SWIR Camera Lens Illumination Guide: How to Choose Wavelength, Irradiance, Angle and Lighting Geometry for 900–1700 nm Inspection

A high-performance SWIR camera lens cannot create material contrast that never reaches it. In industrial short-wave infrared imaging, illumination is therefore not an accessory added after the camera and lens have been selected; it is one of the primary elements that determines whether the inspection works at all. A 900–1700 nm SWIR system may be intended to differentiate two visually identical materials, reveal moisture, inspect through selected polymers, identify contamination, examine semiconductor structures or classify a product according to spectral response, but every one of these applications depends on delivering useful photons to the target at wavelengths where the required materials behave differently. The illumination wavelength, irradiance, angle of incidence, distance from the target, spatial uniformity, reflection or transmission geometry and synchronization with exposure collectively determine the signal arriving at the lens. The Kyptec Automation® SWIR Camera Lens collection provides 8.5 mm, 12.5 mm, 25 mm, 35 mm and 50 mm focal-length options designed for 900–1700 nm imaging, giving OEMs a focused optical platform around which the illumination geometry can be engineered. Current product information confirms representative Kyptec Automation® SWIR lenses with 2 MP resolution, 2/3-inch sensor format, F1.4 aperture and C-Mount.

SWIR Illumination Should Be Selected From Material Contrast, Not Maximum Brightness

The most common illumination mistake is to choose the brightest available SWIR source rather than the wavelength that produces the strongest difference between the acceptable and defective states. Industrial SWIR imaging works because materials can absorb, reflect or transmit different fractions of incident radiation at different wavelengths. Two materials may look almost identical at 1050 nm but separate strongly at 1450 nm, while another material pair may behave in the opposite way. This means the goal is not maximum image brightness; it is maximum decision contrast.

A useful development sequence begins with representative good and defective samples illuminated at several candidate wavelengths. Measurements should then compare the separation between the two populations while considering image noise. The wavelength producing the greatest repeatable difference under realistic production variation becomes a stronger candidate than the wavelength simply generating the highest grayscale level. Industrial SWIR illumination commonly uses discrete wavelengths within the approximate 900–1700 nm region precisely because wavelength selection can exploit differences in material absorption and transmission.

Wavelength Selection Should Begin With the Inspection Question

Different inspection objectives require different spectral strategies. Moisture-sensitive inspection may benefit from a wavelength where water absorption becomes strong. Through-silicon semiconductor inspection requires a region where silicon becomes sufficiently transmissive for the hidden feature to influence the image. Polymer differentiation requires wavelengths where the materials' spectral responses diverge. Hidden-content inspection requires both package transmission and product contrast.

The practical sequence should therefore be:

inspection defect → material interaction → candidate wavelength → illumination → lens → camera response → production validation.

Reversing that sequence and selecting illumination solely because it is commonly available can produce an optically bright but diagnostically weak image.

Why Approximately 1450 nm Is Important for Moisture-Sensitive SWIR Inspection

Water exhibits strong absorption around approximately 1450 nm, making this region particularly useful for moisture-sensitive SWIR imaging. Water-rich regions can consequently return a substantially weaker signal than adjacent dry material under suitable illumination, creating contrast that may be difficult to obtain in visible imaging.

This does not mean every moisture application should automatically use 1450 nm. If the surrounding material also absorbs strongly at the same wavelength, both wet and dry regions may become too dark. The correct band should maximize the difference between the wet and dry states while still leaving sufficient signal above the camera noise floor.

Reference Wavelengths Can Make Material Decisions More Reliable

Sometimes one wavelength is strongly affected by the property of interest while another is relatively insensitive. Comparing both can reduce false variation caused by illumination intensity, surface distance or general reflectivity. For moisture inspection, for example, a water-sensitive band can be compared with a reference wavelength at which the same material remains relatively stable.

A normalized two-band feature can be represented as:

R = (Iref − Itarget) / (Iref + Itarget)

where (Iref) is the response at a reference wavelength and (Itarget) is the material-sensitive wavelength. The exact bands must be established experimentally. Real industrial implementations have used a water-sensitive wavelength around 1450 nm together with a reference wavelength to improve discrimination while simplifying the lighting architecture.

