The campaign brief
Originally prepared 21 October 2026. Interface labels, model availability and credit costs can change. Check the controls shown in your account.
Today you are making one original fashion campaign, one decision at a time. Follow the steps in order. The goal is one clean poster, not a perfect collection. You do not need to invent the concept while you are live.
The creative decisions are already made
Concept: VELA, a fictional label for this creative demonstration. An adult woman with a short dark curly bob wears a cobalt blue tailored coat and trousers over an ivory high-neck blouse. She carries a small ivory handbag. Oversized coral fabric folds create the set. The final poster says VELA and FORM IN MOTION.
The reference inspires the process, not a copy of its model, styling, photography or branding. This is a new campaign. VELA is a working concept name, not a claim of trademark clearance.
Your only job before the show
1. Open ImagineArt in your browser and this guide side by side.
2. Create a local folder named VELA campaign.
3. Use the Copy prompt buttons below to copy the exact text for each step.
4. Make sure image generation and reference-image controls are available. Do not buy anything or change plans to follow this guide.
5. Take a sip of water. Start with Step 1. Everything else can wait.
The simplest finish line
Required: approved model, styled model, bag, set, hero and final poster. Optional: the character sheet, extra detail images and separate wordmark. If time gets tight, skip the optional work. Keep one good result from each required stage.
How to use the prompts
Copy only the paragraph under Paste this prompt. Before generating, attach exactly the images listed for that step and remove unrelated references. A reference tells the model what to preserve; the prompt tells it what to change. Save a keeper before moving on. Do not regenerate a good image just because another option might be better.
Back to navigation ↑The generation routine
Use the same image model throughout when it supports the needed reference images. The controls checked in October 2026 are listed below. This is an image workflow, not a promise that ImagineArt duplicates the reference platform’s node canvas.
Verified interface route
1. Open https://www.imagine.art/image or choose Image from the left sidebar at imagine.art/home. Use Create. Sign in to your ImagineArt account first if prompted.
2. Open the combined model and ratio control beside the prompt, shown as “Nano Banana - Lite / 1:1” during inspection. Choose Nano Banana 2 as the starting model if available to your account. Keep it for the sequence if it supports the required references.
3. Set the step’s aspect ratio under Settings. 4:5 and 1:1 were present. Turn Prompt Enhancer off if enabled so your pasted instructions remain the intended brief.
4. Check Variations. The inspected screen showed 2; use the smallest count offered. Check Visibility before uploading: Public was selected. Choose Private if available and desired. Do not upload anything you are unwilling to share publicly while Public is selected.
5. Close Settings. Choose Add references, then Choose Files, and add the listed files. Paste into “Describe the image you imagine.” Check the references and cost, then select Generate.
6. Save a keeper using the result’s actual download control. The signed-in result screen was unavailable during inspection, so its exact control label is not assumed here.
The repeatable routine and reference map
Repeat for every step
1. Read the Attach line. Remove the previous step’s references if they are not listed.
2. Add the listed saved images in the stated order.
3. Paste the prompt. Set the aspect ratio listed for this step if that option is available.
4. Start with one output or the smallest available batch. Read the displayed credit cost before generating.
5. Check the result against the Keep it if line.
6. Download the keeper and give it the exact short filename shown. Keep the original resolution.
7. If it fails, use one targeted repair. If that still fails, simplify and move on.
Reference map
01_model → 02_sheet → 03_styled
04_bag and 05_set are independent images
03_styled + 04_bag + 05_set → 06_hero
06_hero → 07_face and 08_bag_detail
06_hero + optional 09_wordmark → 10_poster
The names describe files you will create during the show. No campaign images are supplied with this guide. Attach your saved outputs, not screenshots of the LinkedIn slides. If a screen accepts only one reference, follow the fallback on the last page.
Keep the settings boring
Use 4:5 portrait for campaign images and the final poster where available. Use 1:1 square for the character sheet, bag and logo. If 4:5 is absent, use the nearest portrait ratio offered and leave generous margins for a later crop. Do not invent a seed, reference-strength slider or negative-prompt control if the interface does not show one.
Back to navigation ↑Step 1 Create the model
Attach: nothing. Ratio: 4:5 portrait. Save as: 01_model.png.
This creates an original adult model. You do not need to hunt for someone else’s photograph or upload a real person’s likeness.
Paste this prompt
Create a photorealistic studio casting portrait of an original fictional adult woman, approximately 35 years old. Warm medium-brown skin, oval face, expressive dark brown eyes, natural eyebrows, short chin-length dark curly bob with a soft side part. Calm confident expression, subtle natural makeup. Fully clothed in a simple ivory high-neck long-sleeve blouse. Head and upper torso visible, face turned only slightly from camera. Plain warm light-gray backdrop, soft large-window light, natural skin texture, accurate anatomy, editorial photography. No jewelry, no bag, no text, no logos, no collage. Leave comfortable space around the head.
