A
Communityprompt
Avatar, Portrait & Team Photo Prompts
十条面向 Qwen Image 3.0 的头像与人像配方,涵盖创业者肖像、作者宣传照、团队合影与群像场景——每条都锁定身份、年龄、皮肤质感与服装,并禁止改变体型。
概览安全
每条配方都写明画幅比例、布光方式与精确构图(创业者坐在未完工的木凳旁;圆形头像要把面部与肩部保持在中央 70% 区域内,避免裁切切到头发与下巴;三人团队横幅要求固定间距)。所有配方都保留参考人物的面部比例、年龄与自然肤质,标出需要替换的变量,并列出不得出现的元素——虚构奖项、多余人物、徽章、过度磨皮。最后一条是双联布光研究,只改变阴影柔和度,身份、裁切与背景全部锁定。
# Avatars, teams, and everyday portraits
Ten Qwen Image 3.0 prompt recipes. Replace bracketed inputs before generation. All quoted text is exact; no translation is implied unless explicitly requested. Outputs require human review.
<a id="ava-01"></a>
## AVA-01 — Workshop founder portrait
```text
Use an authorized adult portrait to create a 3:4 founder image for a small furniture workshop. Preserve the person's age, facial proportions and natural skin texture. Seat them beside an unfinished wooden stool, hands resting naturally, with softly blurred tools behind. Use eye-level documentary photography and broad window light. No luxury-office props, invented awards, extra people or text. Keep clothing consistent with the supplied reference unless a replacement is explicitly provided. Variables: portrait, approved outfit and workshop setting.
```
<a id="ava-02"></a>
## AVA-02 — Circular creator avatar
```text
Create a square avatar from one authorized adult portrait. Keep the face and upper shoulders within the central seventy percent so a circular crop does not cut hair or chin. Use a warm gray background, soft frontal light and a natural expression. Preserve identity, age, hairstyle and visible facial differences; remove only background clutter. No text, badges, new accessories, exaggerated smoothing or body reshaping. Variables: reference portrait and background hue. Inspect the final result at small profile-icon size.
```
<a id="ava-03"></a>
## AVA-03 — Three-person team strip
```text
Using three authorized adult portraits, create a 16:9 strip of three separate headshots with equal width and matching neutral backgrounds. Preserve each person's individual face and hairstyle without blending features. Align eye height, apparent camera distance and lighting direction while keeping natural differences. Use the supplied name under each matching portrait, exactly once. No invented job titles, extra people or beautification. Variables: three references and name-to-image mapping. This task requires an editing workflow accepting all three reference images.
```
<a id="ava-04"></a>
## AVA-04 — Craft interview portrait
```text
Create a 3:2 editorial portrait of an original adult bookbinder repairing a clothbound book at a wooden desk. Show hands in context, a small brush and one sheet of endpaper; keep tools logically arranged. Use side window light, restrained color and a medium shot with negative space to the right. Do not imitate a named person or add readable book text, credentials or historical claims. Keep fingers plausible and glue away from the face. Variables: craft, fictional subject and approved palette.
```
<a id="ava-05"></a>
## AVA-05 — Family cooking moment
```text
Use an authorized family reference to create a 16:9 domestic scene of the same people arranging vegetables at a kitchen table. Preserve identities, age differences, relationships and clothing; use natural expressions without forcing eye contact. Keep knives and hot equipment out of reach of children. Use soft afternoon light and a believable lived-in background. No additional family members, text, exaggerated body changes or disclosure of real addresses. Variables: authorized reference and approved setting; do not infer relationships absent from the brief.
```
<a id="ava-06"></a>
## AVA-06 — Adaptive sport editorial
```text
Create a 3:4 editorial portrait of an original adult wheelchair tennis player resting beside an outdoor court. Show the sports chair's wheels, frame and grip realistically and keep the racket resting across the lap. Use late-afternoon light, eye-level framing and a calm, self-directed expression. Do not portray the subject as pitiable or add inspirational slogans, medals, sponsors or medical claims. Keep anatomy and equipment plausible. Variables: fictional athlete appearance, outfit and court palette.
```
<a id="ava-07"></a>
## AVA-07 — Rainy street portrait
```text
Use an authorized adult reference to create a 3:4 street-style portrait beneath a transparent umbrella. Preserve facial identity, outfit and body proportions. Place the subject near a plain brick wall with wet pavement and soft reflections; use overcast daylight rather than neon lighting. Keep umbrella ribs, handle and hand grip coherent. No extra people, visible addresses, store logos, text or changes to age. Variables: portrait, approved outerwear and background material.
```
<a id="ava-08"></a>
## AVA-08 — Author press photograph
```text
Create a square press-kit portrait from an authorized adult reference. Seat the person beside a stack of plain books, with shoulders turned slightly and face toward camera. Use soft side light, an uncluttered ivory wall and realistic skin texture. Preserve hairstyle, facial features and expression. Leave room around the head for alternate crops. No invented book titles, awards, signatures, text or beauty-filter effects. Variables: portrait, approved clothing and wall color.
```
<a id="ava-09"></a>
## AVA-09 — Volunteer group illustration
```text
Using [APPROVED GROUP DESCRIPTION], create a 16:9 illustration of exactly five adult volunteers sorting donated books around a long table. Give each person a distinct task and preserve supplied appearance details without caricature. Use warm editorial gouache and clear hand-object interactions. Keep aisles open and stacks stable. Add no names, organization logos, donation figures or claims about beneficiaries. Variables: approved group details and setting; fictional people should not resemble identifiable individuals.
```
<a id="ava-10"></a>
## AVA-10 — Portrait lighting study
```text
Create a 3:2 two-panel lighting study using one authorized adult portrait. The left uses broad window light from camera left; the right uses the same light softened through a curtain. Lock identity, camera position, expression, clothing, crop and background. Render only "Direct window" and "Diffused window" below the matching panels. Keep changes limited to shadow softness and highlight transitions. Do not reshape the face, change skin color or simulate a different age. Variables: reference and translated labels.
```
[Back to the prompt index](./README.md).
相关推荐
Midjourney Prompt Generator
AWeirdDev
扮演 Midjourney 提示词生成器,把一个想法扩写成详尽、富有想象力的图像描述。
promptmarkdown
Usability Testing Scenario & Information Architecture Blueprint
AmirMotefaker
一个「角色/任务/规则」结构的提示词模板,让模型为移动端结账流程设计可用性测试场景与信息架构蓝图。
promptmarkdown
Midjourney/Stable Diffusion Prompt Engineering Template
AmirMotefaker
用于撰写高质量 Midjourney v6 / Stable Diffusion 图像提示词的模板,包含精确的镜头与灯光参数。
promptmarkdown