Broadband and Narrowband SWIR Illumination Solve Different Problems

Broadband illumination provides energy across a wide spectral region and is useful during development when engineers need to understand how a target behaves across many wavelengths. It can also support systems that acquire several spectral bands through filtering. The disadvantage is that wavelengths with strong material contrast can become averaged together with wavelengths providing little useful information.

Narrowband illumination concentrates optical energy around a selected wavelength. Once the important spectral region has been identified, narrowband lighting can improve selectivity and simplify production classification. It can also reduce unnecessary spectral information and improve rejection of ambient light when paired with appropriate optical filtering. The best production system is therefore not necessarily the one with the broadest spectral illumination; it is the one providing enough optical energy exactly where the required material decision is strongest.

Irradiance Determines Whether Spectral Contrast Becomes Usable Signal

Correct wavelength selection alone is not sufficient. If the target receives too little optical power, the resulting sensor signal may remain close to the noise floor. Increasing camera gain can brighten the displayed image, but it also amplifies noise and cannot recreate photons that were never captured. Adequate SWIR irradiance therefore provides the foundation for reliable material classification.

The illumination requirement should be evaluated at the actual production working distance, exposure time and aperture. A laboratory image acquired with a one-second exposure says little about a production line that permits only 200 microseconds. The available optical signal must support the required line speed.

Illumination Distance Has a Major Effect on Available Signal

Moving a compact source farther from the target generally reduces irradiance substantially as the emitted energy spreads over a larger area. Real industrial illuminators do not behave as perfect point sources at every distance, but the inverse-square concept remains useful for understanding why increasing working distance can dramatically reduce available photons.

If a prototype operates successfully with illumination positioned 100 mm from the object, relocating the light to 400 mm for mechanical clearance can produce a much weaker signal. The final illumination distance should therefore be established during system design rather than changed after exposure settings and classification thresholds have been finalized.

Working Distance Must Be Designed for Both Lens and Lighting

Camera working distance is often calculated only from field of view, but the lighting geometry may require additional space. A transmission system needs room behind the target. Angled reflection lighting may need significant lateral clearance. Diffuse illumination may occupy more physical volume than direct lighting.

The strongest OEM design considers the camera lens, target, illumination and machine structure together. Longer focal lengths within the Kyptec Automation® portfolio can be useful where the camera must remain farther away while maintaining a controlled field of view, whereas shorter focal lengths can support broad scenes when the machine permits closer positioning.

Reflection Illumination Is Useful When Only One Side of the Product Is Accessible

In reflection geometry, illumination and camera observe the target from the same general side. The SWIR source strikes the material, and the lens captures the portion that returns toward the camera. This is practical for conveyor inspection, surfaces, powders, composites, produce, industrial components and many applications where backlighting is impossible.

The difficulty is that the measured signal contains both material response and surface geometry. Glossy targets can return strong specular highlights, while rough targets scatter energy broadly. Illumination angle should therefore be chosen to emphasize the material-dependent signal rather than the direct mirror-like reflection.

Transmission Illumination Can Reveal Information Hidden From Surface Imaging

Transmission geometry places the illumination on the opposite side of the object from the SWIR camera lens. Radiation must travel through the target before reaching the imaging system. This can be powerful for selected polymers, packaged products, thin materials and semiconductor structures because internal absorption or hidden contents can modulate the transmitted signal.

The important design variable is total optical attenuation. If the target absorbs nearly all energy at the selected wavelength, transmission becomes signal-starved. If it transmits almost everything with little difference between good and defective states, contrast may also be weak. The strongest wavelength is often somewhere between these extremes.

Backlighting Should Be Uniform Across the Full Inspection Field

A transmission system can generate false defects if the backlight is brighter in the center than at the edges. A material classifier using raw intensity may then conclude that identical objects near the edge are compositionally different from objects in the center.

Illumination uniformity should therefore be measured across the complete field. Flat-field correction can reduce residual variation, but strong non-uniformity should be addressed optically before relying on software correction. The better the physical illumination, the less aggressively the algorithm needs to compensate.

