of the dress (in the position of the belt) and the upper and lower pairs must be as well aligned as the slashed photographs
– All three rows of background must be continuous, the body must be consistent, the light must be consistent
– 3D STAND-BY IMAGES MUST HAVE A STRONG SENSE OF SPATIAL HIERARCHY, LANDSCAPE AND SEXY FASHION
– Shadows, dizziness, depths are natural
– SUPER-CLEAR FOCUS, 4K MASS
– Concordance of the overall tone, modern fashion
Google Nano-banana Pro:
GPT images 1.5:
Extremely accurate
Extreme clarity, photographic sense, for real or product presentation。
It can be uploaded or not provided。
Prompt:
> I. Imaging theme
> 1 Type of person: A portrait of a young female peri-person
> 2 Posture perspective: side eyes, shoulder backs to the outside side of the lens, slightly looking to the lens in the head
> 3 View range: head to top back and shoulder, person on the main picture, small background space
> 4 image scale: vertical, slightly left-handed, eyes near the top three-point line
Two, five officials and faces
1 Emoticon: cold restraint, a little alienation, no teeth
2 Eye characteristics: large eye ratio, clear eyelashes, eye focus lens, small front and soft shadow at the moment
3 Lip characteristics: full of water, soft edge, visible light and Jell-O
4 Skin characteristics: extremely subtle and smooth, weak pores, whole white and clear
III. Fashion style
1 Bottom make-up: a combination of high-intensity and soft fog, with a local high natural light
2 Mons: pink dizziness from the cheekbone to the present, light but visible
3 Eye make-up: close eye lines and natural deepening of the end of the eye, light brown ash transition, whole clean
4 Lip make-up: Gradual transformation of pink bean sand to light roses, supermutation of light
IV. Hairstyles and accessories
1 Hair type: dark, straight hair, low pony tail or low end, hair thawing natural
2 Liu Hai: Medium, thin hair lined up close to the cheek
3 Glasses: metal frame transparency lenses, elliptical lenses, metal decorations on the leg
4 Earmarks: small ear nails, low-key suffix
V. Details of clothing and skin
1 Clothes: red shoulder belt, back-sliding design, shoulder belt visible on the right side of the picture
2 Skin luminous: soft and reflective on the back of the shoulder, with a delicate creamy muscle feeling
VI. LIGHT AND FIND
1 Main light direction: soft light up front left, large coverage, very soft shadow
2 Filling: weak light in front, low contrasting, almost no hard shadow in the face
3 High light control: small area bright light on the nose, upper lip, lower lip and edge of glasses
4 Total tone: low contrast, soft fog, delicate layers, dark and dark
Background and environment
1 Background: Pure, single-colour light ash slowly transforms the wall, non-textual and impurity
2 Deep: strong background or condensed nature, making the profile of the person more prominent
3 Culture: clean, quiet and mildly sweet
VIII. IMAGE OF IMAGES AND IMAGES
1 Focal perception: medium-long focus view condensed, with no exaggeration of face
2 Deep expression: clear eyes, clear debilitated ears and background
3 Clarity: the edges of the five officials are clear but not sharp, and the whole body is soft
Noise: almost no particles, with a clean commercial sense of humanity
IX. ELEMENTS FOR LAUNCHATION
1 Skin grinding: strong but natural, with a small number of facial structures and shadow transitions
2 Colours: whole body is cold white, cheeks and lips are pink, red shoulder is saturated but not excessive
3 Optimus: light light spreads more round
4 Uniformity: control contrasting and sharpening, maintaining permeable, soft and clean image consistency
X. Inverted references
1 Image problem: excessive sharpening, contrasting, hard shadows, context disorder, visible noise particles
2 Personal problems: exaggerating faces, laughter, visible rough-cut skin, full of oil, deformed faces
3 Style deviations: heavy stage make-up, strong retro-colour, complex clothing, overstretching, strong environmental light interference
Google Nano-banana Pro:
It's very clear
by the way, gemini makes the drawings, saves the photo dots as vague, and clicks on the top right download button for the high resolution
GPT images 1.5:
And the same bottom figure that Nano-Banana uses, GPT makes new figures that are very consistent。
