Practical scanning guide
How to scan black, shiny and clear objects
Separate missing surface data from lost tracking, then check the exact model, setup and permitted preparation.
On this page
Identify the failure before changing settings
Why part of the surface can disappear
Conceptual light paths and possible missing coverage. Not a scan result. Arrows and dashed regions are illustrative; they do not predict where a particular scanner will lose data.
Dark finish
The useful return can be weak. Check the actual mode, exposure and raw coverage on your part.
Open larger diagram (new tab)Shiny finish
A directed reflection can miss the camera or add false data. Compare views and inspect the unfilled surface.
Open larger diagram (new tab)Clear material
Light can pass through the material. Confirm a permitted capture method instead of assuming a closed mesh proves coverage.
Open larger diagram (new tab)Tracking: can the scanner follow the part or the required markers? Coverage: which surfaces were actually recorded before filling? Follow the exact model and mode instructions; several surface effects can occur on the same part.
Original illustration: 3D Scanner Match. CC0 license (new tab). Reuse and image credits. Documentation: SHINING 3D FreeScan preparation (new tab); Artec Ray preparation (new tab).
Dark, shiny and transparent surfaces fail in different ways, and one part can combine all three. No single setting fixes all of them.
Match the symptom to its next check and evidence
| Surface or symptom | Likely cause | Evidence to ask for |
|---|---|---|
| Dark finish, patchy coverage | Too little light returns to the cameras in this setup | An untreated scan plus the exposure, distance or mode changes that were made |
| Polished metal, noisy or false geometry | Reflections send light away from the cameras or create false points | Coverage from several angles and a view of the raw data before cleanup |
| Clear plastic or glass, missing or shifted surface | Light passes through or reflects instead of scattering from the surface | The exact capture method; a coated demo does not prove untreated capture |
| Smooth shape, tracking loss | Too few features or markers for the scanner to follow | The marker layout and a check for doubled surfaces or misalignment |
| Deep recess, hole that never closes | The scanner cannot see into the area | An access plan, or a plain statement that another method is needed |
A smooth, hole-filled mesh can hide every one of these problems. Ask to see the captured coverage before any filling or reconstruction. For parts that must fit, geometry the software filled in needs its own engineering decision.
Bring your actual part to the demo. A scanner that handles a matte sample tells you little about a polished housing or a glossy black panel. Decide in advance whether you are allowed to coat the part, because that one rule can change which scanner makes sense.
Use the instructions for your exact scanner and coating product. The sources at the end of this page show where the guidance comes from.
What to try first
- Check the scanner's distance indicator and try a different viewing angle.
- Adjust exposure or choose the model's recommended mode for that finish.
- Keep the object stable and give the tracking system the features or markers it needs.
- Try controlled lighting before judging coverage in the workshop.
- If coating is allowed, test the specified product on an approved area and record the preparation time.
Change one condition at a time and inspect raw coverage before smoothing or filling holes.
Compare capture modes on equal terms
"Laser", "infrared" and "structured light" describe parts of how a scanner works, not three separate product classes. Many handheld scanners combine a blue laser mode with an infrared mode. What matters is how the exact model and mode behave on your surface, so check the model's manual rather than assuming a light color guarantees a result.
Set up a fair comparison for each mode
Manufacturers still recommend preparation for difficult surfaces. SHINING 3D's laser-scanning instructions for its FreeScan handheld scanners recommend a matte coating spray, permanent or temporary, for reflective, shiny, transparent and dark objects, plus at least four markers placed randomly. Revopoint and Artec publish similar guidance for their own models. These are model-specific instructions, not proof that any scanner handles an untreated black part.
Ask the vendor to name the active mode, working distance and processing settings. Keep the same part and deliverable for every candidate. If one scanner needs spray and another does not, count the preparation time and the permission it requires in the comparison.
Decide whether spray is acceptable
Temporary sprays that evaporate on their own and coatings that must be washed off solve different problems. Read the technical and safety data sheet for the exact product before using it, and do not assume a disappearing spray is safe on every material.
- Get permission from the part's owner, especially for optical surfaces, museum objects, electronics and functional finishes.
- Test a hidden area first if the product instructions and the owner allow it, and note how the coating is removed or disappears.
- Follow the manufacturer's ventilation, handling and protective equipment instructions. This guide does not replace the safety data sheet.
