I’m nervous lately 🫢 I want people to understand me more, so I’m going to start a YouTube. Subscribe to me there ! Everyone commenting on the @ted video, come talk to me and we can be friends 🤝 緊張してるあたし🫨 さて、近々YouTube始めます。是非登録しておいてねっそして、TEDにコメントしてる皆さん、あたしと直接話しましょう笑🤝 #ted #youtube #imma #あたしCGらしい #ithinkimcgi

imma

@imma.gram · Digital creator

INSTAGRAM · 2024-09-22Source

-1likes

20comments

Remix This

Prompt

[HIGH-GRANULARITY INVENTORY]

Subject(s)
- Count: 1 female-presenting CG character (virtual influencer / 3D human)
- Pose: upper-body visible, turned in profile toward the right; mouth slightly open mid-speech
- Expression: nervous/concerned but composed; soft gaze forward
- Hair: blunt chin-length bob, straight, glossy; saturated bubblegum-pink color; full straight bangs
- Facial details: smooth CG skin, subtle blush, defined lips with pink lipstick; visible teeth slightly; small nose; neutral brows

Clothing & materials
- Top: white ringer T-shirt with black ribbed collar and black sleeve cuffs
- Graphic: large printed collage of black-and-white faces with bold red text across the chest (readable “Imma …” style), additional smaller red/black text elements
- Fabric: cotton jersey, slightly fitted; minimal wrinkles

Props/objects
- No handheld props
- No on-image captions; design interest comes from environment + shirt graphic

Environment
- Monochrome pink interior set
- Architectural elements: rounded/arched doorway opening on the right; curved corridor/room beyond
- Background: smooth matte pink walls and floor; minimal decor, studio-like set design

Composition
- Horizontal frame (landscape)
- Medium shot from chest up; subject placed left-of-center, facing right
- Camera height: around chest/eye level
- Angle: side profile 3/4 from the left, emphasizing silhouette and hair shape
- Negative space: large open area to the right with the archway, clean minimal composition

Lighting
- Soft studio lighting, low contrast
- Gentle key light from front-left, smooth falloff; minimal harsh shadows
- Color temperature: neutral; overall scene tinted pink by environment

Color palette
- Dominant: pink (walls/floor/background)
- Secondary: white shirt
- Accents: black trim and red/black graphic print; natural skin tones

Image style
- High-quality 3D render of a virtual human (semi-realistic)
- Clean shading, smooth skin, subtle specular on hair
- Minimal grain, crisp edges, slightly soft cinematic look


[MASTER PROMPT]
[Subject]
High-quality 3D render of one female-presenting virtual influencer, semi-realistic CG human, bubblegum-pink blunt bob haircut with straight bangs, smooth skin, subtle makeup, mouth slightly open as if speaking, nervous but gentle expression. Side-profile pose facing right.

[Wardrobe]
White ringer T-shirt with black collar and sleeve cuffs, fitted cotton jersey. Large chest graphic print with black-and-white faces and bold red text (streetwear magazine vibe).

[Environment]
Minimal monochrome pink interior set: smooth matte pink walls and floor, rounded arched doorway on the right leading to a curved corridor, no clutter.

[Composition/Camera]
Landscape frame, medium chest-up shot, subject placed left-of-center looking right, plenty of negative space and the archway visible on the right. Eye-level camera, clean silhouette.

[Lighting]
Soft studio lighting, gentle key from front-left, low contrast, smooth shadows. Scene naturally tinted pink from the environment.

[Style/Rendering]
Premium 3D character render, crisp but slightly soft cinematic finish, detailed hair specular, clean textures, no noise.

[Detail constraints]
Only one character. Keep the environment purely pink with the arch doorway. Do not add furniture, props, text overlays, logos/watermarks, or extra characters.


[NEGATIVE PROMPT]
extra people, furniture, posters, clutter, neon signs, text overlay, watermark, logo, distorted face, asymmetrical eyes, uncanny teeth, melted hands, plastic wax skin, heavy pores, harsh shadows, dramatic rim light, film grain, low-res, anime cel shading, photoreal background, outdoor scene


[SUGGESTED PARAMETERS]
- Aspect ratio: 16:9 (landscape)
- Lens feel: 35–50mm
- Depth of field: moderate (background still readable but soft)
- Steps: 28–45
- CFG / guidance: 5–7
- Sampler: DPM++ 2M Karras (or equivalent)
- Seed: 20240922


[DELTA PROMPT STRATEGY — 10 DRIFT FIXES]
1) If the palette stops being monochrome: “entire room is matte monochrome pink, pink walls and floor, minimal set”
2) If hair shape changes: “blunt chin-length bob, straight bangs, bubblegum-pink glossy hair”
3) If shirt becomes plain: “white ringer tee with black trim, large red/black graphic print on chest”
4) If the arch disappears: “rounded arched doorway on the right, curved corridor beyond”
5) If camera angle becomes frontal: “side-profile 3/4 view from left, subject facing right”
6) If lighting becomes dramatic: “soft studio lighting, low contrast, gentle shadows”
7) If skin looks too plastic: “natural CG skin shading, subtle texture, not waxy”
8) If extra props appear: “empty room, no furniture, no decor, no props”
9) If background becomes too sharp: “slightly soft background, keep focus on subject and hair silhouette”
10) If expression becomes wrong: “nervous but gentle expression, mid-speech mouth slightly open”

How imma.gram Made This Virtual Influencer AI Portrait — and How to Recreate It

This image is doing something subtle: it looks like a scene from a minimalist film, but it’s really a conversation starter. A virtual human in a single-color world feels instantly recognizable, and the side-profile “mid-sentence” mouth makes the post feel like a moment you caught—not a posed selfie. That’s exactly the vibe you want when your caption is emotional and personal.

