Lab / Templates / Cinematic Motion Shot Template

Cinematic Motion Shot Template

1 张真人照 + DSLR cinematic prompt → 9:16 全身动态镜头 · 身份保持 + 显式 negative prompt + face lock 直接指令 · golden hour 户外

项目内首验证 2026-05-01 · yuan_kid 4yo + viral cinematic motion prompt (9:16 全身 spinning fountain). 4 unique test 维度全 PASSED: (1) **9:16 aspect ApiMart · FIRST validation** · 直接 accept · 无需 fallback to 4:5 → unlocks 垂直 Reels/TikTok/Shorts 格式 on gpt-image-2 → 之前只验过 1:1/4:5/4:3/2:1 · 9:16 现入 valid set (2) **Negative prompt syntax · STRONG honor** · 'painting/illustration/CGI/plastic skin/...' 反向 anchor 全有效 → gpt-image-2 不像 SDXL 强 negative · 但 显式 'negative prompt:' prefix 有效降低相应 artifact → 新 prompt-engineering primitive · 适合 DSLR / photoreal 场景作 安全网 (3) **Face lock vs 'young woman' anchor 冲突 · FACE LOCK WON** · 4yo ref + 'young woman' subject anchor + 'do not alter face/hairstyle' directive · 结果: identity 保持 (East Asian girl · 略 age-compensate 至 ~7yo · 但仍是 child · NOT young woman) · OPPOSITE of porcelain test (where 'young woman' beat 'use for structure only') → **新 rule**: 显式 imperative ('do not alter face') 强于 descriptive constraint ('use for structure only') → directive 强度 > anchor 数量 (4) **Photoreal aesthetic** · DSLR-quality 完整达成 · golden hour rim glow + water droplets mid-air + 85mm bokeh + skin pores visible + 自然 motion blur on dress edges + wet fabric edges sticking Identity-preserving family · sister templates-business-headshot (1:1 商务化 · 同 'directive 直接指令' 范式) + sister templates-porcelain-sculpture-portrait (medium-transform · 反例参考).

observed_once · ¥0.5 · aspect 9:16 · providers · gpt_image_2 · id · templates-cinematic-motion-shot
Cinematic Motion Shot Template
sample · experiments/cinematic_motion_test/cinematic_motion_v1_yuan_kid_9x16.png
Run Record
2026-05-01T07:49:34+08:00
✓ pass¥0.541.3s wall· 2.3MB
provider · gpt_image_2 (apimart) · aspect_ratio=9:16 · size=9:16 · n=1
prompt · inline (no library entry · 1878 chars)
task_id · task_01KQGCQBP53XBS157NVESXA7BK
script · experiments/cinematic_motion_test/test_v1_yuan_kid.py
view full sidecar yaml
date: '2026-05-01T07:49:34+08:00'
result: pass
prompt:
  text: >-
    ultra-realistic cinematic photograph, 9:16, full body dynamic shot, a young woman joyfully
    spinning in shallow water near a fountain, body slightly turned mid-motion with one hand lifting
    the flowing dress while the other moves naturally, soft joyful smile with relaxed expression,
    face lock: preserve exact facial features and identity, do not alter face or hairstyle, natural
    proportions, skin: soft natural skin with visible pores and smooth tonal transitions, subtle
    micro-texture with natural light reflection, hair: long loose wavy hair, slightly damp with
    dynamic motion, fine strands catching backlight glow, no over-styling, clothing: lightweight
    flowy dress in yellow and pink color combination, soft semi-transparent fabric with subtle
    floral texture, fabric reacting naturally to spin and water resistance, edges lifting with
    motion and partially sticking due to moisture, environment: outdoor fountain during golden hour,
    shallow water splashing around feet with droplets suspended mid-air, water ripples expanding
    outward in concentric patterns, background lush greenery softly blurred with strong sun rays
    piercing through leaves, lighting: warm golden hour backlight creating strong rim glow around
    hair and dress, soft frontal bounce light illuminating face, sparkling highlights on water
    droplets, subtle lens flare and light diffusion for cinematic depth, color: rich warm tones with
    golden highlights and natural contrast, lens: 85mm shallow depth of field with subject tack
    sharp and background creamy bokeh, realism: DSLR-quality, ultra-detailed, photorealistic,
    natural motion blur only on fastest moving edges, no stylization