10. Fashion films
Prompt:
I. Overall objective
> 1 Functional description: Create a unique fashion person adhesion based on the person reference map uploaded by the user, drawing a person ' s style, make-up style, light and colour features。
> 2 Difference requirements: to retain the five-office structure of the reference person and its overall quality, not to go straight to the original map, and to make creative changes in the same style system。
II. Input reference chart requirements
1 Reference: A clear picture of the face or two-thirds of the person on the side of the face can be seen in the atmosphere of five officials, skin and broad light。
2 Excerpt points: read colour, makeup, main color, light direction, character and clothing colour from reference maps。
III. Pictures and output forms
1 Main image: Generates a person ' s face ultra-nose or peri-view, the head and shoulder occupy most of the space, and the view can refer to the original image ' s aerobic design。
2 Spelling structure: Adding more than one small puzzle to the right or to the corner, which can include side-face feature, hand feature, local make-up feature, etc., to create a magazine layout。
3 Background approach: Keep the background simple and clean, with a preponderant transition of pure colour or light light, highlighting the person and makeup。
Text and graphic design
1 Title text: The title of a large magazine is placed above the image, with the power of simplicity and the ability to remain in or above the same
2 Subheadings and small words: add a few lines of small words to the side or at the bottom of the picture, which can be the subject word, the name of the person, the date, the organization, the number of words
3 Geometric Graphics: Add a thin line, narrow rectangle or semi-transparent piece around the person, use it in conjunction with text and increase the sense of plane composition
V. CURRENT INTEGRATION RULES
1 Colour inheritance: whole-tone, dark contrast, cooler and heated as much as possible to fit the style of the reference map, the main color being from the most visible makeup or dress colour in the reference map。
2 Inheritance of light: the orientation and degree of softness of the reference original map, while maintaining a soft light perception if the original map is soft, and retaining a hard sense if the original map is more contrasted。
3 Inheritance of makeup: Maintaining the same level of style according to the makeup characteristics of the reference map (clear, strong, experimental, natural, etc.) and then making detailed changes in eye makeup, cheeks and lips。
4. Inheritance: The emotional atmosphere that perpetuates the reference map, such as coldness, sweetness, rebellion, mystery, etc., echoes expressions and gestures。
VI. Innovative change rules
1 Make-up change: add partial exaggeration to the reference style, e.g., expansion of the range of cheeks, increased lip dizziness, increased freckles or high light。
2 Structural change: The face angle, tailoring scope and small drawing collage can be adjusted so that each generation has a different image。
3 Apparel changes: inclusion of simple, designable accessories in the neck, ear or hand, corresponding to the main color, but not a complete copy of the reference map。
4 Detailed gestures: Increased touching of lips, cheeks, immediate and lower hand movements to make the picture more storytographic and participatory。
VII. COMPREHENSIVE QUALITY AND LAST TYPE
1 Clarity: Keep skin pores and details visible and avoid excessive skin grinding, but soften unfocused areas as appropriate。
2 Hue processing: a uniform overall color, a moderate decrease in the multiplicity of colours, and only a reinforcement of the main colour and a few dots。
3 Emotional expression: Control of the strength of the expression and the eye, based on the quality of the reference map, so that the image retains both its original style and its individual character。
Automatically read and create images in accordance with the above rules, with reference to images of people uploaded by users。
Google Nano-banana Pro:
GPT images 1.5:
11. Queen Punkgot
This hint is for Queen Van, but not for Queen Van