- Record the product, how it was applied and how long ago. Repeated coats change the surface the scanner sees.
- For tight tolerances, account for the coating in your measurement plan. Coating thickness depends on the product and how it is applied.
AESUB, one spray manufacturer, publishes technical and safety sheets for each product. Use the sheet for the can you actually have. If preparation is not allowed on production parts, the demo has to pass without it, or the scanner is not suitable for that job. Its AESUB blue product instructions describe a self-vanishing coating and application procedure. Confirm compatibility for the actual material and keep coating thickness separate from the scanner claim. The product description is not a measurement on your part.
What a coating can do to a dimension
A coating can widen a part and narrow a hole
Illustration: a coating 0.01 mm thick on each relevant surface. This is an assumed thickness, not a measured spray result. The coating is drawn larger for clarity.
Outside faces: dimension grows
- Before coating
- 40.00 mm
- Idealized after coating
- 40.02 mm
40.00 + 2 x 0.01 = 40.02 mm
Open larger schematic (new tab)Inside bore: opening shrinks
- Before coating
- 5.00 mm
- Idealized after coating
- 4.98 mm
5.00 - 2 x 0.01 = 4.98 mm
Open larger schematic (new tab)Original schematic: 3D Scanner Match. CC0 license (new tab). Use this as geometry arithmetic, never as a correction for an actual scan.
A scanner observes the coated surface. Preparation that improves coverage can also change the geometry being measured. Record which reference was measured before or after coating and whether the same prepared condition is used for the independent check.
Read the arithmetic table and prepared-condition checks
Consider an idealized, uniform 0.01 mm coating on each relevant surface. This example thickness is for the calculation, not a measured or claimed AESUB coating thickness:
| Geometry before coating | Idealized coated geometry | Why the direction matters |
|---|---|---|
| 40.00 mm between two external parallel faces | 40.02 mm | Coating on both outside faces adds 0.02 mm to the span |
| 5.00 mm cylindrical bore with both walls coated | 4.98 mm diameter | Coating inside the bore reduces the open diameter by 0.02 mm |
Real application can be uneven and can change with elapsed time. These calculations are not a correction to apply to a scan. Do not subtract a guessed thickness from a dimensional result and call it accepted. Define the prepared measurement condition and an adequate independent reference with the supplier.
For the permitted A/B comparison, retain the required markers and same capture settings. For a mating part, connect the preparation record to the named functional dimensions and the accuracy decision rule.
Separate markers from surface preparation
Markers help the scanner follow its position; coating changes the surface it sees. Retain the markers required by your mode when comparing untreated and coated captures. Keep them clear of the edge or bore you need to measure, and follow the manual's size, placement and visibility rules.
"Untreated" in the comparison below means no coating. A marker-free test is useful only if the selected mode supports it.
Match marker rules to the connection and release
The CrealityScan 4.3.2 desktop notes describe marker-free Raptor-series scanning over USB. The mobile APP guide gives blue-laser marker instructions. Follow the instructions for your exact scanner, connection and installed version.
Record a surface test
Photograph the part and track the same areas in each run. A and B compare surface preparation under a controlled setup. C checks whether the selected process works in the production environment; it does not isolate the effect of preparation.
Download the blank surface demo record (CSV) and instructions (TXT). Use one copy per scanner candidate. The record has fields for settings, raw coverage, filled geometry, reference measurements and total job time. It contains no test results or preset pass criteria; agree those with the supplier before the demo.
Swipe or use Left and Right arrow keys to see every column.
| Run and purpose | Keep fixed | Change and record | Pass/fail evidence |
|---|---|---|---|
| A: untreated baseline | Part, required features, output, reference and pass/fail rule; choose and record mode, exposure preset, distance, lighting, required markers, capture path and processing | No coating. Retain required markers. Save the setup as the baseline for B | Raw coverage, tracking breaks and the exported file, including a failed capture |
| B: controlled preparation comparison | The A setup, operator, alignment rule and processing, including required markers | Only the owner-approved surface preparation. Record product, application, elapsed time and removal | Coverage difference, lost detail and reference checks. If settings also changed, you cannot tell whether the coating or the settings made the difference |
| C: production-environment validation | Same part, required output and pass/fail rule; use the selected preparation and processing | Actual lighting, work area, access and operator handling. Record differences from the controlled setup and repeat captures | Usable output across repeat captures and total job time; a separate validation result, not an A/B preparation claim |
Where exposure or another setting adapts automatically, keep the same preset and record the observed behavior. If the operator changes mode, marker layout, distance or processing between A and B, repeat a matched pair before drawing a preparation-only conclusion. An untreated baseline that cannot be captured is a recorded limitation, not a reason to invent data. If coating is not allowed, omit B and validate the permitted untreated process in C.
