A 2D laser profile scanner triangulates a line of points across a moving target to build a dense 2D cross-section, while a confocal chromatic displacement sensor stacks a single point through a chromatic lens and decodes height from the wavelength that returns to peak intensity [S2][S3]. The two architectures answer different metrology questions: profile scanners measure contour and roughness trend over a line at high speed, confocal sensors resolve single-point height with sub-micrometre repeatability on surfaces that confuse triangulation, such as polished metal, glass, or transparent coatings.
Specifying either technology for a surface finish application comes down to three variables: the required lateral scan length, the target Ra/Rz range, and the optical behaviour of the material (specular vs diffuse vs transparent). Engineers in metal stamping, bearing raceways, and additive-manufactured part inspection routinely compare these two sensor classes side by side, and the decision rarely hinges on a single spec sheet; it is a fit-for-purpose judgment that combines standoff, spot size, and controller integration.
How each sensor actually measures
A laser profile scanner projects a line onto the target, observes the line at an angle through a 2D CMOS array, and converts each pixel of the line into a height value via triangulation; the published 2D framing rates run from roughly 1 kHz to 200 kHz depending on profile count and field of view [S2]. Confocal chromatic sensors, by contrast, focus white light or LED through a lens with controlled longitudinal chromatic aberration so each wavelength focuses at a different distance, and the sensor identifies the peak returning wavelength to derive a single-point height; the technique is non-contact, sub-micrometre capable, and largely unaffected by surface colour, mirroring, or transparency.
In practical terms, the profile scanner is a "line" device: you get a 2D cross-section (X by Z) per scan pass. The confocal sensor is a "point" device: you get a Z value per measurement, and you build surface coverage by scanning the part under the sensor with a motion stage. If the inspection question is "what is the Ra of a 50 mm bearing track?", a profile scanner can capture that in one or two passes; if the question is "what is the coating thickness on a transparent polymer layer?", the confocal sensor can resolve it because the chromatic focus penetrates transparent films and identifies each layer's interface.
Resolution, standoff, spot size: numbers that matter
Triangulation-based profile scanners typically deliver vertical resolutions in the 1-10 micrometre range and lateral resolutions of 10-50 micrometre, with working distances of 30-200 mm, and the OGP Cobra 2D type scanners cited in industrial metrology literature target shop-floor 2D/3D curve capture on machined parts at high throughput [S3]. Confocal chromatic displacement sensors reach sub-micrometre vertical resolution (often 10-300 nm, depending on the controller and range), with spot diameters on the order of 2-20 micrometre, and standoff ranges typically 1-30 mm; the physical measurement principle, decoupled from return-signal intensity, lets them hold calibration on glossy, polished, and transparent targets where triangulation struggles.
Spot size is the most under-appreciated variable. A 20 micrometre spot averages surface features smaller than itself and biases Ra low; a 2-5 micrometre confocal spot resolves micro-roughness but slows areal coverage. For a roughness filter cut-off of 0.8 mm (the standard lambda-c value for many ISO 4287/4288 evaluations), a profile scanner with a 2,000-point line at 50 micrometre pitch easily covers the cut-off length in a single pass, while a confocal point sensor at 1 kHz needs a fast motion stage or risk under-sampling.
Specular, transparent, and layered surfaces
On highly reflective metal surfaces, triangulation can fail because the specular return misses the receiver aperture; on transparent or multilayer substrates, triangulation typically measures the top of the bulk material and cannot separate film thickness. Confocal sensors handle both because they decode height from wavelength rather than return-signal geometry, and the chromatic focus "finds" the strongest interface whether it is a polished surface, a coating, or a buried layer. [S2]
That makes confocal the default for applications such as thickness of paint and clearcoat on automotive body panels, glass and display cover thickness, and stacked-film inspection in electronics; profile scanning is the default for weld bead profile, brake-disc run-out, and edge geometry on stamped or rolled parts where the target is rough or textured enough to give a strong diffuse return.