- Prompt:9:16 vertical frame, 85 mm lens, warm gold sunset. platinum silver hair, ruby gold crowns, leopard fur coats, sharp red glasses, heavy-chain crosses, metal and jewellery are indistinguishable. reverse the light of phnom penh with a crown, a soft face, a high contrast. four moods: 1 face cold to camera 2 side face 45° tattoo 3 under the mirror death eye 4 face closed eyes for holy light. cream excrete, fine film particles, high dynamic range, deep shadows and gold glazed, high-facility。
Google Nano-banana Pro:
GPT images 1.5:
12. Beach background
The girls take their own pictures. You can do the beach
- Prompt:PEOPLE (E.G., FIGURE 1) TWILIGHT BEACHES, FILM EMOTIONS. THE SEA WINDS ARE SHATTERING WITH DARK HAIR, WITH CHEEKS ON THEIR CHEEKS; NAKED RED, FINE GRANULAR SKIN. WHITE WITH A VEST FOR COWBOYS, DARK WAVES AND TWILIGHT CLOUDS, DUSTY LAVENDER IN THE SKY. LOW SATURATION COLD, SOFT CONTRAST, MIST AND PARTICLES, 8K PERISCOPE, BLUE NATURAL LIGHT。
Google Nano-banana Pro:
GPT images 1.5:
13. Mapping of girls
By uploading a picture and entering a hint, you can clearly and clearly analyse the person ' s anger and grief, clothing and objects。
FOR THIS HINT, GPT WAS INITIALLY REJECTED, AND I DELETED SOME OF THE PRIVACY OF THE PHONE, AND IT WAS CREATED。
Prompt:
> Rolle (role setting)
> You're a top-of-the-art game and animation designer (Concept Artist) who's good at making detailed role-setting (Character Sheet). You have the ability to “spread pixels” in order to see the level of the role's dress, capture micro-expressions, and pictify the items associated with them. You're particularly good at character and background stories with women's roles' private objects, side items and life details。
> Task (mission target)
> Generates a ** " Panoramic role depth concept breakdown"** based on the main image uploaded or described by the user. The picture must contain the entire body of the person at the centre and be surrounded by a stratification of the person ' s clothing, different expressions, core props, material features, and a display of the intimate and personal objects with a very high life quality。
> Visual Guidelines
1. Structure (Layout):
• Central position (Center): The whole body or main dynamic position where the role is placed as a visual anchor。
• Surroundings: in the absence of a central person, the sequenced composition of the dismantled elements。
• Visual guidance (Connectors): connects the decomposition of the surroundings with the corresponding part of the centre or the area to which the person belongs (e.g., a package connected to the hand) using hand-drawing arrows or a lead line。
Dismantling content (Deconstration Data) - Core iterative areas:
• Clothes Layer (Cloating Layers) [enhanced version]:
• Distinction of role costumes into individual displays. In the case of multiple layers, the internal state of the coat should be displayed。
• New: Privately contained (Intimate Apparel): Independently displaying the role of underwear, with emphasis on design senses and materials. For example, lace underwear (showing lace strip details), thongs (showing clippings), stockings (showing meat senses and socks design), plastic or safety pants。
Expression Sheet:
• Draw 3-4 different head features in the corner, showing different emotions (e.g. cold, shy, surprised, demented, or focused on theorem in lipstick)。
• Texture & Zoom [enhanced version]:
• Selection of 1-2 key locations for zoom-in features. For example: wrinkles of fabric, texture of skin, hand details。
• New: Quality profiling of items: Increased description of materials for small objects, e.g., salivation of ointment, granular texture of leather bags, fineness of cosmetic powder。
• Associated Items (Relaated Items) [Depth Overlay]:
• This is no longer limited to large-scale props and there is a need to add “life slices” to show roles。
• Pack bags and contents (Bag & Contents): Daily commuting bags or handbags for drawing roles and “opening them” to display scattered objects。
• Make-up and nursing (Beauty & Grouping): displaying their common cosmetic combinations (e.g., lipstick/spatting specials with a specific colour, pancake box with a mirror, perfume bottle design, hand cream)。
• Private living objects (Lifestyle & Intimate Items): items on the hidden side of a visualized role. Depending on the character of the role, it may include: a private diary, a common drug/refill box, an electronic smoke, or a more private object (e.g. an aircraft cup/functory as mentioned by the user, to be presented with an objective view of the design, indicating the model or design characteristics)。
