Composition
Composition is how everything fits within the frame. It’s the balance of distance, angle, and perspective that defines how the viewer reads the scene. It tells the model where to place the subject, how much space surrounds it, and what sits in front or behind.
Small shifts in composition can change the entire message of an image: a person centered against a blank wall feels formal and static, while a person in the lower corner with open sky above feels dynamic and free.
Spatial cues help the model build this structure. Use phrases like centered composition, symmetrical layout, rule of thirds, tight crop, or wide shot to control balance and distance.
Perspective terms such as profile view, three-quarter view, top-down view, or subject facing forward set the camera angle and orientation. If you want a sense of depth, describe layers: foreground flowers, subject in middle ground, mountains fading into distance.
For multi-object scenes, clarify relationships: a child holding a balloon beside a street vendor, a cat sitting under a wooden chair, or a car parked in front of a café. These relationships keep geometry stable and prevent the model from blending elements incorrectly.
Composition language functions like stage direction. It tells the model not only what to show, but how everything fits together inside the frame.