Pic23D Open converter

2D to 3D model: what converts well

Two jobs hide behind this search: giving a picture depth, and inventing a full 3D object. This page explains which one you actually want, and what each can and cannot deliver.

Two very different jobs share one search term

When people search for a 2D to 3D model conversion they usually want one of two things, and confusing them is the main reason people end up disappointed with a converter.

Job one: keep the picture exactly and give it depth. The result is a relief, a lithophane or a plaque. Every pixel of the source is represented, and the part is predictable — you know what you will get before you print it. This is what photo to STL and PNG to 3D model do.

Job two: invent a full 3D object from a flat picture. The result is a figure with a back, sides and guessed geometry. That requires a reconstruction model trained on 3D data, and the output is only as good as its guess — thin parts break, texture is lost, and a single front view leaves the back ambiguous.

Neither is better. They answer different questions, and a good workflow picks one on purpose.

A quick decision path

  1. Do you need the picture to be recognisable? If yes, you need a relief. Reconstruction will not preserve it faithfully.
  2. Do you need to see the object from behind? If yes, you need a reconstruction — no amount of image processing can recover geometry the photo does not show.
  3. Is it a logo, sign, badge or cookie cutter? Extruded relief. Start with the plaque mode.
  4. Is it a photo of a person or a pet you want as a gift? Lithophane. It is cheap, it prints in an evening, and it looks like a photo when lit.
  5. Is it a product part that must fit something? Neither. Model it in CAD; reliefs cannot carry tolerances.

What relief conversion does to an image

Brightness becomes height. That single rule explains almost every result you will see: dark areas are thick, bright areas are thin, and midtones sit in between. It also means flat lighting converts badly — if the photo has no contrast, the relief has no depth. Raise contrast before converting rather than after.

Transparency is a second channel. In plaque mode an opaque pixel is raised and a transparent pixel is base, which is why a clean PNG cut-out gives such a crisp outline.

File formats and what to do next

STL is the language of slicers. OBJ is the language of 3D software. Both come out of the same mesh at real millimetre scale, so pick by what you will open next, not by quality. Our image to OBJ page covers the difference in detail.

Print advice by type

  • Lithophane: upright, 100% infill, 0.12 mm layers, white filament, 3 mm depth range.
  • Plaque: flat on the base, 0.2 mm layers, 3 perimeters, 2 mm relief, brim for thin shapes.
  • Relief tile: flat, 0.16 mm layers, 15% infill; keep the maximum height under 6 mm so it does not need supports.

Where relief conversion stops working

Very small features, hair-thin strokes, deep shadows and glass or smoke textures all fall apart at printer resolution. If the feature would be thinner than 0.8 mm at the finished size, it will not print no matter how high you set the detail slider.

Next step

Pick the mode that matches your job: photo to STL, PNG to 3D model, or image to OBJ. Every one of them runs locally in your browser with no signup.

Questions

What is the difference between a relief and a reconstruction?

A relief keeps the picture and adds depth — brightness becomes thickness. A reconstruction invents a full 3D object and guesses what the photo cannot show. Reliefs are predictable; reconstructions are speculative.

How do I turn a 2D logo into a 3D model?

For a logo or a sign, plaque or relief mode is the right answer and takes about a minute. For a full 3D object from a single photo you need a reconstruction model — this site does not pretend to do that.

Which file format should I export?

STL for slicers, OBJ for Blender, CAD, CNC and game engines. The geometry is identical; the container is the only difference.

What is the minimum size that still reads?

Reliefs work at any size. Lithophanes need to be at least about 60 mm wide and 2 mm deep before the image reads, and 100 mm is where they look best.

Can I convert a photo with fine detail like hair?

If the feature is thinner than 0.8 mm at the finished size, no. That is below two nozzle widths. Scale up the part or simplify the artwork.