Convert a Photo to a Lithophane, Step by Step

A lithophane turns a photograph into a thickness map that becomes readable with light behind it. You do not need to draw its surface by hand. You do need to choose a useful crop, set sensible physical dimensions and inspect the sliced result. This walkthrough uses the lithophane maker. The illustration is an engine render of our original procedural landscape, not evidence of a physical test print.

1. Choose a photograph you can reproduce

Start with an image you own or have permission to use. For gifts, also consider whether the people pictured would be comfortable with the intended use. The modelling tool processes the image locally in your browser; it does not upload the photo. That does not change any permissions needed to print, share or sell the resulting object. Keep the original in a separate location before editing.

A clear face, pet or landscape can work better than a busy scene with many equally small subjects. Look at the image in grayscale before deciding. Two objects with different colours may have nearly the same brightness and merge into one tone. If a face is deeply shaded, a slight exposure adjustment can help separate its features, but extreme brightening may flatten the highlights. Aim for readable tones rather than maximum contrast everywhere.

2. Crop before loading

Crop in an image editor so the important subject occupies the useful area. The generator follows the image proportions; width is not a promise of a square result. Remove unnecessary borders, captions and empty space unless you want them in the printed panel. Do not stretch a face to fit a chosen rectangle. Choose the crop first, then choose dimensions that suit the intended display.

Check the image orientation after editing. Save a normal upright copy rather than relying on a viewer’s temporary rotation. For a portrait, look for a recognizable asymmetric detail, such as a parting in the hair or a sign in the background. You can use it later to verify that you are viewing the intended face of the model. Avoid repeatedly saving a heavily compressed image, because blocky artifacts may become surface changes.

3. Open the tool and choose the image

Open the tool, choose your image and wait for the preview to rebuild. The included Mountain light sample lets you try the controls without selecting a personal file. A slow rebuild on an older phone does not mean the image has been sent to a server: geometry is calculated in a browser worker. Begin with a modest output size and sampling density if device memory is limited.

Read any error before changing several controls. If the image cannot be decoded, export a fresh PNG or JPEG from your editor and try that copy. If the page was interrupted during generation, reload and start with the built-in sample to distinguish a file problem from a device problem. Your downloaded files and the original photo remain separate from the page session.

4. Set the physical panel

Try a 100 mm width, 0.8 mm minimum thickness and 3.2 mm maximum thickness as initial model settings. These numbers need testing with your own filament and light. Keep the maximum greater than the minimum. The picture depends on a meaningful difference between light and dark regions; an almost constant thickness removes that difference even if the preview still has a recognisable outline.

Start flat so you can understand the basic surface. Curved mode adds a curve angle and changes the footprint. Rotate the preview and check the full bounds before deciding how it fits on the printer. A border gives you an area to handle and mount, but it also increases the object’s dimensions. The generator does not include a complete lightbox, fixture or power supply; plan those parts separately.

5. Export and inspect the mesh

Download STL for ordinary slicing. The current free plan allows three STL downloads per day for personal use. Separate-part 3MF export and paid plans are not available yet. A different file format would not make the original photograph higher resolution. Use a filename that records the photo version and model settings, especially if you intend to compare several thickness ranges.

Open the file in your slicer and confirm the units and overall dimensions. Rotate it into the chosen print orientation. Look for a continuous base, intact border and plausible thin areas. Step through the layer preview and inspect how the interior is filled. A shaded mesh can look smooth while the actual toolpaths reveal gaps, visible infill patterns or features too small to print. Resolve those issues before starting the printer.

6. Test a small area first

Choose a crop containing highlights, midtones and shadows, then generate a small test panel with the same thickness settings. Use the filament and backlight planned for the final object. Compare the result at a realistic viewing distance. If the image is hard to read, change one thing at a time: source exposure, thickness range, material or illumination. Keep notes so a useful result can be reproduced.

The settings guide discusses orientation, sampling and internal paths in more detail. Its slicer values are also starting points, not guarantees. Make sure any display light remains suitable for use near printed plastic. A visually pleasing model does not establish thermal or electrical safety. Allow for handling, mounting and heat in the finished display rather than assessing the panel alone.

7. Keep a reproducible project record

After a successful trial, retain the source crop, exported file, dimensions, thickness values and slicer profile together. If you later change the photo crop, regenerate and recheck the panel rather than assuming the earlier result applies. Different filament colours and batches can alter the appearance under light. For repeat gifts, a short record is more reliable than trying to remember a set of slider positions.

If what you really want is a plaque viewed from the front, try Image to STL instead. A relief uses raised geometry as the visible picture rather than transmitted light. For a caption or separate display label, use Text to STL. Each tool solves a different part of the project; combining their outputs still requires deliberate placement and checking in your modelling or slicing workflow.