When a client sends us an object to digitise, the first decision we make is which capture method to use: structured-light 3D scanning, or photogrammetry. Both produce a 3D mesh you can print or reverse engineer — but the tolerances, cost, and turnaround are worlds apart.
Structured-light 3D scanning
Structured-light scanners project a calibrated pattern onto the object and read the deformation with two cameras. On the metrology-grade scanners we run in the Guildford studio, accuracy sits at ±0.05 mm across a 400 mm capture volume, and we typically finish a small part in under an hour.
Best for: - Reverse engineering machined or moulded parts where dimensional accuracy matters - Inspection against a CAD reference model - Small-to-medium objects (up to ~1 m) with matte or lightly sprayed surfaces - Anything you plan to reprint at 1:1 scale
Photogrammetry
Photogrammetry stitches dozens or hundreds of photographs into a mesh using feature matching. The maths is impressive, but the accuracy is bounded by lens distortion, lighting, and the number of overlapping views. You'll typically see 0.5–2 mm accuracy on well-lit objects — an order of magnitude worse than a structured-light scan.
Best for: - Very large subjects (vehicles, buildings, monuments) where a scanner physically can't reach - Colour texture capture for visualisation, film, or games - Sites where you can't bring hardware — a phone camera is enough
How we choose in practice
If the deliverable is a printed part, a CAD model, or an inspection report, we scan. If the deliverable is a coloured mesh for a render or a game engine, and the object is either huge or unreachable, we shoot photogrammetry. On some hybrid jobs we do both: photogrammetry for the shell, a scan pass for the critical mating faces.
If you're unsure, send us a photo and a short description — we'll write back within 24 hours with the right approach and a fixed quote.
