A family New Year’s feast unfolds through close-ups of pork, fish, greens and dumplings being cooked and served in a cinematic food video

Prompt

Theme:
Based on the user-provided photo of a New Year’s Eve family dinner, create a 15-second photorealistic food documentary short film with a cinematic look, in the style of 《舌尖上的中国》. Widescreen 16:9, 24fps, cinematic color grading and fine grain, primarily macro and close-up shots, shallow depth of field, realistic steam and oil sheen.

The scene is a northern Chinese family’s New Year’s Eve dinner kitchen and dining table, based on the photo: warm lighting, a wooden dining table, and white porcelain plates and bowls. The cook’s hands and forearms may appear (show hands only, no face); movements should be realistic and restrained.

Edit as 8 short shots with hard cuts, each lasting 1.5–2 seconds. The final shot must show all three dishes and a plate of dumplings served on the table.

Shot 1: Based on the photo, aromatics sizzle in hot oil: fine bubbles and rippling patterns across the surface of the oil, a light veil of steam rising, clean highlights along the rim of the wok, with a clear “滋滋” sizzling sound.

Shot 2: Red-braised pork is added to the wok and stir-fried until browned; the gelatinous texture of the skin is clearly visible, the oil sheen looks realistic, with a light scraping sound from the spatula.

Shot 3: Based on the photo, fish in spicy broth: the lid is lifted from the steamed fish, the flesh quivers slightly, its aroma seems to rush toward the viewer, and steam billows.

Shot 4: Sichuan peppercorns and dried chilies hit the oil; particles and fibers are clearly visible, splattering is restrained, and the aroma feels vivid.

Shot 5: Before the fish in spicy broth is served, red chili oil is poured over it, flowing in ribbons; tiny droplets of oil spatter where it lands as layers of hot steam billow.

Shot 6: Based on the photo, leafy greens are added to the wok and stir-fried; vivid green, crisp and tender, with a thin layer of wok steam rising, and clean, brisk stir-frying movements.

Shot 7: Dumplings are lifted from the pot; large and small bubbles are clearly visible in the boiling water, droplets cling to the dumplings, and steam rushes toward the camera as the lid is lifted.

Shot 8: Three dishes and a plate of dumplings are served on the table: a plate of glossy red-brown red-braised pork, a large plate of fish in spicy broth with roiling red oil, a plate of fresh, vibrant green stir-fried vegetables, and a plate of plump, smooth, white dumplings; the camera pushes in slightly and holds for 0.5 seconds as the steam rises in layers like a breath.

Audio:
Music is the highest priority. The soundtrack must be documentary music with the feel of 《舌尖上的中国》: warm and restrained, with a distinctly Chinese sensibility and an atmosphere created by strings, woodwinds, and light percussion;
at the same time, include clear, realistic cooking sound effects (aromatics sizzling, sauce bubbling as it reduces, boiling, the spatula, steam as the lid is lifted). Audio must not be omitted.

Voiceover VO is a high priority: add quiet, restrained Chinese documentary narration (not in an advertising style), using short phrases to describe “heat, steam, reunion, and time.” Do not include brands, the program title, or specific place names.

主题：
根据用户提供的年夜饭照片，生成一段15秒电影级写实美食纪录片短片，风格参考《舌尖上的中国》，画面宽银幕16：9，24fps，电影级调色与细腻颗粒，微距与近景为主，浅景深，真实蒸汽与油光。

场景根据照片中的北方家庭年夜饭厨房与餐桌场景，包含暖色灯光、木质餐桌与白瓷盘碗。允许出现厨师的手与前臂（只拍手，不拍脸），动作真实克制。

剪辑为8个短镜头硬切，每镜头1.5–2秒，最后必须三菜一饺子齐上桌。

镜头1：根据照片中的热油爆香，油面细密气泡与油纹波动，薄蒸汽上升，锅沿高光干净，听到清晰“滋滋”声。

镜头2：红烧肉下锅翻炒上色，肉皮胶质细节清晰，油光真实，锅铲轻刮声。

镜头3：根据照片中的水煮鱼，蒸鱼开盖，鱼肉微微抖动，香气扑面而来，蒸汽翻涌。

镜头4：花椒与干辣椒入油瞬间，颗粒纤维清晰，飞溅克制，香气感强。

镜头5：水煮鱼出锅前红油淋下，形成丝带状流动，落点溅起微小油滴，热气翻层。

镜头6：根据照片中的青菜，青菜下锅煸炒，翠绿脆嫩，锅气蒸汽薄层上涌，翻炒节奏干净利落。

镜头7：饺子出锅，沸水大泡小泡层次清晰，饺子表面挂水珠，揭盖蒸汽扑面。

镜头8：三菜一饺子端上桌：红烧肉一盘油亮红棕，水煮鱼一大盘红油翻涌，清炒青菜一盘清爽翠绿，饺子一盘白润饱满；镜头微推定格0.5秒，蒸汽像呼吸一样翻层。

音频：
音乐权重最高，必须生成《舌尖上的中国》气质的纪录片配乐：温润克制、具有中国气韵的弦乐/木管/轻打击乐氛围；
同时必须有清晰真实的烹饪拟音（爆香滋滋、咕嘟收汁、沸腾、锅铲、揭盖蒸汽），音频不可缺失。

旁白VO权重高：加入低声克制的中文纪录片旁白（非广告腔），用简短句描述“火候、热气、团圆与时间”，不出现品牌、不出现片名、不出现具体地名。

• Needs your photo: attach an image to the request — the prompt works with it.
• The prompt is in Chinese: paste it as is — the model understands it.