    negative prompt: painting, oil painting, illustration, CGI look, plastic skin, over-smooth face,
    stiff fabric, frozen unnatural motion, oversharpening, flat lighting, distorted anatomy, extra
    limbs, blur, noise, text, watermark
refs:
  - source: yuan.studio/public/images/yuan_kid_history/raw_frame_001.png
    sha256: null
    error: file not found at write time
    upload_url: >-
      https://personalized-video-refs.tos-cn-beijing.volces.com/4a21e4ba8a91/raw_frame_001.png?X-Tos-Algorithm=TOS4-HMAC-SHA256&X-Tos-Credential=AKLTYjM3ZTcxNjI5ZjFkNGZmYjgyMTNjNWRjZjU4N2IzNjY%2F20260430%2Fcn-beijing%2Ftos%2Frequest&X-Tos-Date=20260430T175741Z&X-Tos-Expires=86400&X-Tos-SignedHeaders=host&X-Tos-Signature=e00dec5d3246089dde91813f6c7914ef6c2b1d22eaca22aff4207093249a22d0
provider:
  id: gpt_image_2
  relay: apimart
  config:
    aspect_ratio: '9:16'
    size: '9:16'
    'n': 1
output:
  path: ./cinematic_motion_v1_yuan_kid_9x16.png
  bytes: 2458006
  wall_seconds: 41.3
  task_id: task_01KQGCQBP53XBS157NVESXA7BK
script: experiments/cinematic_motion_test/test_v1_yuan_kid.py
cost_yuan: 0.5
notes: >
  3 unique test results · all PASSED:


  (1) 9:16 aspect ApiMart · FIRST validation · accepted directly (no fallback to 4:5)
      → unlocks vertical Reels/TikTok/Shorts formats on gpt-image-2

  (2) Negative prompt honor · STRONG effect · model honored:
      · no painting/CGI look (DSLR-realistic)
      · natural skin (no plastic) · pores visible
      · dynamic fabric (not stiff)
      · warm lighting (not flat)
      · no text/watermark
      → gpt-image-2 DOES respect negative prompt syntax

  (3) Face lock vs 'young woman' conflict · FACE LOCK WON (yuan_kid identity preserved · ~7yo
  apparent age)
      · slight age compensation (4yo ref → looks 7-8) but still child · NOT young woman
      · 'do not alter face or hairstyle' directive overrode subject anchor
      → OPPOSITE of porcelain test (where 'young woman' beat 'use for structure only')
      → directive strength wins · explicit imperative ('do not alter') beats descriptive ('use for structure only')

  (4) Photoreal aesthetic · DSLR-quality achieved · golden hour rim glow on hair · water droplets
      mid-air · 85mm bokeh background · skin pores visible · natural motion blur on dress edges
prompt not yet extracted to library
Verbatim prompt lives in the experiment script. Pending migration to recipes/image_gen/gpt_image_2/prompts/.

Cross-Reference

methodology · — no dedicated recipe doc · referenced in memory only —
axis · §Templates · 参数化 prompt 产品 · 用户填占位符 / 加 ref · 商业化路径明确 · 商务头像 / NYC 旅行 / GRWM 等

