Your images, ready for the wall
Shmexy Relief Canvas turns imported photographs, illustrations and artwork into framed canvases in Blender. Create a lightweight Flat Plane for interior scenes, or use 3D Relief to give the visible surface a shallow, editable depth.
Import. Shape. Frame.
- Bring your own image: use portraits, landscapes, illustrations and artwork you have permission to use.
- Two canvas modes: Flat Plane needs no depth runtime; 3D Relief estimates depth locally from your imported image.
- Live relief adjustments: generate depth once, then refine the result without running depth analysis for every adjustment.
- 28 frame presets: choose a molding profile and finish, with automatic fitting to the canvas proportions.
- Editable Blender objects: keep working with the canvas and frame in your scene, and save or export your result.
A focused canvas workflow
This public edition does not include text-to-image or image-to-image generation. You supply the source image. There is no ComfyUI requirement, image-generation model, or 24-part image-engine download. Automatic depth estimation remains an optional AI-assisted feature.
Requirements and setup
Validated in Blender 4.5.13 and 5.2.1 on Windows. Automatic 3D Relief depth requires the separate Windows x64 local depth package (approximately 155 MiB; allow 1 GB free space). Install it once from the add-on, then process images locally. Flat Plane works without that package. A single image produces a shallow visible-surface relief, not a complete reconstruction of hidden geometry.
What is included
Installable add-on, editable Python source, setup guide, license notices and 28 procedural frame presets. The optional depth runtime is supplied separately for Windows x64.
About the gallery
The room renders show canvases created from example artwork, including images generated externally. They illustrate framing and scene placement; this edition does not generate those source images. Room furnishings and decorative props are presentation context and are not included.
Development used substantial AI assistance, followed by creator review and testing. The depth model is Depth Anything V2 Small, with Apache 2.0 license and provenance notices included.