The pink set is not just aesthetic. It’s a control system. When everything is one color, the audience stops scanning the background and starts reading the face. The graphic ringer tee adds a touch of everyday realism, which keeps the character from feeling like a fashion render. You get futurism (CG person) with familiarity (casual shirt), and that blend is what makes people lean in.

Why it spreads (signals you can copy without copying the character)

Signal	Evidence (from this image)	Mechanism	Replication Action
Controlled world	All-pink room with a single archway, no clutter	Less visual noise = more attention on expression and message	Lock a monochrome set (one hue) and explicitly forbid furniture/props
“Caught mid-thought” energy	Side profile, mouth slightly open like speaking	Reads as candid, which pairs well with vulnerable captions	Prompt “mid-speech, mouth slightly open, soft nervous expression”
Iconic hair silhouette	Blunt pink bob + straight bangs	Memorability: the audience recognizes the post in a scroll	Give your character one unmistakable silhouette (hair/hat) and keep it constant
Everyday grounding detail	White ringer tee with black trim and a bold graphic	Balances the “CG” feel with streetwear familiarity	Add one casual wardrobe cue (ringer tee, hoodie) to reduce “sterile render” vibes

Best-fit scenarios

Channel launches: announcements feel more human when the image looks like a moment, not a poster.
Vulnerability posts: nervous/excited updates land better with soft lighting and a calm set.
Virtual persona building: consistent palette + hair silhouette becomes your signature.
Creator community invites: “come talk to me” posts perform when the visual tone is approachable.

Not ideal

High-info education: this frame is mood-first; it won’t carry dense text overlays.
Product detail shots: a character-led composition will steal attention from the product.
Hard comedy: the minimal art direction can read too serious unless you add a deliberate twist.

Transfers (exactly 3)

Recipe 1: New persona, same art direction
- Keep: monochrome room, side-profile mid-speech pose, soft studio light
- Change: character identity, hair color, shirt graphic theme
- Slot template: “{virtual_persona} in a monochrome {color} room, side-profile mid-speech, wearing {casual_top_with_graphic}”
Recipe 2: Color-series system
- Keep: same camera angle and set geometry (archway), same wardrobe silhouette
- Change: room color per emotion (pink = nervous, blue = calm, red = bold)
- Slot template: “same scene layout, monochrome {emotion_color} set, expression: {emotion}, caption tone: {announcement}”
Recipe 3: Swap the ‘mid-sentence’ beat
- Keep: minimal set, negative space, soft lighting
- Change: micro-expression (surprised, shy smile, thoughtful) and mouth position
- Slot template: “{persona} in monochrome room, {micro_expression}, {mouth_position}, side-profile composition”

Aesthetic read: what makes it feel premium instead of ‘random 3D’

The trick is that the environment is designed like a brand identity: one hue, one architectural motif, and nothing else. That archway creates depth without adding objects. The lighting is gentle, so the face reads first. And the shirt is intentionally ordinary—if everything looked futuristic, the audience would keep their distance.

If you’re building a virtual creator format, consistency beats novelty. Use one camera angle, one room geometry, one palette, and one hairstyle silhouette as your “signature.” Then let the caption carry the story changes.

Observed	Evidence in the image	Recreate instruction (prompt knob)
Monochrome set discipline	Walls + floor are the same pink; no decor	“empty monochrome {color} room, no furniture, matte surfaces”
Depth via architecture	Rounded archway and curved corridor	“arched doorway, curved corridor, minimal geometry”
Signature silhouette	Blunt bob + straight bangs is instantly recognizable	“iconic blunt bob haircut with straight bangs, glossy hair specular”
Soft lighting, low contrast	No harsh shadows; skin reads smooth	“soft studio key light, low contrast, gentle shadow falloff”
Grounding wardrobe cue	Ringer tee with a bold print breaks the pink field	“white ringer tee with black trim, bold graphic print, cotton jersey”

Prompt technique breakdown (treat it like a controllable scene)

Prompt chunk	What it controls	Swap ideas (EN, 2–3 options)
Character realism level	Whether it reads as premium CG vs uncanny	“semi-realistic CG human”, “stylized 3D doll look”, “hyper-real virtual model”
Hair silhouette	Recognizability and brand memory	“blunt bob with bangs”, “high ponytail”, “short pixie cut”
Set palette	Brand identity and noise control	“monochrome pink room”, “monochrome blue room”, “monochrome cream room”
Set geometry	Depth without props	“arched doorway”, “curved hallway”, “rounded window cutout”
Pose micro-beat	Candid vs posed energy	“mid-speech mouth slightly open”, “half-smile”, “thoughtful pause”
Wardrobe anchor	Relatability and contrast	“ringer tee”, “oversized hoodie”, “simple tank top”

A prompt skeleton to start from

premium 3D render of {virtual persona}, {iconic hair}, side-profile mid-speech,
wearing {casual top}, empty monochrome {color} room with an arched doorway,
soft studio lighting, low contrast, clean cinematic look

Remix steps (converge fast)

Baseline lock

Palette: commit to one set color and keep every surface inside it.
Camera: keep the same side-profile angle and framing.
Signature feature: lock the hairstyle silhouette before you experiment with anything else.

One-change rule

Change one knob per run: expression OR set color OR wardrobe. You’ll keep the “system” intact while learning what actually moves engagement.

Example 4-step iteration sequence

Run 1: match the pink-room template with a ringer tee and mid-speech pose.
Run 2: change only the set color to blue and keep everything else identical.
Run 3: change only the expression (shy smile) and keep the set locked.
Run 4: change only the wardrobe silhouette (hoodie) while preserving the camera angle.