When the demo does not pass
Do not accept the result if a required face is still missing, if filled-in holes are presented as captured data, or if the preparation that made it work is not allowed on production parts. Call it inconclusive if the reference measurement or setup was not good enough.
Take both the original and processed files with you. Ask the vendor to mark every area rebuilt by hand and to finish your downstream task. A printable shell and a CAD reference for a fitted part need different pass/fail checks.
For dimensions, use the accuracy check sheet. For the full scanner and software quote, see 3D scanners for reverse engineering. The car-parts guide adds examples with painted panels, housings and deep recesses.
Diagnose tracking loss before repairing the mesh
Start with one documented failure and change one condition at a time. Keep the failed segment before repair.
Open the tracking and processing diagnosis checklist
| Observed problem | Next controlled check | Evidence to retain |
|---|---|---|
| Capture stops or jumps | Selected mode, working-distance indicator, exposure, speed and visible reference features/targets | Failed segment, settings and what changed |
| Doubled edge or seam | Stable object/reference arrangement, overlap and pass alignment | Separate passes and registration transform |
| Smooth area disappears | Actual reflectance, permitted preparation and optical access | Unedited coverage before filling |
| Color looks complete but geometry is missing | Geometry view without the texture | Raw depth/points and the named feature |
| Detail disappears after processing | Compare raw/full mesh with reduced or smoothed copy | Operations, parameters and both files |
Change one condition at a time. Check tracking recovery as well as an uninterrupted run. Record whether a failure was a reference, surface, access or processing problem before trying another device.
Revopoint's tracking guidance , checked October 5, 2026, discusses working distance, steady movement, exposure and recovery. Its global-marker guidance warns against moving the established object/reference arrangement. Use the instructions for the exact scanner and mode; no target-count rule applies to every scanner.
Check the line of sight
Coating and markers cannot expose a hidden pocket. Check whether the scanner can see the required wall from enough angles. If a fixture carries targets, keep it rigid relative to the part and test any repositioning step.
A hidden surface needs another view
Even a matte surface can be hidden behind a rim. Follow one viewing line in this illustrative cross-section.
This illustrates access only. It does not predict coverage or accuracy for a scanner. Check the actual part, mode and raw geometry before filling a missing area.
Original illustration: 3D Scanner Match. CC0 license (new tab).
For transparent or deformable parts, compare an owner-approved preparation or another measurement route. Filled-in geometry cannot replace a missing measurement.
Calibration and a reference check are separate
Follow the model's calibration procedure and handling conditions. Record scanner identity, software release, calibration accessory and result, temperature/environment and any transport or impact that calls for another check. Do not substitute instructions from another product family.
Revopoint's Trackit manual , last edited August 20 and checked October 17, 2026, calls for calibration before first use, after a change in scanning-environment temperature, or after transport. These are Trackit instructions, not a schedule for every new scanner. Follow your exact model's manual.
A successful calibration routine is not independent verification of your delivered dimension. After setup, capture a suitable reference or independently measured feature, preserve the result and apply the agreed accuracy decision rule. If calibration fails, retain the error and logs for the vendor; do not relabel a failed routine as complete.
Frequently asked questions
Can you 3D scan black objects?
Some can, but results depend on the exact scanner, mode, finish and lighting. Start with your own uncoated part while retaining any markers the selected mode requires. If preparation is allowed, compare a matched prepared capture; do not infer untreated performance from a coated sample.
Can you 3D scan shiny or transparent objects?
Shiny metal can often be captured from several angles, but reflections create noise and gaps. Clear parts can need a permitted coating or another capture method. When coating is prohibited, require an actual untreated result on the required surface and feature before buying.
Sources
- SHINING 3D: preparation for laser scanning (FreeScan) , checked October 1, 2026.
- Revopoint POP 3 Plus: hard-to-scan objects
- Artec Ray: scanning preparation
- AESUB technical and safety downloads , with the blue product instructions checked October 14, 2026.
- How we compare
The demo record above is a planning tool, not a product certification.