Style & Notes:
• PAINTING STYLE: MAINTAIN HIGH-QUALITY 2D ILLUSTRATION OR CONCEPTUAL DESIGN SKETCH STYLE WITH CLEAN LINES。
• Background: To create an atmosphere for the design of manuscripts using yellow, parchment or light grey texture background。
• WORD DESCRIPTION: SIMULATE HAND-WRITTEN NOTES AT EACH DECOMPOSITION ELEMENT, OUTLINING MATERIAL (E.G., “SOFT LACE”, “SAND LEATHER”) OR BRAND/MODEL IMPLICATION (E.G., “COMMON COLOUR #520”, “SUPPLEMENTATION”)。
Workflow (implementation logic)
When the user provides a picture or description:
4. Analysis of the core characteristics, style and potential character of the subject。
5. Extracting decomposable first-stage elements (jacket, shoes, big face)。
6. Brain supplementation and design of secondary Depth elements (how does she wear her underwear? What kind of lipstick does she have in her bag? What does she use when she's alone。
7. Generate a combination of all these elements to ensure that the vision is accurate, that the light is uniform and that the comments are clear。
8. USE CHINESE: ENGLISH TAG, HIGH-RESOLUTION 4K HD OUTPUT
Google Nano-banana Pro:
GPT images 1.5:
14 -- The Magician
This hint allows people to play different directions。
- Prompt:ON THE BASIS OF [UP-TO-DATE IMAGES OF PEOPLE] AND KEEPING THEIR FACES ABSOLUTELY UNCHANGED, A HIGH-QUALITY 3 X 3 KAUPULE COLLAGE PHOTOGRAPH WAS PRODUCED, WITH NINE SEPARATE IMAGES SHOWING THE PERSON IN: THE FLYING BLUE TANG DYNASTY (CLASSICAL GARDEN BACKGROUND); THE ABLE AND DEEP BLUE SKY HOSTESS UNIFORM WITH A SILK SCARF (THE BACKGROUND OF THE MODERN AIRPORT TERMINAL); THE INK EMBROIDERED SILK FLAG ROBE (THE INDOOR BACKGROUND OF THE SHANGHAI WIND IN THE SHUG SEA); THE WIL KUI-GIN GUARD FISH FLYING SUIT WITH AN EMBROIDERY KNIFE (THE BACKGROUND OF THE TWILIGHT PALACE); THE WHITE MODERN NURSE ' S DRESS (THE BRIGHT AND WARM HOSPITAL BACKGROUND); THE DAILIEST-DRY-WEAVE TANG (THE BACKGROUND OF THE TRADITIONAL TEA COMPLEX); THE BLUE JK OF THE GREEN NAVY ' S UNIFORMS (THE BACKGROUND OF THE CHERRY BLOSSOMING CAMPUS); THE HANDSOME-SPIRITED PILOT ' S JACKET WITH A SUNGLASS MIRROR (THE BACKGROUND OF THE HANGAR AIRCRAFT); AND THE TIME-STYLE WIND-COVERED SUIT (THE BACKGROUND OF THE MODERN URBAN SCENE) REQUIRING A PICTURE OF FILM, DETAIL AND HARMONY. WIDGET RATIO 9:16
Google Nano-banana Pro:
GPT images 1.5:
15. Personal details
Prompt:
Overall image setting
> The image is a vertical scale of three to four, with a single image of a nine-gauge layout, three rows, each of which is an independent lens。
> The overall style is overwritten real-time photography, with a sense of life, light film particles and soft and dark angles, which can be published directly as today ' s wear。
> The whole picture ends with nine lenses of a set of well-dressed stories with a whole set of shapes, body priorities and life details。
Reference chart binding rules
The above reference map is the only visual basis。
Model 5 officers follow a reference map in full proportion to their body size, hairdressing and makeup, and do not change their identity。
The clothing layouts, colours, folding layers and accessories are all inherited reference maps, including earpieces, bags, shoes, nails, etc。
The type of scene, light orientation and overall color are based on a reference map and are only slightly extended at the same location or within the same scenery system。
The model in the nine grids is always the same person, the same outfit, and only changes the space, the structure and the view。
Nine-gauge spectroscope structure
Topline: holistic perspective and spatial sense
The picture is a face-to-face whole, the model displays a whole set of clothes from head to toe, with a natural attitude that allows for a slight walk or turn, and the environment and light follows a reference map and is used as a reference for the overall model。
The second is an overlooking angle, with a slightly higher altitude than a human being, highlighting the deep- and ground-based relationship between the body and the body for a whole or large body view, and facilitating the reading of the overall contours as a proportion of space。
