The optimal f-number is the aperture range that supplies enough focus tolerance and exposure while keeping diffraction below the inspection’s usable resolution threshold. That aperture and diffraction trade-off must be tied to a real defect, edge, code, or measurement—not merely to whether an image looks sharp on a monitor.

The Aperture Trade-Off in Machine Vision

Depth of Field is the object-distance interval within which defocus blur remains below a selected circle-of-confusion limit. A higher f-number means a smaller aperture: focus tolerance grows, but less light reaches the sensor and diffraction increases. A lower f-number admits more light and supports shorter exposures, but reduces focus tolerance. These relationships are documented in the machine-vision aperture overview from 1stVision.

Machine Vision Aperture Trade-Off Table: What Changes When You Open or Stop Down the Lens

A full stop reduction in aperture halves the admitted light, according to the cited machine-vision source. Whether that matters depends on available illumination, sensor response, line speed, and the exposure ceiling imposed by motion.

Define Inspection Requirements Before Choosing an F-Number

Calculate the required depth range from object-height variation

Record the nearest and farthest acceptable object surfaces relative to the lens, including fixture tolerance, part warp, conveyor runout, and changeover error. Focus calculations describe geometric limits; they don’t prove that an algorithm can detect a low-contrast defect at those limits.

Set the nominal focus plane deliberately. Centering it between the near and far surfaces is a useful starting inference, not a universal rule, because near and far focus tolerance need not be operationally equivalent.

Set the smallest feature and acceptable image-space blur

Translate the smallest required feature into sensor pixels at the selected magnification. Then decide how much contrast loss the detector, code reader, or metrology routine can tolerate. One-, two-, or four-pixel blur allowances can be trial criteria, but none is a universal pass threshold, as the Commonlands machine-vision focus guidance explains.

Account for magnification, sensor, and working distance

Magnification determines how object-space detail maps to the sensor. Pixel pitch affects sampling, while sensor format and working distance constrain focal length and field coverage. Resolve those variables before aperture; the workflow for machine vision lens selection by field of view, sensor, and working distance covers that earlier decision.

Include conveyor speed and allowable motion blur

Set an exposure ceiling from the distance a feature may move during capture. Faster motion requires shorter exposure or controlled strobed illumination. If stopping down forces exposure beyond that ceiling, the extra focus tolerance has not improved the inspection—it has exchanged defocus risk for motion blur.

How to Balance Aperture and Depth of Field Step by Step

1. Start with the widest practical aperture. Establish a setting that meets exposure and motion requirements without saturation. Wide-open performance may be limited by lens aberrations, so treat it as the starting boundary rather than the presumed answer.
2. Stop down only until the required Depth of Field is covered. Test the nearest, nominal, and farthest object positions. Don’t add unused focus tolerance at the expense of light and contrast.
3. Check the diffraction budget. Estimate the Airy-disk diameter and compare it with pixel pitch, sampled feature width, and required contrast. Reject a setting when diffraction consumes the task’s resolution margin.
4. Recheck illumination at production speed. Include source variation, controller limits, exposure time, gain, and thermal constraints. Avoid validating with a brighter bench setup than the production station can sustain.
5. Validate center-to-edge performance. Use representative parts at field center, corners, and any algorithm-critical region. Real-lens aberrations can reduce sharpness even where a geometric depth calculation predicts acceptable focus.

The output should be a bounded setting such as “approved from f/5.6 through f/8 under the specified lighting and line speed,” not “use f/8.” The example is a documentation format, not a recommendation for every system.

Identify the Diffraction-Limited Aperture for Your Inspection Task

Relate diffraction blur to pixel size and feature contrast

For a circular aperture, Airy-disk diameter can be estimated as:

d ≈ 2.44 × λ × N

where d is image-plane diameter, λ is wavelength, and N is f-number. At 550 nm, the estimated diameters are about 5.4 µm at f/4, 10.7 µm at f/8, and 14.8 µm at f/11. These are screening calculations from the verified formula, not automatic pass/fail limits.

