Ultra-Detailed AI Prompt and Hand Gesture Control Guide for Heisei Retro Japanese Portraits
A breakdown of an ultra-detailed AI prompt capturing late Heisei-era Japanese film photobook aesthetics, spatial lighting, kimono drapery, and hand gesture cont
On September 11, 2026, AI image creator 蛋黄堡 (@Hamburgerai) shared an exceptionally detailed prompt engineering workflow designed to bypass artificial studio lighting and overly polished plastic skin. Drawing inspiration from late Heisei-era Japanese private photobooks (shishashin) and vintage film aesthetics, the prompt reconstructs the subtle mood of a traditional wooden washitsu room illuminated by the warm evening sun while navigating the persistent challenges of multi-finger hand gestures in diffusion models.
Image source: @Hamburgerai via X
The newly released prompt focuses on capturing the lyrical, deeply personal atmosphere of Japanese editorial portraiture from the late 1990s through early 2000s. Rather than relying on standard commercial strobe setups or aggressive beauty retouches, the composition emphasizes architectural shadows, authentic skin texture with visible pores and downy hair, and organic fabric weight. Alongside the prompt, the creator provided practical commentary on handling one of the most notoriously fragile generation targets in current multimodal systems like GPT-Image 2.5: complex arm and finger positioning behind the subject's back.
Late Heisei Photobook Mood: Circular Sunset Lighting in a Wooden Washitsu
The spatial core of this prompt lies in rejecting artificial studio backdrops in favor of a lived-in, grounded Japanese interior.
The setting is established inside a traditional timber-framed washitsu. Authentic cultural details—tightly woven tatami flooring, delicate shoji paper sliding doors, low floor cushions (zabuton), and casually scattered patterned fabric pieces—populate the room to suggest an intimate, private moment rather than a commercial shoot. The overall environment maintains a muted, low-key ambiance dominated by dark brown cedar tones, allowing the luminous subject and her wardrobe to serve as the undisputed visual center.
The defining visual anchor of the frame is the dramatic lighting projected onto the shoji screen behind the subject. Either from the low angle of the setting sun or a warm directional continuous light, a large circular golden beam appears with soft, feathered edges, evoking the poetic shape of a harvest moon or a glowing solar disc. This circular projection provides high-contrast rim lighting that separates the subject's silhouette from the shadowy room, delivering the rich dynamic range and nuanced tonal transitions characteristic of 35mm portrait film. Shot with a 35mm to 40mm equivalent prime perspective positioned slightly below shoulder level, the framing establishes an unforced proximity that feels genuine and unguarded.
Anatomy and Fabric Contrast: Delicate Lace Against Draped Silk Kimono
The second pillar of the prompt is the uncompromising description of human anatomy paired with an intentional clash between modern and traditional textiles.
The subject is depicted as an adult East Asian woman in her twenties, defined by a soft oval face, balanced facial proportions, and gentle natural fullness along the cheeks rather than the exaggerated jawlines often favored by commercial AI presets. Her eyes are relaxed, almond-shaped, and framed by distinct natural eyelashes, complemented by a delicate, rounded nose tip and unmanipulated lip contours. Crucially, the prompt instructs the model to preserve genuine skin micro-textures—real pores, fine vellus hair, and gentle epidermal subsurface scattering—actively suppressing artificial airbrushing and 3D-rendered plastic highlights.
The styling introduces a deliberate contrast between two distinct textile weights. On the upper torso, she wears a pale mint-green lace bra featuring thin straps, narrow back bands, floral lace cups, and translucent mesh. The fabric is instructed to fit the body naturally, demonstrating physical tension, edge indentations, and micro-creases under gravitational pull. In sharp contrast, a lavish deep teal furisode kimono has slipped downward, pooling heavily across her hips, legs, and the tatami floor. Thick, lustrous silk satin embroidered with pink, coral, and white florals spills outward with realistic draping and folds, showing its white lining and red piping to emphasize the natural reality of adjusting one's attire in privacy rather than a theatrical performance.
Complex Hand Gesture Control and Model Limits: Reaching Behind the Back
Accompanying the prompt, the creator offered an honest operational assessment: "GPT2.5's control over complex hand gestures like this remains rather average."
The target pose calls for the subject to kneel or sit sideways on the tatami, twisting her pelvis slightly while rotating her upper torso three-quarters toward the camera. Simultaneously, both arms reach behind her back, with her fingers adjusting or fastening the clasp of her lingerie. In current image generation models, occluded limbs and fingers intersecting behind the torso frequently trigger severe anatomical hallucinations, including fused digits, extra knuckles, or phantom wrists melting into the skin.