Angle of Incidence Can Completely Change a SWIR Image

The same material illuminated from 0°, 30° and 60° can produce significantly different spatial intensity because surface reflection depends on geometry. For glossy or curved products, relatively small angle changes can move a specular highlight directly into or away from the lens.

During system development, illumination angle should be treated as an optimization variable just like wavelength. A technically strong prototype should compare several candidate angles and measure defect-to-background contrast rather than selecting the easiest mechanical position.

Oblique Illumination Can Emphasize Surface and Edge Features

Low-angle or oblique lighting can make shallow surface irregularities, layer edges or textured defects produce stronger spatial contrast because their local orientation changes the reflected signal. In applications where material information and surface structure are both important, carefully controlled oblique SWIR illumination can therefore provide more useful data than normal-incidence lighting.

However, oblique illumination can also increase shadowing. The approach should be used when directional surface information is desirable, not as a universal SWIR lighting rule.

Diffuse Illumination Can Reduce Position-Dependent Glare

A highly directional source can produce intense reflections from curved or glossy objects. Diffuse illumination distributes incident angles more broadly, which can reduce the dominance of one specular direction and produce more stable regional intensity.

For material-classification systems, stability is often more valuable than peak brightness. A slightly lower but spatially consistent signal can generate better production separation than a bright image containing strong glare that shifts whenever the object rotates.

Kyptec Automation® KL-1408 Can Support Wide Illuminated Inspection Fields

Where a large conveyor, broad material web or multiple objects need to fit into one image, the Kyptec Automation® KL-1408 8.5 MM SWIR Camera Lens can be evaluated for wider coverage. A wide lens also means the illumination must cover a larger target area uniformly. If the light source is narrow while the camera sees a broad field, edge regions may become under-illuminated and produce false classification differences.

The lighting dimensions should therefore be selected from the actual FOV rather than from product width alone. Every area contributing to an inspection decision needs adequate and repeatable irradiance.

Kyptec Automation® KL-1410 Can Balance Illumination Area and Spatial Detail

For medium-width fields, the Kyptec Automation® KL-1410 12.5 MM SWIR Camera Lens provides a useful intermediate geometry. Its verified live specifications include 12.5 mm focal length, 900–1700 nm operation, 2 MP resolution, F1.4 aperture, 2/3-inch sensor format and C-Mount. This type of configuration can reduce the illuminated area compared with a very wide system, helping concentrate available optical power while still retaining useful coverage.

F1.4 Helps Convert Illumination Into Sensor Signal

The maximum F1.4 aperture specified on current Kyptec Automation® SWIR lenses can be valuable where illumination is limited or high-speed acquisition demands a short exposure. Opening the aperture allows more SWIR energy to reach the sensor, potentially improving signal without increasing exposure duration.

The trade-off is depth of field. Products whose distance from the lens changes substantially may benefit from stopping the aperture down while increasing illumination. In production design, aperture and illumination should therefore be optimized together rather than independently.

Pulsed Illumination Can Support High-Speed Inspection

Where the product moves rapidly, a short intense illumination pulse synchronized with camera exposure can reduce motion blur. Instead of illuminating continuously at modest irradiance, the system delivers greater instantaneous energy during the brief acquisition interval.

The timing must be stable. If the illumination pulse shifts relative to exposure, image brightness can fluctuate frame by frame and create false material differences. Trigger synchronization should therefore be verified across maximum production speed.

Multi-Wavelength Lighting Requires Exposure Balancing

Two wavelengths rarely produce identical detected intensity because source efficiency, material reflectance, lens transmission and sensor responsivity can differ. Using one exposure setting for every wavelength may cause one band to saturate while another becomes too dark.

Each wavelength can therefore require independent illumination intensity or exposure timing. The goal is not identical grayscale values; it is sufficient dynamic range in every band used by the classifier.

Ambient SWIR Radiation Can Reduce Repeatability

Industrial environments may contain sunlight, hot objects or other infrared sources that contribute unwanted optical energy. If ambient radiation changes between shifts or when a machine door opens, the material classifier can drift.

Enclosures, optical filtering, controlled illumination and synchronized acquisition can reduce this dependency. During qualification, the system should be tested under realistic environmental changes instead of only in a dark laboratory.

