A photograph can become the starting point for dimensional artwork, but a camera records colour and brightness—not physical depth. Creating a convincing 3D relief means interpreting that image as a heightmap, controlling the surface range and editing features that an automatic conversion cannot understand. The result is a shallow solid designed to be viewed through reflected light. This workflow sits between image editing, sculpture and manufacturing preparation. A good result depends on deciding which subject projects forward, which background recedes, how much physical depth the panel supports and where light will strike the finished surface. Keeping the conversion editable allows painting, smoothing, text and shapes to correct assumptions made by a simple brightness-to-height mapping before the surface becomes manufacturing geometry.
Photograph, heightmap, bas-relief and lithophane
A photograph is a two-dimensional arrangement of colour and brightness. A heightmap is a single-channel image in which each pixel represents a relative surface height. A bas-relief is physical geometry that projects a limited distance from a backing plane. It creates form through raised surfaces, recesses and the shadows they cast.
A lithophane also varies thickness, but it is intended primarily for transmitted light. Thin regions appear brighter when backlit. A relief is normally read by light striking its surface. The same source image can produce very different geometry depending on which visual result is intended.
If your goal is a backlit image rather than raised surface artwork, read what a lithophane is and how it works.
How grayscale becomes surface height
A basic conversion maps dark-to-light pixel values across a chosen depth range. Depending on the desired result, white may become the highest point and black the lowest, or the mapping may be inverted. Intermediate values form slopes between those limits.
This creates geometry, but it does not infer the real distance from the camera to a face, tree or building. A white shirt is not necessarily closer than a dark face. A shadow on a flat wall is not a physical recess. Grayscale is therefore a useful artistic signal, not automatic three-dimensional reconstruction.
Photo-to-relief conversion should be understood as image interpretation. It cannot recover hidden surfaces or reliable real-world depth from one ordinary photograph.
Choosing a suitable source photograph
Images with a clear subject, separated tonal regions and manageable backgrounds are easier to interpret. Side-lit portraits can provide useful facial modelling, while harsh frontal flash may flatten features. Heavy compression, motion blur and digital noise can turn into unwanted surface texture.
A busy background competes with the subject for the same relief range. Hair, fur and foliage contain fine contrast that may become rough rather than detailed at the final physical size. Cropping and simplifying the image before conversion is often more effective than trying to repair every region afterward.
- Use the highest-quality original available
- Crop around the intended subject
- Avoid clipped highlights and featureless black shadows
- Remove distracting backgrounds where practical
- Judge detail at the intended printed size, not only while zoomed in
Contrast, smoothing and inversion
Contrast spreads or compresses the available height range. Increasing it can separate major forms, but excessive contrast produces plateaus and abrupt cliffs. Lower contrast keeps transitions gentle but may make the finished relief look flat.
Smoothing reduces pixel noise and small texture. It can improve broad facial or landscape surfaces, yet too much erases eyes, lettering and edge definition. Inversion swaps high and low interpretation and is useful when the source polarity does not match the intended lighting. These controls should be adjusted together while watching both the heightmap and 3D preview.
- Start with moderate contrast
- Smooth enough to remove noise, not identity
- Inspect edges for abrupt height jumps
- Use inversion intentionally
- Keep a lossless editable raster for later changes
Why manual painting, text and shapes matter
Automatic conversion treats every pixel alike. Manual editing lets a person correct that limitation: flatten a distracting background, raise an important outline, soften a harsh edge or remove isolated noise. Painting directly on the heightmap changes the underlying surface rather than merely changing its screen colour.
Persistent text and vector-like shapes should remain editable above the base raster. That makes it possible to reposition a name, resize a border or change relief height without flattening those elements permanently into the photograph.
Prompt2Part's Relief Artwork Studio processes the image locally and provides raise, lower, smoothing, flattening, text and shape tools alongside a 3D preview.
Dimensions, relief depth and backing thickness
Physical width and height determine how much detail each pixel can become. A feature visible on a monitor may be too small to manufacture or see on a 100 mm panel. Larger artwork can preserve more spatial separation, but it also requires more material, production time and structural support.
Relief depth controls the distance between the lowest and highest surface values. More depth is not automatically more realistic; excessive depth can distort faces and create steep, poorly supported walls. The backing provides a continuous foundation. It must remain thick enough for handling and production while fitting the intended display method.
The studio preview is a design aid. Final dimensions, orientation, material and manufacturability still require review before production.
Lighting and material selection
Relief artwork is revealed by reflected light. A light placed above or to one side creates shadows that describe the raised geometry. Light directly in front can flatten the appearance, while very low-angle light may exaggerate small noise. The installation location should influence how the design is edited.
A light-coloured matte material often makes surface shadows easy to read, but colour and finish are creative choices. Material selection must also consider stiffness, temperature, sunlight, impact and mounting. The appearance of a screen preview does not guarantee the same contrast after printing and finishing.
Print orientation, layers and finishing
A relief panel printed flat can provide a stable backing but builds shallow slopes through layer steps. Printing upright changes how surface detail aligns with layers but creates a taller job and may need careful support or adhesion planning. Geometry, size and available process determine the better orientation.
Layer height influences visible stepping and production time. Sanding can soften artifacts, but it can also remove fine relief detail; coatings can fill shallow texture. Any finishing plan should be tested on representative detail before committing to the final piece.
- Check that the surface and backing form one closed solid
- Avoid isolated spikes and pits
- Review minimum features at physical scale
- Plan mounting before finalizing backing
- Prototype unfamiliar finishes on a small sample
Applications and realistic expectations
Photo-derived reliefs can become portraits, commemorative pieces, decorative panels, signage, custom gifts and mixed-media inserts. They work best when the source, depth mapping, lighting and final dimensions are designed as one composition.
Faces and hair deserve special attention because viewers notice small distortions. Backgrounds may need flattening so they do not compete with the subject. A successful relief is often deliberately edited rather than being a literal one-click conversion.
Frequently asked questions
Can one photograph produce accurate 3D depth?
Not by grayscale conversion alone. Brightness is influenced by colour and lighting, so the resulting heightmap is an artistic interpretation rather than measured depth.
Should white areas be raised or recessed?
Either mapping can be used. Choose the polarity that suits the subject and intended lighting, then inspect the 3D surface rather than relying only on the source image.
How much relief depth should I use?
Use enough depth to create readable surface variation without distorting the subject or creating impractical slopes. The appropriate range depends on panel size and image content.
Is relief artwork the same as a lithophane?
No. Relief artwork is usually viewed with reflected light and visible surface shadows. A lithophane is designed around light transmitted through changing material thickness.
Design Studio
Turn your photo into relief artwork
Upload an image locally, adjust the heightmap, paint surface changes, add editable text or shapes, and inspect the relief in 3D before requesting review.
Open the Relief Artwork StudioRelated guides