Keep it if
You like the face; both eyes look natural; the hair has a distinctive readable shape; the neckline is modest; there are no strange accessories. Pick on identity, not background polish.
If it needs one repair
Keep this same adult woman, facial proportions, skin tone, eye color and short curly bob. Correct only the eyes so both pupils and eyelids are anatomically natural. Preserve the pose, blouse, lighting and background. Do not beautify or redesign her face.
Use that repair only when the eyes are the problem. Replace the eye instruction with the single actual defect if needed.
Say this on stream
“I’m choosing our model first. Once I like the identity, I’m going to stop changing her face and build the campaign around her.”
Step 2 Make an optional identity sheet
Attach: 01_model only. Ratio: 1:1. Save as: 02_sheet.png.
Using the attached portrait as the sole identity reference, create a clean four-panel character reference sheet of exactly the same adult woman: front headshot, left profile headshot, three-quarter headshot, and full-length standing view. Preserve her face, warm medium-brown skin, dark brown eyes and short dark curly bob in every panel. She wears the same simple ivory high-neck long-sleeve blouse and plain charcoal straight-leg trousers. Neutral expression, even studio lighting, pale-gray background. Clear equal panels with space between them, no labels, no text, no accessories.
Check each face against 01_model. If the sheet creates different women, discard it and use 01_model directly in Step 3. A bad sheet is worse than one good portrait.
Back to navigation ↑Step 3 Lock the styling
Attach: 01_model, then 02_sheet only if all faces match. Ratio: 4:5. Save as: 03_styled.png.
Paste this prompt
Create a full-length fashion fitting photograph of the exact same adult woman in the attached identity reference images. Preserve her face, skin tone, short dark curly bob and age. Dress her in a saturated cobalt-blue tailored long coat, matching straight-leg trousers, an ivory high-neck blouse, and simple ivory low-heeled ankle boots. The coat is neatly structured, worn closed at the waist, with long sleeves and a modest fully covered neckline. Small polished gold stud earrings only. She stands naturally facing camera with both arms relaxed, both hands visible, feet fully in frame. Plain warm light-gray seamless studio background, soft even editorial lighting, realistic fabric seams and drape. No handbag yet. No text, no logos, no extra people.
Keep it if
It is unmistakably the same woman. The coat is cobalt rather than navy. Trousers, boots and both hands are visible. Nothing is sheer, cropped or revealing. The silhouette is clean enough that you can recognize it in another shot.
If the face changes
Attach 01_model first and the failed styled image second. Paste:
Use image 1 as the identity authority and image 2 only for clothing, pose and framing. Restore the exact woman from image 1, including her facial proportions, skin tone and short curly bob. Keep the cobalt coat, trousers, ivory blouse and boots from image 2. Change nothing else. Do not average the faces.
If the clothing changes later
Return to this styled keeper as the wardrobe authority. Do not use the failed later image as your only reference.
Say this on stream
“Now I’m locking the wardrobe. The blue silhouette, the face and the hair are our continuity anchors. We can change the shot without reinventing the person.”
A useful stopping rule
This does not need to be your finished hero image. A straightforward full-length reference with a reliable face and clean clothing is more useful than a dramatic pose that hides everything.
Back to navigation ↑Steps 4 and 5 Build the bag and set
Step 4 The bag
Attach: nothing. Ratio: 1:1. Save as: 04_bag.png.
Photorealistic luxury product studio photograph of one original small structured ivory leather top-handle handbag. Compact gently rounded rectangular body, single rigid curved top handle, one simple small brushed-gold rectangular clasp centered on the front, neat ivory stitching. No shoulder strap, no chain, no monogram, no lettering, no visible brand. Three-quarter front view at eye level, entire bag visible, warm pale-gray seamless background, soft controlled studio light, realistic leather grain and a subtle grounding shadow. No people, no hands, no extra objects, no collage.
Keep it if: there is exactly one handle, one clasp and a believable bag opening. The front and side are clear. Do not add a brand to the bag today; that is one less consistency problem.
Step 5 The sculptural set
Attach: nothing. Ratio: 4:5. Save as: 05_set.png.
Photorealistic empty fashion editorial studio set. Oversized sculptural coral-orange fabric ribbons curve and fold around the left and right edges of a warm ivory studio, like a large abstract folded textile installation. Soft peach undertones, tactile matte fabric, elegant architectural scale. The central standing area and floor are clear and unobstructed for a full-length fashion model. Leave the upper quarter calm and mostly ivory for a future wordmark. Broad soft directional light from upper left, gentle grounded shadows, frontal camera at human chest height. Sophisticated bright color, clean composition. No person, no mannequin, no handbag, no text, no logo.