Illumination Uniformity Should Be Measured, Not Assumed

A light source described as uniform can still produce gradients at the specific working distance used by the machine. The illumination field should be imaged using a stable reference target, and center-to-edge variation should be quantified.

If the inspection requires detecting a 3% material-intensity difference while illumination varies 15% across the field, raw classification will be unreliable. Physical lighting correction and flat-field normalization should reduce the background variation well below the defect signal.

The Kyptec Automation® KL-1412 Can Support Controlled High-Contrast SWIR Illumination

For smaller inspection regions where stronger spatial sampling and concentrated illumination are desirable, the Kyptec Automation® KL-1412 25 MM SWIR Camera Lens provides a tighter field than shorter focal-length options. Concentrating the optical system on a smaller region can also make illumination design easier because available SWIR energy is distributed across less physical area.

This can be valuable during feasibility studies where wavelength and material contrast need to be evaluated carefully before the inspection is expanded to a wider production field.

The Illumination Spot Should Extend Beyond the Decision Region

If the inspection region touches the edge of the illuminated field, small product-position changes can create significant brightness shifts. A stronger design includes illumination margin around the complete decision area so normal positioning tolerance does not move the target into a lower-irradiance zone.

The same principle applies to multiple objects. If objects can appear across a wide conveyor, every permitted position should receive adequate illumination.

Product Height Variation Changes Irradiance and Focus Simultaneously

A product moving closer to the illumination can receive more optical energy while also changing its working distance relative to the lens. This creates two simultaneous variations—brightness and focus—that can imitate material differences.

The illumination and lens should therefore be qualified across the full expected height tolerance. Where product-height variation is large, normalized spectral ratios or diffuse lighting may help reduce sensitivity to distance.

Longer Focal Lengths Can Help When SWIR Lights Occupy the Space Near the Target

Some machines cannot position both the camera and illumination close to the product because the lighting occupies significant physical space around the target. The Kyptec Automation® KL-1414 35 MM SWIR Camera Lens and Kyptec Automation® KL-1416 50 MM SWIR Camera Lens provide narrower optical options where a controlled field must be maintained from greater stand-off. Their current product pages verify 900–1700 nm coverage, F1.4 aperture, 2 MP resolution, 2/3-inch format and C-Mount.

These lenses do not reduce the amount of illumination required by material physics, but they can simplify machine packaging by allowing camera placement farther from the lighting assembly.

Illumination Should Be Qualified With the Final Lens and Camera

A wavelength study performed with laboratory equipment may identify promising material contrast, but the final production system includes the actual illumination spectrum, lens transmission, camera sensitivity, filter response and working distance. Those components collectively determine detected signal.

The final system should therefore be validated as one optical chain:

illumination output × material response × optical transmission × sensor response = usable image signal.

This relationship is why choosing a good SWIR camera lens and a good SWIR light independently is not enough; they must work together at the wavelength required by the application.

Why Kyptec Automation® Is a Strong Optical Platform for SWIR Illumination Design

The Kyptec Automation® SWIR Camera Lens collection provides five focal-length choices—8.5 mm, 12.5 mm, 25 mm, 35 mm and 50 mm—across the 900–1700 nm range. Current product pages confirm representative models with 2 MP resolution, 2/3-inch format, F1.4 aperture and C-Mount and describe the SWIR portfolio as suitable for material identification, moisture detection, semiconductor inspection and industrial quality control.

This range is especially useful because illumination geometry and lens geometry are interdependent. Wide inspection fields need wider and more uniform illumination, tighter fields can concentrate irradiance on smaller regions, and longer-working-distance installations require sufficient illumination power to preserve signal. Kyptec Automation® therefore gives machine builders a consistent SWIR-focused optical platform from which illumination wavelength, angle and field coverage can be engineered around the actual inspection problem.

Frequently Asked Questions About SWIR Camera Lens Illumination

1. What illumination wavelength should I use with a 900–1700 nm SWIR camera lens?

There is no universal best wavelength. The strongest choice is the wavelength where the required good and defective materials produce the greatest repeatable difference while still providing sufficient signal. Representative samples should therefore be tested at several candidate wavelengths before illumination hardware is finalized.

