Methodology
v1 · 19 September 2026
The same test for every scanner
Every scanner in this database scans the same set of objects and is described with the same criteria. A new scanner adds a column to the same table instead of being a separate, incomparable review. When the set of objects or criteria changes, the methodology version changes.
What you get for every scanner
- Photos of the device.
- Scans of every test object, published as renders and as a small 3D mesh you can rotate in the browser, shaded or as wireframe. The mesh is decimated to keep files small; the renders show the full-resolution result.
- Scores from 1 to 5 for each criterion, plus measured values where a script can compute them.
- Date and software version of the test. Published results are never edited; a re-test after a software update is a new record.
Scan mode: infrared or laser
Every scan says which mode it was made in: infrared structured light (NIR), visible structured light or laser lines. This is not a technical footnote: one hybrid scanner does the same object both ways and the results can be very different. The laser usually wins on black, shiny and fine detail, while infrared is faster and more often works without markers.
So each mode is a separate test with its own scores. A hybrid scanner gets one column per mode in the comparison table and one block per mode on a test object page. Scores from two modes are never merged into one number.
Score scale
5 excellent, 4 good, 3 acceptable, 2 weak, 1 unusable. A score describes only what happened with that object on that scanner; it is not a ranking against other scanners.
Measured values
Where a number can be computed, it is computed by the scripts in the repository on the mesh exported at full resolution. Nothing calculated by vendor software is used. The sphere bar gives the sphere spacing error (accuracy) and the sphere form error (surface noise).
Settings
Manufacturer default settings, no smoothing and no hole filling before export. Each object is scanned up to three times; the best attempt is published.
Disclosure
Some links are affiliate links. They are marked and have no influence on scores.
Test objects
Scan modes
A hybrid scanner scans the same object in both modes, so each mode is a separate test with its own scores. Every scan below says which mode it was made in.
- infrared
- Infrared structured light (NIR)An invisible infrared pattern. Usually faster and markerless, weaker on fine detail.
- blue light
- Visible structured light (blue)A visible blue pattern. Good on detail, harder on shiny and very dark surfaces.
- laser
- Laser linesThe strongest mode on black, shiny and fine detail; usually needs markers and more time.
Criteria
Every scanner is described with the same criteria. A score is 1 to 5 by the scale below; a measured value comes from a script run on the exported mesh, never from vendor software.
Score scale: 5 excellent, 4 good, 3 acceptable, 2 weak, 1 unusable.
| Criterion | Kind | Evidence |
|---|---|---|
| precise | score 1 to 5 | sphere |