To counteract these degradation patterns, the creator avoided vague posing commands and instead anchored the biomechanical kinetics across the entire upper body. The prompt explicitly specifies the tension of the scapulae (shoulder blades), the indentation of the spinal groove, the lateral muscle contraction along the waist, and the relaxed angle of the wrists. By forcing the model to generate the kinetic chain of the back and shoulders first, the limbs and fingers are far more likely to follow a plausible physical trajectory.
The complete original prompt is provided below for reference and direct application:
超真实成年东亚女性日式和室写真，整体呈现平成末期写真集、realistic japanese photography、日系私写真、轻复古胶片摄影与克制的和风 editorial 气质，画面不是商业棚拍，而像摄影师在传统榻榻米房间里利用傍晚自然光完成的一张真实人物写真；场景为木结构和室，榻榻米地面、障子纸门、低矮坐垫与花纹布料自然散落，人物身后的障子门被夕阳或暖色定向灯照出一个巨大而边缘柔软的圆形金色光斑，形成类似月轮或落日投影的视觉中心，房间其余区域保持低照度与深棕木色，让人物皮肤、浅薄荷绿色内衣和深青色花纹和服形成鲜明层次。
人物是一位明确成年的二十多岁东亚女性，脸型偏小的柔和鹅蛋脸，额头饱满，面中偏短，下颌纤细圆润而不尖，脸颊保留轻微自然肉感；眼睛为横向稍长的杏眼，眼睑放松、睫毛自然清晰，鼻梁细直、鼻尖圆润小巧，嘴唇饱满柔软，唇峰自然，侧脸轮廓干净秀气，皮肤白皙但保留真实毛孔、细小绒毛与轻微肤色起伏。深棕近黑色长发盘成低位和式发髻，厚薄适中的齐刘海自然贴额，两侧留出细碎鬓发与弧形发丝，发髻一侧插着体积较大的暗红色花朵发饰并垂下细小花瓣装饰，整体既有传统和风感又带平成写真时期的轻熟氛围。
妆容采用自然日系淡妆，柔雾底妆不过度磨皮，眼尾轻微加深，腮红集中在颧骨与脸颊中部，嘴唇为低饱和豆沙粉或玫瑰裸色并保留微弱润泽感。身材为纤细骨架与明显柔软曲线并存，肩窄、锁骨自然、上背纤薄，腰部纤细，胸部体量明显丰满丰盈，具有真实重量感与侧面投影，胸腰比例清晰但保持人体真实结构，臀胯线条圆润自然，坐姿与扭转动作让腰臀关系产生真实软组织形变。
人物穿浅薄荷绿色蕾丝文胸，细肩带、窄背带、花卉蕾丝杯面与半透网纱结构清晰可见，面料紧贴身体并形成真实承托、轻微拉伸、边缘压痕与褶皱，避免硬挺塑料质感；下半身仍包裹着一件已经松散滑落的深青绿色振袖或华丽和服，布料为厚实柔亮的丝缎质感，大面积粉红、白色、珊瑚橙与浅绿色花卉纹样分布在衣摆，白色里衬与红色滚边从腰臀附近自然翻出，和服松垮堆叠在榻榻米与腿部周围，产生大量真实重量褶皱、折痕与垂坠感，强调“脱下部分传统服饰后仍停留在私人整理衣着瞬间”的生活感，而不是夸张情色表演。
人物背对镜头跪坐或侧跪在榻榻米上，骨盆略向一侧转，腰背形成明显S形曲线，上半身向左后方扭转约四分之三角度，左肩更靠近镜头，头部再向后转成清晰侧脸轮廓，眼神没有直接挑逗镜头，而是越过摄影者斜上方停留，眼睑略垂，嘴唇微张，神态安静、自持、略带疏离和亲密距离中的暧昧感；双臂绕到背后，手指正在整理或扣合文胸背扣，手腕与手指姿态自然，背部肩胛骨、脊柱沟和腰侧因为动作产生真实受力关系。
采用接近人物的35mm镜头或约40mm等效焦段，竖幅构图，机位略低于人物肩部并处于腰背后方近距离拍摄，人物从臀部至发髻完整入画；光影兼具傍晚自然漫射光与后方障子门透出的柔和金色圆光，形成细腻的边缘轮廓光与深色和室环境的明暗过渡，35mm胶片颗粒质感细腻，色彩层次丰富且克制。
Original source
- Original post: 蛋黄堡 (@Hamburgerai) on X