Sister capabilities · same axis

NYC Travel Vlogger · No-Ref (Mood Vibe)
observed_once
9-panel mood-driven travel collage · pure T2I · 模型造虚构 vlogger · creative pre-viz / mood vibe / 灰区"假旅行账号" 用途 · JSON-structured prompt · landmark precision · polished influencer aesthetic (NOT phone-diary)
NYC Travel Vlogger · Personal (with Ref)
observed_once
1 张真人照 + JSON multi-scene prompt → 9-panel "你在 NYC 旅行" 海报 · identity 跨 9 panel 保 + landmarks 9/9 准 + NO age-up · personal branding / LinkedIn / 母亲节礼物 / 童装 marketing
Porcelain Sculpture Portrait Template
observed_once
真照 + selective identity copy prompt → 3D relief mixed-media 雕塑 (porcelain + rose-gold + pearl filigree) · 显式 disentangle structure-vs-pose · gallery / gift / luxury brand collateral
Surreal Mural Portrait Template
observed_once
复合 prompt (subject + 超现实 merge + 环境 mural + cursive 文字) → suburban 墙上的写实壁画 · subject 融入 environment · 街景 + 玫瑰 + 泡泡糖 + Instagram-ready
Pet Crayon Poster Template
observed_once
宠物名 + 复合 prompt → 蜡笔/naive sketch 风海报 · centered + warm yellow + vintage print + 中文 title · 礼物/SNS 头像/宠物 marketing
Photo Annotation Overlay · JP Vlog Style
observed_once
真照 + 中文 annotation rules → 源照 100% 保留 + 手绘 doodle 注解 + 日记 tone 短句 + Japanese-kawaii aesthetic · 小红书/RED viral vlog 模板
Oriental Ink Wash Painting Template
observed_once
东方水墨 template (含 [SUBJECT][COLOR1][COLOR2] 占位符) → 真 水墨画 · 自动补印章 + 落款诗 + 远山 + 松 · museum-quality · 工艺品 / 装饰画 / 文化 marketing
Watercolor Landscape Template
observed_once
风景 prompt → loose expressive watercolor 风景画 · 多元素平衡 (建筑+自然+海+天空) · gallery-worthy 油画质感 · 印刷品/家居装饰/旅游 marketing
Pencil Sketch Portrait Template
observed_once
JSON-structured prompt → 真 graphite-on-paper portrait · 7 anchor 全 honor (subject/pose/hair/clothing/medium/strokes/paper grain) · production-grade illustration · gift / commission / SNS / brand
One-Line Gesture Redraw Template
observed_once
单 ref portrait → 单条连续黑线 (不抬笔) · loose / 负空间大 / 极简姿态画 · pack avatar-redraw-2026q2 variant
MS Paint Stick Figure Redraw Template
observed_once
单 ref portrait → 童画 stick-figure / MS Paint mouse-drawn 感 · 跟 pathetic 同祖 但显式 stick limbs · pack avatar-redraw-2026q2 variant
Torn Paper Collage Redraw Template
observed_once
单 ref portrait → 撕边彩纸拼贴 · 不齐叠层 · 跨 medium 物理材质 · 出片 gallery-quality · pack avatar-redraw-2026q2 variant
Notebook Margin Doodle Redraw Template
observed_once
单 ref portrait → 圆珠笔走神课堂涂鸦 + 自动生成 annotation/arrows/side jokes · 模型 invent text · pack avatar-redraw-2026q2 variant
Pasta-Line Flat Lay Redraw Template
observed_once
单 ref portrait → spaghetti pieces 拼出轮廓 · 跨 medium 物理材质 · sparse gap-heavy · pack avatar-redraw-2026q2 variant
Whimsical Children's Book Redraw Template
observed_once
单 ref portrait → 童趣手绘绘本风 · 黑线轮廓 + 柔和 pastel 色块 + 抽象背景装饰 · 完整 DNA prompt · pack avatar-redraw-2026q2 variant
Pathetic Redraw Template
observed_once
1 张真人照 + viral "pathetic / clumsy / MS Paint" prompt → 刻意烂的儿童画版你 · meme / 自嘲 / SNS 分享 · viral GPT Image 2 trend
Business Headshot Template
observed_once
1 张真人照 + 模板 prompt → 商务化档案头像海报 · 身份保持 + 自动翻译 (中文姓名→EN) + 信息条 (姓名/职位/机构) · 1:1
High-Angle Fisheye Fashion Portrait Template
observed_once
pure T2I 5-item wardrobe + 高角度 fisheye + studio 渐变 → 杂志封面级 editorial streetwear · 无 ref · 1:1
Function-Syntax Weighted 2×2 Creature Variations Template
observed_once
Function-style prompt + 数字权重 layer + 跨域 translation + meta-prompting → 2x2 grid 4 个 creature variants 带 per-cell label · 一次出 4 个 viral concept
Real Human + Named Anime IP Photo Template
observed_once
Real-human ref (face lock) + 命名 anime IP (e.g., Ninja Hattori) 同框 · model 智解 stylized 为 mascot suit performer · "我和 X" 周边/SNS · 1:1
Pastel Cyberpunk Portrait · Anti-Neon Template
validated_2plus
Anti-cyberpunk-cliché · 暖 pastel 调 + 显式 'no neon' negative + gallery aesthetic · 高端品牌 / 时尚 / 二次元 IP 现实化 · 4:5 portrait
Selective Color Portrait with Heavy Brand Text Template
observed_once
Selective color (face / frames in color · rest B&W) + 拥挤 brand-heavy 背景 + 6+ readable text strings + advanced "object-as-mask" 概念 · pure T2I · 1:1
Lego Brick Concept Poster Template
observed_once
主题占位符 → 原创积木风格字体 + 概念视觉隐喻 + 多 zone 信息图 (字体样本 + 模块结构 + 颜色图例 + 细线标注) + 高级克制非玩具感 · 收藏级 design exhibition poster