From Random Ai Dance Generations To Polished K-Pop
提示词
From random AI dance generations to polished K-pop performances This might be one of the best workflows for AI creators right now. Start by building a consistent AI influencer, create a solid first frame, turn it into a 16-frame storyboard, then use Seedance 2.0 to generate smooth, perfectly synchronized dance videos. No more dealing with faces changing every few seconds, outfits randomly switching, or the choreography falling apart halfway through. Full tutorial and every prompt you need below. Composition: a 4:5 frame with the subject placed just off center, head and shoulders given generous room, and the background kept simple so the face carries the image. The camera sits at eye level with a short telephoto feel, the shoulders angled slightly while the gaze meets the lens, and the edges of the frame cropped tight enough to feel intimate without crowding the subject. Lighting: golden-hour light angled low through a window, laying a warm wash across the cheek and hair while the background sinks into a cooler, darker tone. The contrast is gentle and flattering, the skin keeps its real texture with visible pores and fine lines, and a subtle rim of light separates the hair from the backdrop. Materials and mood: skin rendered with honest texture rather than smoothing, fabric showing its weave — soft knit, crisp cotton, brushed wool — and hair kept as separate, glossy strands instead of a single mass. The palette stays warm and restrained, the mood quiet and assured, and the whole frame feels like a portrait taken on a good afternoon rather than a studio setup. Style anchor: editorial portrait photography with honest skin texture, natural color, and a quiet, confident mood — polished but never plastic. Details: fine strands of hair catching the light at the temples, the subtle texture of the knit or cotton, a small piece of jewelry or a collar detail that gives the eye a second place to rest, and a background kept to one or two soft shapes — a plant leaf, a window frame, a plain wall — so nothing competes with the face.
参数卡
| 最佳模型 | Nano Banana 2.1 |
| 画幅 | 4:5 |
| 商用许可 | 可以——生成物许可取决于你使用的模型 |
为什么这条词有效
Subject: From random AI dance generations to polished K-pop performances This might be one of the best workflows for AI creators right now. Composition: a 4:5 frame at eye level with the face in the upper third, the background dissolved into soft blur, and the gaze carrying the image. Lighting: soft window light wrapping one side of the face, with open shadows and bright, round catchlights.
怎么用一条提示词
- 复制提示词
- 打开上方标注的模型
- 生成出图