The third is a viewing angle with a slightly lower altitude than that of a human being, strengthening the feeling of a falling line and clothing for a whole or large body, while bringing out an extension of buildings, smallpox or the sky to create air and stereo。
Middle row: physical focus details
The four images show the details of the face, which are focused on the head and neck, and are presented in a make-up atmosphere, five official features, hair and sea, as well as in local beautifications in the ear, with a natural nature of the skin and the light in the same direction as the reference map。
Image 5 shows the details of the structure of the chest and the top shirt, covering the clavicle, chest and upper torso, highlighting the size of the shirt, the shape of the collar, the shoulder line, the thickness of the fabric and the folding, and showing wrinkled and constrictive positions, reinforcing the upper half。
Image VI shows the details of the hips and the lower-assembly structure, which is concentrated below the waist to the upper edge of the thigh or the area of the hip, showing the lower cut, asswipes and fabrics in a manner that clearly shows the position of the pockets, the direction of the stitches and the direction of the wrinkles, with emphasis on the lower half of the body to the scale。
Bottom line: Details of body and accessories
Image VII shows the details of the feet and shoes, which are concentrated in the ankles to the soles of the shoes, with the generals of the shoes, the material on the face of the shoes, the height of the shoes and the thickness of the soles of the shoes, with the material and light of the ground, and with the socks, the legs of the pants or the skirts。
Image VIII shows the details of the hand and movement, which are concentrated in the hand and forearms, and shows natural gestures, such as handbags, ablutions or the taking of small items, the colour of nails, the texture of fingers and nail polish, and the appearance of rings, bracelets, watches, etc。
Image 9 shows key accessories and material features, focusing on one or two key accessories, such as earpieces, necklaces, bags, or belts, with emphasis on the high light and texture of materials such as metals, leather, fabrics, stones, etc., and a simplified background to ensure that the vision focuses on the accessories themselves。
Clothing layout and fabric
Main wearing structures and colours strictly follow the reference diagrams and do not change individual items at will。
The key contours, such as shoulder lines, waist lines, pants or skirts, and coating lengths, are reinforced on multiple occasions in different lenses。
The knitted fabric displays a clear sense of needle foot and a slight velvet, the thin coat and shirt displays soft and high light and natural wrinkles, the cowboy fabric displays texture, paints and wires, and the leather displays fine high light and creases and dark layers。
Apparel and Micro Details
All accessories are based entirely on reference maps, including earpieces, necklaces, rings, bags, belts, watches, etc。
The edges of the metals are clear, high and soft and reflective, and can be seen in small grafts and structural details。
Packet sutures are even, edges are delicately processed, hardware is of clear quality and signs are clearly fixed。
Shoelines, the thickness of the soles and the head curves are clearly visible and can be slightly worn or fractured to enhance the sense of truth。
Hair and skin
Hairstyle and hair colours, which fully inherit the reference map, change the filament flow and the light of light only because of a change of angle, see the light of light and a little darker。
The skin is natural and visible, with small pores and small defects, avoiding excessive skin grinding, and the overall state is consistent with the atmosphere of the reference map。
Environment & Hue
The type of scene is mainly based on reference maps, while street filming is extended to the same block or the same outdoor environment and indoors to the same space or room of the same style。
At the top line, more environmental information is kept and the middle and bottom line background is moderately reduced, focusing on the details of the person and the line。
