OpenZhenZhen AI 프롬프트 갤러리 · gpt-image-2.5 · image

구도를 고려한 21:9 캔버스 확장

Composition-Aware 21:9 Canvas Extension

‘구도를 고려한 21:9 캔버스 확장’은(는) gpt-image-2.5로 만든 작품입니다. OpenZhenZhen에서 원본 프롬프트와 생성 예시를 살펴보고 새로운 창작에 활용해 보세요.

구도를 고려한 21:9 캔버스 확장 — AI image creation
구도를 고려한 21:9 캔버스 확장

프롬프트

중국어 프롬프트

将完整附件图像 ref-10a-v1 同时作为待转换照片和唯一身份参考。将画布扩展为 21:9 横向比例。保持完整原图居中且内容不变：保留成年女性可辨识的面部及五官、头发、姿势、黑色细肩带上衣、纯青绿色背景、所有精细细节，以及原有光线和构图。不要裁切、非等比缩放、扭曲、重绘或修饰原图内容。只将现有纯青绿色背景延伸到新增的两侧区域，并匹配原有颜色、色调、纹理和光线，使接合处没有可见边界、光晕或颜色偏移。不要臆造未显示的身体部位、物件、场景、文字或标志。如果原图中存在排版文字或品牌标记，应保持其外观准确，并保留相对于原图的位置。扩展后的画布应围绕主体保留自然留白和视觉平衡。输出一张协调统一的 21:9 图像，使其看起来原本就是按该比例构图，光线方向一致，背景衔接自然无缝。

영어 프롬프트

Use the complete attached image ref-10a-v1 as the photo to transform and the sole identity reference. Expand it to a 21:9 landscape canvas. Keep the entire original image centered and intact: preserve the adult woman's recognizable face and features, hair, pose, black thin-strap top, plain teal background, all fine details, and the original lighting and composition. Do not crop, scale non-uniformly, distort, regenerate, or retouch the original content. Extend only the existing plain teal background into the new side areas, matching its color, tone, texture and lighting so the joins have no visible seams, halos or color shifts. Do not invent unseen body parts, objects, scenery, text or logos. If any typography or brand mark is present in the source, retain its exact appearance and position relative to the original image. Preserve natural negative space and visual balance around the subject on the wider canvas. Output one coherent 21:9 image that looks naturally composed at that ratio, with consistent light direction and seamless background continuation.

API 모델 보기

더 많은 이미지·동영상 프롬프트 보기 · 이미지 생성 · 동영상 생성 · AI 캔버스에서 만들기