Both modes start from the same thing: a greyscale copy of your image in which every pixel becomes a height. The difference is polarity — which brightness adds material — and that one choice changes the print settings, the lighting, and whether the result makes sense at all.
The difference in one line
- Heightmap (relief) — bright pixels are high. You get a carved plaque you look at under normal room light.
- Lithophane — dark pixels are thick. You get a thin panel you look through, against a lamp or a window.
Desk, keyring, fridge magnet, wall sign: heightmap. Lamp insert, window hanger, light box: lithophane.
What each one expects from the source image
A relief is read by reflected light, so contrast has to come from shadows cast by the raised surface. That favours hard-edged art — logos, line drawings, silhouettes, maps, text. A photograph works, but a face becomes a bumpy mask: noses and cheeks stick out while the silhouette is the outline of the head.
A lithophane is read by transmitted light, so mid-tones survive. That is exactly why portraits and landscapes work: the soft gradients that make a relief look muddy become the whole point when light passes through.
The practical test: if the image is mostly continuous tone, use a lithophane. If it is mostly flat colour regions, use a relief.
The slicer settings are not interchangeable
| Setting | Heightmap relief | Lithophane |
|---|---|---|
| Infill | 15–20% | 100% |
| Layer height | 0.16–0.2 mm | 0.08–0.12 mm |
| Orientation | flat on the bed | standing on the long edge |
| Walls | 2–3 | 2–3, solid through the thickness |
| Supports | none | none |
| Speed | 40–60 mm/s | 30–40 mm/s |
| Light | room light | 10–30 mm backlight behind a diffuser |
Printing a lithophane flat is the classic mistake: the layers then run parallel to your line of sight and the picture turns into a stack of bands. Printing a relief standing up wastes filament on supports that the flat base never needed.
Base and height numbers
In relief mode you control base thickness and relief height separately. A base of 0.8–1 mm with 2–4 mm of relief gives a plaque that is stiff enough to handle and still shows a visible profile under a desk lamp. Below about 1.5 mm of relief the carving looks flat; above 5 mm the vertical walls get steep enough that you start needing slower speeds and better cooling.
A lithophane is the opposite: you are working with total thickness, base plus relief, and the usable window is 2.5–4 mm. That is a much narrower range than a relief, because both too thin and too thick ruin the contrast.
What neither mode is
This is a height map, not photogrammetry. There is no depth data invented and nothing to orbit around: the mesh is a closed solid with a flat base and a height-mapped top surface. People searching for image to 3D model often expect a sculpture and get a plaque — that is the correct output for this kind of converter.
When to convert twice
Polarity is baked into the mesh. You cannot invert it later by mirroring the model in your slicer, and a relief printed standing up does not become a lithophane. If you want both a plaque and a lamp panel from one photo, run the conversion twice — it takes seconds and costs nothing.
One more case worth knowing: engraving. Some projects want a flat plate with the image cut into it rather than raised. That is the inverse of relief mode, which is why the converter has an invert switch: same image, same thickness, material removed instead of added.
Quick decision
- Logo, sign, keyring, name plate → relief (or text to STL for lettering)
- Family photo you plan to backlight → lithophane via the lithophane maker
- Terrain or depth data → heightmap to STL, relief polarity by default
- Art with transparent background → PNG to STL, then pick polarity to match the artwork
If you already have the STL and just want to check it, drop it into the STL viewer — it reads both binary and ASCII files and reports the real size in millimetres.