A 10.7 µm disk spans more than two 5 µm pixels, but that fact alone doesn’t disqualify f/8. Pixel integration, lens aberrations, sampling, feature size, illumination spectrum, and the algorithm’s contrast requirement all affect the result.

Why smaller pixels narrow the usable range

As pixels get smaller, diffraction blur covers more pixels at the same wavelength and f-number. The diffraction-limited aperture may therefore arrive earlier in pixel terms, even though smaller pixels provide denser sampling. Compare complete-system contrast at the required spatial frequency rather than equating pixel count with resolved detail.

Separate sampling, lens, and diffraction limits

• Sampling: the sensor lacks enough samples across the feature.
• Lens performance: aberrations or field variation reduce contrast.
• Diffraction: stopping down spreads point energy over a larger area.

The practical limit is their combined effect. Lens MTF data at relevant apertures and field positions can narrow the candidates, but representative-image testing remains necessary.

When diffraction can be acceptable

A setting beyond a pixel-based rule of thumb may still pass when the inspection uses large, high-contrast features. Conversely, a nominally acceptable Airy diameter may fail on faint scratches or precision edges. The decision boundary is feature detectability and measurement repeatability, not an isolated optical number.

Choose the Optimal F-Number for Machine Vision by Constraint

The optimal f-number for machine vision is therefore application-specific. If the candidate ranges for focus, motion, lighting, and contrast do not overlap, aperture adjustment alone cannot close the specification.

Lens, Camera, and Lighting Choices That Expand the Range

A suitable lens needs the required sensor coverage, resolution at the working distance, usable aperture control, and acceptable field performance. Compare available Machine Vision Lenses against those requirements rather than buying by focal length and mount alone.

When dimensional stability is the governing requirement, evaluate telecentric or low-distortion optics and verify their measurement behavior over the specified depth range. Before selecting a smaller aperture, improve illumination geometry, coupling efficiency, source placement, or strobe timing where feasible. More efficient delivery may preserve both exposure time and focus tolerance without excessive thermal load.

Changing magnification, working distance, or pixel pitch can create a usable aperture overlap when the original architecture cannot. For short-wave infrared designs, wavelength changes the diffraction estimate; use a wavelength-specific process such as SWIR lens selection for InGaAs inspection systems.

If no standard product meets the combined sensor, wavelength, envelope, focus, and resolution limits, assess optical lens assemblies for OEM vision systems rather than forcing the requirement onto an unsuitable catalog lens.

Common Aperture Selection Mistakes

• Stopping down to the minimum opening: geometric focus tolerance rises, but diffraction and illumination demand may erase useful contrast.
• Opening solely for brightness: exposure improves while part-height variation may move required surfaces outside acceptable focus.
• Trusting the aperture-ring marking at close focus: request manufacturer data or test the configured lens instead of assuming the nominal marking fully describes system behavior.
• Testing only at image center: corner performance may govern a code, measurement edge, or defect zone.
• Approving optics at reduced line speed: later speed increases can invalidate exposure and motion margins.
• Using a monitor as the acceptance tool: pass the production algorithm and repeatability criteria, not a subjective sharpness review.

Validate the Selected Aperture Under Production Conditions

Build a test matrix covering candidate f-numbers, nearest/nominal/farthest positions, relevant field locations, production line speed, and expected lighting variation. Measure feature contrast, detection rate, code readability, or dimensional repeatability using the actual acceptance criterion. Calculated near and far limits are only starting points; depth-of-field and circle-of-confusion fundamentals show that the result depends on the chosen blur criterion.

Record the lens identifier, focus lock, aperture setting, working distance, illumination settings, exposure, gain, wavelength, line speed, and approved adjustment limits. This turns aperture from an installer preference into a controlled system parameter.

For an aperture-range and lens-suitability review, submit the sensor format and pixel pitch, field of view, working distance, feature size, object-height variation, wavelength, and maximum line speed. Those inputs allow the focus, diffraction, exposure, and lens constraints to be reviewed together.