A cylinder is the friendliest shape for optical measurement and the least informative. A silhouette gives you two edges and therefore a diameter at one axial position. Everything a machinist cares about beyond that — taper, roundness along the length, runout — is either another station or another instrument.
Outer diameter at the imaged axial position, and length if both ends are in the field of view. That is the honest scope of one silhouette. It is a large scope — diameter and length cover a great deal of incoming inspection — but it is not roundness, not taper, and not runout.
The reason telecentricity earns its place here is diameter variation: shafts arrive across a tolerance band, and a shaft at the top of the band presents its surface closer to the lens than one at the bottom. Object-side telecentricity removes the resulting magnification difference; a standard lens turns diameter variation into a reading error.
Unlike a flat part, a shaft on a fixture has physical bulk around it: centres, a V-block, a chuck, a conveyor edge. The lens has to clear all of it. This is why the models below are ordered by working distance rather than by resolution.
There is a hard trade: at a given magnification, longer working distance requires a larger front element and a larger, more expensive lens. Specifying WD generously "just in case" is a real cost.
| Model | Magnification | Working distance | Resolution | Depth of field | Distortion | Telecentricity | Sensor | Mount |
|---|---|---|---|---|---|---|---|---|
| PMS-LTFM0106-468 | 0.106X | 468mm | 34.95μm | — | 0.02% | — | 4/3″ | F-mount |
| PMS-LTFM0206-468 | 0.206X | 468mm | 21.16μm | — | <0.01% | <0.02° | 35 mm | F-mount |
| PMS-LTCM0063-416-G | 0.063X | 416mm | 63.8μm | — | 0.01% | <0.02° | 4″ | C-mount |
| PMS-LTFM017-416 | 0.17X | 416mm | 31.56μm | 22.14mm | 0.01% | <0.02° (typ. 0.05°) | 35 mm | F-mount |
| PMS-LTCM0061-398 | 0.061X | 398mm | 87.99μm | 171.97mm | 0.01% | <0.02° | 2/3″ | C-mount |
| PMS-LTCM0088-398-G | 0.088X | 398mm | — | — | 0.01% | — | 1″ | C-mount |
| PMS-LTFM024-398 | 0.24X | 398mm | 30.76μm | — | 0.01% | — | 35mm | F-mount |
| PMS-LTCM0118-261-G | 0.118X | 261mm | 31.85μm | 32.18mm | 0.01% | — | 1″ | C-mount |
| PMS-LTCM016-189 | 0.16X | 189mm | 25.17μm | 18.75mm | 0.01% | <0.02° | 1″ | C-mount |
| PMS-LMC03-185 | 0.3X | 185mm | — | — | 0.02% | — | 2/3″ | C-mount |
Magnification from 0.061X to 0.3X, ordered by working distance, longest first. Shaft fixtures — centres, V-blocks, chucks and conveyor edges — need clearance, and at this magnification the field of view covers the diameters that incoming inspection usually gates.
A shaft on centres that is not perfectly concentric rotates its silhouette as it turns, so a single-shot diameter is a diameter at one angular orientation. If the part is rotated and multiple shots taken, the spread is real roundness information — but only if the axis of rotation is the datum the drawing uses.
For long slender shafts, sag between centres is larger than the diameter tolerance. Measure near a support, or support the part along its length.
A silhouette of a cylinder gives two edges. Two edges cannot distinguish a circle from a three-lobed form, and they cannot detect lobing at all. What is sometimes sold as "roundness measurement" from a single view is a diameter measurement at one orientation.
Genuine roundness needs either rotation with many samples or a dedicated roundness instrument. See when a telecentric lens is the wrong choice.
To measure diameter at more than one axial position you need either a larger field of view (lower magnification, worse resolution) or a moving stage. A common configuration is a telecentric lens on a linear stage with encoder feedback, sampling diameter at defined positions. The stage accuracy then becomes part of the measurement.
Collimated backlight. A cylinder lit by a diffuse backlight produces a bright rim at each edge that shifts the extracted edge inward, and the shift depends on surface finish — so a ground shaft and a turned shaft read differently at the same true diameter. The lighting guide covers this in detail.
Not from a single view. A silhouette gives two edges, which cannot distinguish a circle from a lobed form. Real roundness needs rotation with many samples or a dedicated roundness instrument.
Because a diffuse backlight produces a bright rim at each edge and the extracted edge is pulled inward by an amount that depends on surface reflectivity. Use a collimated backlight and keep the lighting constant between part families.
Enough to clear the fixture — centres, V-block, chuck or conveyor edge — not the part alone. Specifying WD generously costs money, because at a given magnification a longer WD requires a larger front element.
One silhouette gives the diameter at one axial position. Taper needs either a larger field of view at lower magnification, or the lens or part on a linear stage with encoder feedback, in which case the stage accuracy becomes part of the measurement.
Yes, because shafts arrive across a diameter tolerance band and a shaft at the top of the band presents its surface closer to the lens. A standard lens converts that into a magnification error; object-side telecentricity removes it.
Sag between centres typically exceeds the diameter tolerance. Measure near a support, or support the part along its length, and treat the support positions as part of the measurement setup.
Specification arithmetic gets you to a shortlist, not to a verdict. Send one part with the drawing and the tolerance you have to hold, and we will run it and report what the system returns — including the features that come back as unmeasurable. A written feasibility note is worth more than another datasheet.
Simply enter your email to receive the latest news and insights from Pomeas. Stay connected with Pomeas and be the first to discover new innovations in optical excellence.