Keep it if: the middle can hold a person without a fabric fold running through her face. There is a floor and a believable horizon. The top is quiet enough for text.
If the set is too busy
Simplify this same studio set. Keep just two large coral fabric folds near the side edges. Clear the entire central standing area and upper quarter to warm ivory. Preserve the lighting, material texture and floor. No people, products or text.
Say this on stream
“The bag and the set have separate jobs. I want the bag to stay recognizable, and the background to support the subject instead of fighting her.”
Back to navigation ↑Step 6 Bring the campaign together
Attach in this order: 03_styled, 04_bag, 05_set. Ratio: 4:5. Save as: 06_hero.png.
Paste this prompt
Create one photorealistic full-length fashion campaign photograph using the three attached references. Image 1 is the exact model and wardrobe authority: preserve her face, short dark curly bob, skin tone, cobalt tailored coat and trousers, ivory high-neck blouse and ivory boots. Image 2 is the exact handbag authority: preserve its ivory leather body, single curved top handle and small gold rectangular clasp. Image 3 is the set and lighting authority: use the coral sculptural fabric folds and warm ivory studio. Place the same woman standing slightly right of center, facing camera with a relaxed confident expression. Her right hand naturally holds the bag by its top handle beside her right thigh, with the bag front turned toward camera. Her left arm rests naturally at her side. Keep both feet, the entire bag and the entire head in frame. Match light and contact shadows so person, product and set belong in one photographed scene. Leave the upper quarter calm and mostly ivory for typography. Eye-level camera, natural perspective, refined editorial finish, realistic skin and fabric. No text, no logos, no extra people, no extra bags.
Check before you do anything else
1. Face: still the same woman?
2. Wardrobe: same blue coat, trousers and ivory neckline?
3. Bag: one handle and one clasp, the same shape?
4. Hand: a believable grip, no fused fingers?
5. Feet: both on the floor with sensible shadows?
6. Layout: enough clean space above the head for VELA?
The easiest rescue
If the grip fails twice, remove the hand interaction. Keep the model standing and put the bag on a low ivory plinth beside her. That is a valid art-direction choice, not a failed demo.
Preserve this exact woman, wardrobe, face, composition, lighting and coral studio set. Change only the bag placement: place the same ivory handbag upright on a simple low ivory plinth beside her, with its front fully visible. Both of her arms rest naturally at her sides. Remove any duplicate bag. Keep realistic contact shadows.
Say this on stream
“This is the first time all the decisions meet. I’m checking identity, product shape and physical contact before I worry about making it more dramatic.”
Back to navigation ↑Step 7 Optional campaign details
Do these only after 06_hero is a keeper. Two detail images are enough. Skip them if you are behind.
The face and fabric detail
Attach: 06_hero, then 01_model. Ratio: 4:5. Save as: 07_face.png.
Create a tight editorial portrait from the same fashion campaign as image 1. Image 2 confirms the exact face identity. Preserve this same adult woman, short dark curly bob, warm medium-brown skin, subtle makeup, gold stud earrings and cobalt coat over an ivory high-neck blouse. Frame from upper chest to above the head, three-quarter face angle with eyes toward camera. Keep the coral sculptural studio background softly out of focus and the same upper-left soft light. Natural skin texture and realistic cobalt fabric. No bag, no text, no new jewelry, no change of identity.
Keep it if: the face matches the hero; the neckline and earrings match; it looks like another camera position in the same session, not a new campaign.
The bag detail
Attach: 04_bag, then 06_hero. Ratio: 4:5. Save as: 08_bag_detail.png.
Create a close-up product editorial photograph for this same campaign. Image 1 is the handbag design authority; image 2 supplies the cobalt wardrobe, coral-and-ivory color palette and soft lighting. Show the exact ivory structured handbag resting upright on a low ivory plinth, with the single curved handle, small gold rectangular clasp, neat stitching and leather grain clearly visible. A small softly blurred edge of cobalt fabric may appear in the background, but no person or hand touches the bag. Simple coral fabric fold out of focus behind it. Natural perspective, believable scale, crisp product detail. No extra straps, no extra clasps, no lettering, no new logo.
Keep it if: the bag matches 04_bag exactly in its major design details. If it has a different handle or clasp, reject it even if it looks beautiful.
Say this on stream
“The wide shot tells the story. The detail shots prove we can keep the same world, face and product when the camera moves.”
Save time without generating
You may simply crop the original hero for a detail preview if the resolution holds up. Call it a crop, not a newly generated matching shot. Never sacrifice the finished poster to chase optional images.
