#花譜 × #CHiCOのコラボWリリース 2025年6月25日(水)配信 ───────────────── 花譜×CHiCO 「撃って」作詞・作曲・編曲：#清竜人 CHiCO×花譜「Story Tellers」作詞：MOMIKEN(#SPYAIR)/CHiCO 作曲：UZ(SPYAIR) 編曲：tasuku ─────────────────

花譜

@virtual_kaf · ai-influencer

INSTAGRAM · 2025-06-27Source

CHiCOのコラボWリリース花譜清竜人 SPYAIR)/CHiCO

3.7Klikes

4comments

Remix This

Prompt

HIGH-GRANULARITY VISUAL INVENTORY

Subject(s)
- Count: 2 subjects, standing back-to-back

Subject A (left)
- Style: anime/CG character
- Hair: pastel pink hair styled into a low bun with a braided wrap and side braid detail
- Pose: side profile facing left, head slightly down, hands raised in front (as if holding or examining something)
- Outfit: dark navy/purple futuristic jacket/hoodie with graphic accents; orange circular target-like emblem on the upper sleeve/shoulder area

Subject B (right)
- Style: photoreal human (live-action look)
- Hair: medium-length brown hair with soft layers, slight wave; profile facing right
- Pose: back-to-back with Subject A, hands raised in front mirroring the left subject
- Outfit: white long sleeves with a blue overall/pinafore-style dress; darker outer layer near shoulders (black/charcoal)

Environment
- Background: lush green forest/park with heavy bokeh and soft depth (out-of-focus trees and foliage)
- Lighting: soft natural daylight, gentle highlights, no harsh shadows
- Atmosphere: dreamy, cinematic, slightly hazy

Composition
- Square album-cover layout
- Two subjects centered, back-to-back, occupying the mid frame
- Shallow depth of field; background strongly blurred
- Color contrast: green background vs pink hair and blue clothing

Typography / graphic overlay
- Top center: small spaced-out text “CHiCO × 花譜” (white)
- Bottom center: large elegant serif title “STORY TELLERS” (white), spanning across both figures

Color palette
- Dominant: deep greens (forest)
- Accents: pastel pink hair, navy/purple jacket, blue dress, white text

Image style
- Mixed-media composite: anime character + photoreal subject in the same scene
- Cinematic album cover, clean, high resolution

MASTER PROMPT (English)

[Scene]
A cinematic square album-cover image set in a lush green forest with strong bokeh and shallow depth of field, soft natural daylight, dreamy atmosphere.

[Subjects]
Two subjects standing back-to-back in the center:
1) Left subject: an anime-style girl with pastel pink hair in a low bun with braided wrap detail, side profile facing left, wearing a dark navy/purple futuristic jacket with an orange circular target emblem on the sleeve.
2) Right subject: a photoreal young woman with medium-length brown hair, side profile facing right, wearing a blue overall/pinafore dress with white long sleeves.
Both hold their hands raised in front in a mirrored gesture.

[Composition/Camera]
Square 1:1, centered duo, shallow depth of field, background heavily blurred green foliage, cinematic framing.

[Typography]
Add white album-cover typography: small spaced text at top “CHiCO × 花譜”; large elegant serif title across the bottom “STORY TELLERS”.

[Style/Rendering]
High-quality mixed-media composite (anime + photoreal) with consistent lighting and depth-of-field, clean, high resolution.

[Detail constraints]
Exactly two subjects back-to-back; keep the forest bokeh; keep the mixed anime + photoreal pairing; keep the white typography placement.

NEGATIVE PROMPT
Extra people, front-facing poses, city background, harsh sunlight, messy typography, illegible text, low-res, artifacts, watermark, over-sharpening, mismatched lighting between subjects.

SUGGESTED PARAMETERS (starting points)
- Aspect ratio: 1:1
- Resolution: 1024x1024 (or match 1080x1080)
- Steps: 30–45
- CFG / guidance: 5–7
- Sampler: DPM++ 2M Karras (SDXL) / high-quality default (FLUX)
- Lens feel: 50–85mm, shallow DOF, cinematic bokeh
- Seed: 250627

DELTA PROMPT STRATEGY (top drift risks → corrective micro-prompts)
1) Mixed-media becomes all anime → “right subject photoreal, live-action look”
2) Lighting mismatch → “consistent soft daylight lighting on both subjects”
3) Background loses forest bokeh → “lush green forest bokeh, heavily blurred foliage”
4) Back-to-back pose breaks → “two subjects standing back-to-back, mirrored side profiles”
5) Pink hair style changes → “pastel pink low bun with braided wrap detail”
6) Outfit details vanish → “blue pinafore dress with white sleeves; navy/purple jacket with orange target emblem”
7) Hands/gesture distorts → “hands raised in front, natural fingers, gentle gesture”
8) Typography missing → “album cover typography: top small artist text, bottom large serif title”
9) Too sharp background → “strong shallow depth of field, background heavily out of focus”
10) Composition shifts off-center → “centered duo, symmetrical balance”

The Story Tellers Album Cover: How virtual_kaf Built This AI Art

This cover works because it turns a collaboration into a visual metaphor. Two figures back-to-back, facing opposite directions, sharing the same light. One reads as illustrated/anime, the other reads photoreal—yet the scene holds together because the environment and depth-of-field are consistent. It feels like one world, not two layers pasted together.