The global tone is uniform, and nine images can be fine-tuned in the dark and in the light, but the overall atmosphere and color orientation are consistent with the reference map, and are always in visual focus。
Google Nano-banana Pro:
GPT images 1.5:
16. Animal City selfie
- Prompt:In Chinese: 24 mm wide self-portrait, cinema microlight. Asian girls with long black hair sit in the middle, with green shirts with purple-and-white ties and smiles on their cell phones; right shoulder with orange fox Nick's arm, and left side Judy Grey Bunny leans on her head. The three men, with hair and skin in the same light, the light is cold and warm, and the background poster Mad Animal City is visible on the screen. Outside the light view of the butter, the film-grade blue oranges, 4K fibers, no watermarks, no cartoons, warmth。
It works better in English。
> **Prempt:**
> {
>title: "Cinema Selfie with Zootopia Characters – Policealistic",
>language: "en",
> "Description": "High-quality special selfie of a young Asian woman with Nick Wilde and Judy Hopps from Zootopia, finematic watching, side-angle light.",
“Physical”: “for example, a person who is a member of an American organization who is not a member of the United Nations, who is not a member of the United Nations and who is not a member of the Office of the High Commissioner for Human Rights.”,
>Negative_prompt: “low question, cartoonish environment of the human, watermark, logo overplays, text analysis image, out-of-focus face, bad anatomy, incorrupt lighting, obvious CGI acts, extra limps, muta-hands, oversacurated skins, over-sharping, noise, motion , motion blurded perspect, complexes on first reading”,
> “characters”:
{
"name": "Human subject",
"Attributes": "Young Asian women, long black hairs, milking, green shirts, white shorts, purple-and-white striped, natural makeup"
},
{
"name": "Nick Wilde.",
"attributes": "tall orange fox, unhroporpic, technically expression, realistic shadow, armed around human"
},
{
"name": "Judi Hopps",
"attributes": "smalt radit, anthroporphic, grey/white fur, leaving head against man, exceptive eyes"
}
],
"compositation": {
"camera": "wide-angle 24mm equivalent",
"Framing": "there-quarter self-operation, for the ground humans lightly left of center, Nick to right, Judy close to center, movie screen witness behind",
"aspect_ratio": "3:2"
},
"lighting": {
"primary": "soft directional key from movie clear",
"fill": "subtle warm fill from above ambitt lights,",
"rim": "this warm rim light sparing subjects from background,",
"mood": "cinematic, warm nostalgic"
},
"rendering_interactions": {
quality: "ultra-high.",
"style": "photoeconomic; cinematic",
"color_grading": "filmic contras, subtle-real-orange gade, natural skin tones,",
"post_processing": "lens bokeh, film plain 0.5, subtle vignette, dodge & burn for facial highlights"
},
"safety_and_rights": {
"containers_copyried_characters": true,
“notes”: “User intends to develop Copyarighters (Nick Wilde, Judy Hopps)
},
"output_settings": {
“resolution”: “4096 x 2730”,
“guidance_scale” : 7.5,
"sampler": "ddim",
“steps”: 28,
“seed”: “random_or_provide_seed”,
"samples": 1,
"format": "PNG"
}
}
Google Nano-banana Pro:
GPT images 1.5:
17. Christmas specials -- My Christmas, with myself
- Prompt:16:9 Double-gauge: Two left and right, one face intact. The left green sweater was wearing a Christmas hat, with a right turn on the body, with one end of the gun in hand; the right was changed to the red sweater, with one left, and the same side of the gun. Footside gift boxes with snowy dolls, yellow/green background, Christmas lights around. No shadows, no fun “cloning” interacts as if it was in the mirror。
Google Nano-banana Pro:
GPT images 1.5:
End
THAT'S WHAT I PICKED THIS WEEK
TO BE HONEST, WHEN I PUT TOGETHER THESE DOZEN SETS OF HINTS, I MYSELF WAS FRIGHTENED BY THE RATE OF EVOLUTION OF AI NOW -- A LOT OF TIMES, IT'S REALLY HARD TO TELL WHETHER IT'S A PHOTO OR AN AI BIOGRAPH。
Here's a little advice:
When you get the hint, don't just copy it。
You try to change the color, the weather, the light, you'll find the door open to the new world。
I hope these hints really inspire your creation。
I'VE PACKED IT ALL. THE OLD RULE: PAY ATTENTION TO THIS PUBLIC CALL, BACKSTAGE REPLY. IF YOU DON'T KNOW, YOU CAN ALSO ADD "HAND HANDS TO TEACH YOU TO USE AI FOR VIDEO-REALIZATION" TO THE CHILDREN WHO COME TO SCHOOL AND ASK ME。