2. Is brighter SWIR illumination always better for machine vision?

No. Additional irradiance can improve signal when the image is photon-limited, but excessive illumination can saturate reflective regions and eliminate useful contrast. The correct intensity is one that keeps both acceptable and defective states within the usable dynamic range of the imaging system.

3. Should I use narrowband or broadband SWIR illumination?

Broadband lighting is useful during material characterization or when several spectral bands are required, while narrowband lighting can be stronger for production once the most useful wavelength has been identified. Narrowband illumination also allows the system to concentrate optical energy around the wavelength carrying the required material information.

4. Why does changing SWIR illumination wavelength change the appearance of the same product?

Materials absorb, reflect and transmit different fractions of radiation at different wavelengths. Changing illumination therefore changes which material property contributes most strongly to the image. This wavelength-dependent behaviour is the foundation of SWIR material inspection.

5. Why is 1450 nm commonly used for moisture inspection?

Water exhibits strong absorption around approximately 1450 nm, so water-rich regions can appear substantially darker than surrounding dry material under suitable SWIR illumination. The final application should still compare this wavelength with nearby reference bands because substrate absorption and available signal vary by material.

6. Can the same SWIR illumination wavelength be used for every material-identification application?

No. Different material pairs have different spectral relationships. A wavelength that separates Polymer A from Polymer B may provide poor contrast between two other materials. Wavelength selection should therefore be based on the hardest material pair in the actual application.

7. How does illumination distance affect SWIR inspection?

Greater source-to-target distance usually reduces irradiance because the emitted energy spreads over a larger area. A system moved farther from the product may therefore require higher source power, longer exposure or a wider aperture. Illumination should always be tested at the final production distance.

8. Should SWIR lighting be placed in front of or behind the object?

Use front-side reflection illumination when the relevant information comes from returned radiation and only one side is accessible. Use transmission or backlighting when the material permits SWIR radiation to pass through and internal absorption or hidden contents provide the desired contrast. The correct geometry depends on the target rather than on the lens alone.

9. How does illumination angle affect reflective products?

Glossy products can produce strong specular reflections when illumination is positioned at an angle that directs reflected energy into the camera. Changing the angle can dramatically reduce or increase glare. The illumination position should therefore be optimized using the real surface finish and orientation tolerance.

10. Why does a SWIR inspection work in the laboratory but fail at production speed?

Laboratory experiments often use longer exposures and stronger effective photon collection. Production motion requires shorter exposures, which can drastically reduce captured signal. A reliable system must therefore prove sufficient illumination at the maximum line speed rather than extrapolating from stationary tests.

11. How uniform should SWIR illumination be across the field of view?

It should be sufficiently uniform that residual brightness variation is small relative to the material contrast the system must detect. If the defect changes the signal by only a few percent, illumination gradients need to be controlled much more tightly than in an application where the defect creates very large contrast.

12. Can I correct non-uniform SWIR lighting entirely in software?

Flat-field correction can compensate for stable gradients, but severe or unstable non-uniformity should be fixed optically first. Software cannot recover adequate signal from a region that receives too few photons, and aggressive correction can amplify noise.

13. When is the Kyptec Automation® KL-1408 useful for SWIR illumination-based inspection?

The Kyptec Automation® KL-1408 8.5 MM SWIR Camera Lens is useful to evaluate where a broad inspection field must be captured. The accompanying SWIR illumination should cover that complete field with enough margin and uniformity that objects near the edges receive comparable irradiance to those near the center.

14. When can the Kyptec Automation® KL-1412 be useful for wavelength-development tests?

The Kyptec Automation® KL-1412 25 MM SWIR Camera Lens can be useful where a controlled material sample or smaller production region should occupy a larger part of the sensor. Tighter framing allows illumination to be concentrated over a smaller physical area while preserving stronger spatial sampling of the material response.

15. Does an F1.4 SWIR camera lens reduce the illumination power required?

Opening a lens to F1.4 allows more available optical energy to reach the sensor than using a smaller aperture, which can help when the system is signal-limited. It does not remove the need for sufficient target illumination, and depth-of-field requirements may prevent permanent operation at maximum aperture. The current Kyptec Automation® SWIR portfolio specifies F1.4 on its representative models.

