Output quality

    Outpainting

    Also called Uncrop, Generative expand.

    Outpainting extends an image beyond its original borders, generating new content that continues the scene outward.

    The picture stays as shot and the canvas grows around it. The model reads the existing edges — perspective, horizon, light direction, texture — and continues them into empty space, which is how a tightly framed photo becomes a wide one that never existed.

    Its main job is reformatting without loss. Taking a square asset to a full-screen vertical by cropping means throwing away the sides; taking it there by outpainting means keeping every original pixel and inventing headroom and floor. For a campaign that has to run in three frame shapes from one hero image, that difference is the whole workflow.

    The further you go, the less the original constrains it. A modest extension inherits perspective convincingly; a large one drifts, because there is progressively less real information anchoring the invention.

    In practice

    • Extend in stages rather than in one jump — each pass has more context to work from.
    • Extending along one axis at a time is more reliable than expanding on all four sides.
    • Architecture and straight lines are the hard case; skies, foliage and water are the easy one.

    The mistake to avoid

    Expecting a subject to be completed correctly. Outpainting continues a scene; it does not know what the rest of a cropped person or product actually looks like.

    Where you will run into it

    Related terms

    The all-in-one AI studio for creators. 60+ models for video, image, voice, music and lipsync in a single app.