Comparison matrix: profile scanner vs confocal displacement sensor
Engineers can line up the two technologies on four decision criteria: measurement type, vertical resolution, best-fit surface type, and integration effort. Profile scanners measure a 2D line (X-Z cross-section) at 1-10 micrometre vertical resolution, fit diffuse to moderately reflective surfaces, and integrate with a simple encoder + line-rate trigger. Confocal sensors measure a single Z point at sub-micrometre resolution, fit specular, polished, transparent, and multilayer surfaces, and require XY motion or a scanning bridge plus a controller with chromatic decoding electronics. Cost, throughput, and calibration overhead scale accordingly: profile scanners typically need only a flat-field calibration, while confocal sensors require a wavelength-calibrated reference target and careful thermal management because the chromatic focus can drift with temperature. [S2]
For shop-floor full-field inspection of formed-metal parts at high throughput, the laser displacement sensor family (which includes profile scanners) remains the high-volume choice; for precision point metrology on surfaces that break triangulation, the confocal displacement sensor class is the high-accuracy choice. A related concept worth understanding is the broader displacement sensor family, of which both are sub-types; understanding the parent category clarifies why a profile scanner and a confocal sensor can both be called "displacement" devices even though their physics differ.
Standards, calibration, and traceability
Surface-finish reporting in regulated industries is typically tied to ISO 21920 (the current roughness-evaluation standard, superseding ISO 4287 in many new specifications) and ISO 25178 for areal parameters; stylus-type traceable references remain the calibration baseline against which non-contact sensors are correlated. Users should expect any non-contact sensor to be referenced against a calibrated roughness artefact (Ra reference, typically 0.1-3.2 micrometre class) and a flat/step height standard for Z linearity, with documented correlation against a displacement position reference before any shop-floor release.
Calibration artefacts for both sensor classes are widely available, but the periodicity differs: a profile scanner with a closed optical head on a vibration-isolated bench typically needs recalibration every 6-12 months, while a confocal chromatic sensor with a fixed optical head and LED source can hold calibration longer if the controller temperature is controlled, though drift on the order of a few hundred nanometres per 10 K is common and should be budgeted in the metrology loop. For process-control release decisions, ISO 9001 / IATF 16949 quality systems will look for documented correlation between the in-line sensor and a periodic tactile reference; that correlation file, not the sensor spec sheet, is what matters at audit time.
When each technology is the wrong choice
Profile scanners are the wrong choice when the surface is mirror-finish and the specular return cannot be guaranteed, when the target is transparent or multilayer, or when the required lateral resolution is finer than roughly 5 micrometre. Confocal sensors are the wrong choice when the application needs full 2D coverage at line rates above a few kHz, when the part geometry forces long lateral travel incompatible with the available motion stage, or when the budget cannot absorb the controller and chromatic decoding cost per measurement point. [S2]
For large-area coverage of stamped or welded sheet metal at automotive line rates, the laser distance sensor class (time-of-flight or triangulation) is often a better fit than either; for high-speed full-field 3D surface capture, structured-light or stereo-camera systems become the appropriate answer. Knowing the failure modes of each non-contact class is what keeps engineers from over-promising resolution or coverage to a quality team.
Selection checklist and field signal
A practical starting heuristic: if the part is rough or textured and the question is "what is the 2D profile, weld contour, or Ra trend over a 10-100 mm line", use a 2D laser profile scanner. If the part is polished, transparent, or coated, and the question is "what is the height or thickness at a specific point with sub-micrometre repeatability", use a confocal chromatic displacement sensor. Trackable next signals to watch: published correlation studies between confocal sensors and ISO 21920 stylus references on Ra values below 0.2 micrometre, and factory-side updates to laser profile scanner controller firmware that add areal parameter (Sa, Sz) reporting to match ISO 25178 areal conventions. For comparison criteria on adjacent non-contact metrology, see the Surface Roughness Tester vs Vision Measuring Machine for Calibration Artifact reference. [S2]