16. Can I use one exposure for several SWIR wavelengths?

Sometimes, but it should not be assumed. Illumination output, material reflectance, lens throughput and camera sensitivity can vary significantly with wavelength. Independent exposure or source-intensity settings may be required so every band remains within useful dynamic range.

17. How do I choose SWIR lighting for a high-speed conveyor?

Start with the minimum defect, line speed and maximum acceptable motion blur. This establishes the longest practical exposure. Then determine the illumination irradiance needed to provide adequate signal during that exposure at the chosen wavelength and working distance. The final test should be performed at full production speed.

18. What should I test before purchasing illumination for a SWIR camera lens?

Test the real good and defective materials, candidate wavelengths, reflection or transmission geometry, final working distance, required FOV, expected surface orientation, line speed and ambient-light conditions. Illumination should be selected only after the system demonstrates sufficient contrast under these realistic constraints.

19. How should I select focal length together with SWIR illumination?

First define the required physical FOV and minimum defect size, then choose a focal length that provides sufficient sensor sampling. After that, size and position the illumination so the complete decision area receives adequate and repeatable irradiance. Wide fields require broad uniform lighting, while smaller fields can often use more concentrated illumination.

20. Why is Kyptec Automation® a strong choice when designing a 900–1700 nm illuminated inspection system?

Kyptec Automation® provides a dedicated SWIR Camera Lens collection covering 8.5 mm, 12.5 mm, 25 mm, 35 mm and 50 mm focal lengths for 900–1700 nm imaging. Current verified specifications show representative models with 2 MP resolution, 2/3-inch sensor format, F1.4 aperture and C-Mount. This focal-length range allows OEMs to design the illumination field and optical FOV together for wide conveyor inspection, controlled laboratory-to-production development, localized material analysis and longer-working-distance industrial systems.

Conclusion

A SWIR camera lens illumination system should be designed from the material decision outward. The wavelength determines which absorption, reflection or transmission differences become available; irradiance determines whether enough photons reach the imaging system; illumination angle controls glare, shadowing and surface response; and lighting geometry determines whether the inspection emphasizes reflection, transmission or another optical path. The lens then has to capture that material-sensitive signal with enough spatial resolution and throughput for the required production speed.

The strongest engineering workflow therefore begins by testing good and defective samples at candidate wavelengths, identifying the band with the largest stable decision margin, choosing reflection or transmission according to the physical target, determining the exposure required at production speed and then calculating the irradiance needed at the final working distance. Illumination uniformity, object-height variation, surface angle, ambient infrared, multi-wavelength exposure balance and machine clearance should be introduced during validation rather than treated as commissioning problems. In moisture-sensitive applications, strong absorption around approximately 1450 nm illustrates how wavelength can fundamentally change inspection contrast; in other materials, completely different bands may be optimal.

The Kyptec Automation® SWIR Camera Lens collection provides the optical flexibility needed to translate these illumination decisions into practical machine geometry. Wide 8.5 mm and 12.5 mm configurations can support larger illuminated fields, the 25 mm option can concentrate inspection on more controlled regions, and 35 mm or 50 mm focal lengths can help where the camera must remain farther from lighting assemblies or process hardware. With current representative models specified for 900–1700 nm, 2 MP resolution, 2/3-inch sensor format, F1.4 aperture and C-Mount, Kyptec Automation® offers a focused SWIR lens platform for OEMs building material-sensitive industrial inspection systems.

For machine builders and industrial buyers, the central principle is straightforward: do not ask how brightly a SWIR target can be illuminated; ask which wavelength and lighting geometry create the largest reliable difference between the states the machine must separate, then select the SWIR camera lens and irradiance needed to preserve that difference at production speed. When wavelength, irradiance, angle, working distance, field uniformity, aperture and lens geometry are engineered as one optical system, Kyptec Automation® SWIR Camera Lenses provide a strong foundation for repeatable 900–1700 nm inspection across material identification, moisture-sensitive analysis, hidden-content imaging, semiconductor inspection and advanced industrial quality control.