Back to navigation ↑Steps 8 and 9 Finish the poster
Step 8 Optional separate wordmark
Attach: nothing. Ratio: 1:1. Save as: 09_wordmark.png.
Create a clean typography proof for a fictional fashion label. The only text is exactly VELA in uppercase. Elegant high-contrast black serif lettering, generous but balanced letter spacing, one centered horizontal wordmark on a plain pure white background. Crisp letterforms, no icon, no mockup, no texture, no shadows, no extra text. This is a flat graphic, not a photographed sign.
Check the four letters individually: V E L A. This is a raster concept wordmark, not a vector logo. White is not transparency. If it is wrong after one repair, skip it and use the text instruction below.
Step 9 The final campaign poster
Attach: 06_hero first; add 09_wordmark only if correct. Ratio: 4:5. Save as: 10_poster.png.
Paste this prompt
Finish image 1 as a refined fashion campaign poster while preserving the existing photograph: the exact same woman, face, cobalt outfit, ivory bag, coral set, framing and lighting. Do not redesign or replace any photographic element. In the clean ivory space at the top, add the word VELA in large black elegant high-contrast serif uppercase lettering. If a second image is attached, use it only as the wordmark design reference, not as a white rectangle or a photographed object. Center a small black uppercase line near the bottom reading exactly FORM IN MOTION, clear of the feet and bag. These are the only two text elements. Maintain generous safe margins and clean editorial hierarchy. Spell every letter correctly. No additional copy, no watermarks, no dates, no website, no border.
Final check at full size and thumbnail size
VELA is spelled correctly. FORM IN MOTION is spelled correctly. Face, hands and bag survived the finishing pass. Text has contrast and does not touch the head, feet or bag. The poster works when reduced to phone size. No invented watermark or unwanted logo appeared.
If the generator changes the photo or misspells text
Return to 06_hero and finish with editable type in your familiar layout editor. Place VELA centered in the top quiet area, and FORM IN MOTION centered near the bottom. Keep a margin of about 5% of the canvas on every edge. Start with a high-contrast serif font for VELA and a clean sans serif for the small line. This manual typography route is the reliable fallback; its exact editor controls are not part of the verified ImagineArt steps.
Download the final at its original resolution. Keep the text-free hero too. Do not claim it is print-ready unless resolution, dimensions and print requirements have actually been checked.
Back to navigation ↑Your live show plan and rescue prompts
A calm 60 minute run of show
00–05: Show the inspiration and explain the separate-decision approach.
05–12: Create the model. Make the identity sheet only if the model is ready quickly.
12–20: Lock the styling.
20–30: Make the bag and empty set.
30–43: Combine the hero. Spend your repair time here.
43–48: Optional details or wordmark, not necessarily both.
48–56: Finish the poster and check the lettering.
56–60: Show the before-and-after sequence and recap what you locked at each stage.
Generation wait time varies. If you reach minute 40 without a hero, skip all optional images. If the poster text fails, show the clean hero and finish type manually. A strong image with honest process is better than rushing ten broken outputs.
One problem at a time
Always attach the latest good image, not just the failed result. Name what must stay fixed, then name one change. Do not stack five repairs into one prompt.
Preserve the exact face, hairstyle, clothing, bag design, background, camera position and lighting of the attached image. Change only [describe one visible defect and its correction]. Do not add any new people, objects, accessories or text.
If only one image reference is supported
Use 03_styled as the reference for the hero. In the hero prompt, replace the numbered-image instructions with: “Preserve the model and wardrobe in the reference. Add a small structured ivory leather handbag with one rigid curved handle and one small gold rectangular clasp. Place her in a warm ivory studio framed by two oversized coral fabric folds.” Expect less exact bag and set matching. State that tradeoff on stream. Do not pretend all three references were used.
If generation is slow or unavailable
Talk through the prompt and the decision each reference controls. Review the keepers you already have. Do not switch to a paid plan or an unverified product feature mid-demo. If you have no keepers yet, explain the concept and use the guide as a live planning exercise rather than presenting nonexistent results.
The closing line
“The useful part is not one magic prompt. It’s deciding what stays fixed, changing one thing at a time, and keeping the images that actually belong together.”
Reference
Inspired by Vira Anikieienko’s nine-slide IPVISE fashion-poster process, inspected on 21 October 2026: https://www.linkedin.com/feed/update/urn:li:activity:7514003951227224064/
Original concept and prompts in this guide are for your ImagineArt demonstration, not a reconstruction of undisclosed settings from that post.
Create in ImagineArt. Referral link. I may earn a commission if you purchase through it. Disclosure