If you want to recreate this kind of “two styles, one story” cover, the key is to treat the background as glue. A forest with heavy bokeh is perfect: it hides hard edges, gives you cinematic softness, and lets the typography sit cleanly on top.

Why it spreads: it visualizes the collab in one glance

Collaboration posts succeed when the visual makes the concept instantly obvious. Here, the pairing is readable even at thumbnail size: two characters, back-to-back, different worlds, same mood. That clarity makes it shareable—and it also creates a “double take” loop as viewers notice the mixed media and zoom in.

The typography choice matters too. A large serif title across the bottom signals “album cover” immediately. People don’t just see an image; they see a release. That context increases clicks, saves, and reposts.

Signal table

Signal	Evidence (from this image)	Mechanism	Replication Action
Collab clarity	Two subjects paired back-to-back	Instant “two artists / two sides” read	Use a symmetrical duo pose (back-to-back, mirrored profiles)
Mixed-media novelty	Anime character beside a photoreal subject	Creates a second-look moment	Lock lighting + DOF so the mix feels intentional, not pasted
Cinematic softness	Forest bokeh and shallow depth of field	Hides compositing seams and feels premium	Choose a bokeh-friendly background (forest, city lights, stage haze)
Album-cover framing	Large serif title + small artist line	Signals “official release” immediately	Use a two-tier type system: small top credit + big bottom title

Use cases & transfers

Best-fit scenarios

Music collabs: visualize “two voices, one track” as a duo composition.
Anime x artist promos: keep the illustrated character iconic; let the environment unify.
Dual-release campaigns: make two variants by swapping which side is illustrated.
Story-driven singles: back-to-back profiles suggest tension, partnership, or parallel worlds.

Not ideal

Text-heavy announcements: cover layouts need breathing room for title type.
Complex group photos: mixed media gets messy beyond 2–3 subjects.
Hard-edged backgrounds: sharp architecture makes compositing seams obvious.

Transfers (3 recipes)

Recipe 1: Background glue
- Keep: heavy bokeh + shallow DOF + consistent daylight
- Change: {background} = forest / city night bokeh / stage haze
- Slot template: “two-subject album cover, mixed media, {background}, shallow DOF, clean serif title”
Recipe 2: Style inversion
- Keep: back-to-back pose and typography system
- Change: swap which subject is {illustrated} vs {photoreal}
- Slot template: “left {style A}, right {style B}, consistent lighting, album cover typography”
Recipe 3: Gesture anchor
- Keep: mirrored hand gesture
- Change: {gesture} = holding a small object / reaching / clasping hands
- Slot template: “back-to-back duo, mirrored {gesture}, cinematic bokeh, release cover”

Aesthetic read: consistency beats realism

You don’t need the anime and photoreal parts to match perfectly. You need the camera story to match: same depth-of-field, same direction of light, same softness. The forest bokeh creates a unified “lens look,” and that’s what makes the cover feel coherent.

Keep your palette controlled too. Green background, one pink hair accent, one blue clothing accent, and white typography. That’s enough. Extra colors will make mixed media feel messy fast.

Prompt technique breakdown (lego blocks)

Prompt chunk	What it controls	Swap ideas (EN)
“two subjects back-to-back, mirrored side profiles”	Collab readability and symmetry	“shoulder-to-shoulder”, “face-to-face”, “split frame”
“lush green forest bokeh, shallow depth of field”	Cinematic glue and softness	“night city bokeh”, “stage haze”, “sunset field”
“left anime character, right photoreal subject”	Mixed-media novelty	“both illustrated”, “both photoreal”, “3D + photo”
“consistent soft daylight lighting”	Prevents pasted look	“rim-lit sunset”, “overcast soft light”, “spotlight key”
“album cover typography: small top credit, large serif title”	Signals ‘official release’	“minimal sans title”, “no type”, “centered title block”

Starter prompt

Cinematic square album cover in a lush green forest with heavy bokeh and shallow depth of field, soft natural daylight. Two subjects stand back-to-back centered in frame with mirrored side profiles: left is an anime-style girl with pastel pink hair in a low bun with braided wrap detail, wearing a dark navy/purple jacket with an orange circular emblem; right is a photoreal young woman with medium-length brown hair wearing a blue pinafore dress with white long sleeves. Both hold hands raised in front in a gentle mirrored gesture. Mixed-media composite with consistent lighting and DOF. White typography: small spaced artist line at top, large elegant serif title across bottom reading “STORY TELLERS”. High resolution, clean, album-cover layout.

Remix steps: a fast convergence loop for mixed media

Baseline lock

Pose: back-to-back mirrored profiles
Lens look: shallow DOF + bokeh background
Type system: small top credit + big bottom title

One-change rule (example 4 runs)

Run 1: lock composition and background bokeh.
Run 2: adjust only lighting consistency across both subjects.
Run 3: adjust only style balance (more illustrated vs more photoreal) while keeping DOF fixed.
Run 4: adjust only typography placement and scale.