How to find the fittest optical lenses for your laser or image machine

             Eight Steps to Selecting Optical Lenses: From Application to Coating

In real optical systems, focal length should generally not be determined in isolation from other conditions. For example, with the same 25 mm focal length lens, the priority of parameters will be significantly different when used in machine vision, laser focusing, LED collimation, and fiber coupling.

A more practical approach is to first determine the application and operating wavelength, then discuss working distance, field of view, focal length, NA/F/#, aberrations, materials, and coatings. First, the crucial question: What exactly is this lens intended to do? Different applications require different sets of parameters. Machine vision imaging primarily considers: field of view (FOV), working distance (WD), sensor size, resolution, distortion, and MTF. Laser focusing focuses on: wavelength, input beam diameter, focal length, NA, and target focal spot. If the laser power or energy density is high, further considerations are needed regarding materials, coatings, and laser damage threshold. Laser collimation requires consideration of: light source divergence angle, input beam conditions, focal length, effective aperture, and output wavefront quality. Fiber coupling depends on the specific fiber type, considering: fiber NA, core diameter or mode field diameter, target spot, and alignment tolerances. Therefore, before selecting a lens, answer a more fundamental question: What optical function do you want this lens to perform? This is often more important than immediately asking "how many millimeters of focal length are needed?"

II. What does focal length determine, and what does it not determine? Focal length is certainly important. However, a single focal length cannot explain the entire system. For imaging systems, assuming a fixed sensor size, it can generally be understood as follows: the shorter the focal length, the larger the field of view; the longer the focal length, the smaller the field of view; and with the same sensor size and target field of view, a longer focal length usually corresponds to a larger working distance. Therefore, when selecting a machine vision lens, at least the following should be provided: sensor size + working distance (WD) + target field of view (FOV) + required resolution. Simply knowing "25 mm lens" is not enough to determine the final area that can be captured. Moreover, real machine vision lenses are usually composed of multiple optical elements, and the position of the principal plane, distortion, and internal optical structure will all affect the actual results. Simple formulas can be used for preliminary estimations, but they cannot replace the actual optical parameters of the lens.

III. Why are NA and F/# indispensable? When dealing with focusing, microscopy, imaging, or fiber coupling, it's essential to understand two other crucial parameters: NA (Numerical Aperture) and F/# (F-number). These are not independent performance metrics; both are related to the system's cone angle, aperture, and luminous flux. Under specific paraxial conditions, they can be approximated. In engineering terms, a larger NA means the system can receive or output a larger beam cone. Under ideal diffraction constraints, a larger NA typically means stronger focusing power, a smaller theoretical focal spot, and higher theoretical spatial resolution. However, real systems are also limited by factors such as spherical aberration, diffraction, surface accuracy, beam quality, and assembly conditions. Therefore, the notion that "a larger NA is always better" is also incorrect. Optical design is essentially a trade-off process. IV. Plano-convex, biconvex, aspherical: how to choose? One important reason for different lens shapes is to accommodate different conjugate relationships and aberration requirements. 1. Plano-convex Lens: Plano-convex lenses are very common monolithic positive lenses, suitable for simple focusing and collimation. When focusing collimated light, to reduce spherical aberration, the curved surface is usually oriented towards the collimating beam, and the flat surface towards the focal point. Conversely, if used to collimate an approximate point source, the flat surface is generally oriented towards the point source, and the curved surface towards the collimating output direction. A seemingly simple installation orientation can actually affect spherical aberration performance. 2. Biconvex Lens: Biconvex lenses are also positive focal length lenses. When the conjugate distances of the object and image sides are close, such as when approaching symmetrical conjugate conditions, a biconvex structure is usually more suitable than a simple plano-convex structure. However, whether better image quality can be obtained still depends on curvature, material, aperture, conjugate ratio, and working distance. Performance cannot be judged solely by "symmetry." 3. Plano-concave and Biconcave Lenses: Negative lenses are mainly used for: beam divergence; changing the system's optical power; beam expansion; and forming beam shaping systems with positive lenses. Many laser beam expanding systems are essentially combinations of positive and negative optical elements. 4. Aspherical Lenses: The value of aspherical lenses lies in controlling geometric aberrations, especially spherical aberration, that are difficult for spherical lenses to handle. Therefore, they are commonly used in: high NA systems; compact optical paths; laser collimation and focusing; and designs where a reduction in the number of lenses is desired. However, there is an easily confused issue: aspherical lenses ≠ automatic chromatic aberration elimination. Aspherical lenses primarily address surface shape and geometric aberration issues. One of the root causes of chromatic aberration is the variation of the material's refractive index with wavelength, i.e., material dispersion. V. Why do broadband systems often require achromatic design? A single refractive lens typically does not have exactly the same focal length for different wavelengths of light. Simply put, blue, green, and red light may not focus at the exact same position. If the system uses: white light; broadband LEDs; multi-wavelength illumination; visible light imaging; spectral detection, chromatic aberration needs to be carefully considered. Common achromatic cemented doublets combine positive and negative lens elements with different refractive indices and dispersion characteristics, ensuring that the two design wavelengths have the same or very close focal points, thereby correcting the main longitudinal chromatic aberration. However, it's important to note that achromaticity does not mean the complete absence of chromatic aberration across the entire spectral range. The system may still exhibit residual secondary spectra and other aberrations. Therefore, simply replacing the glass with a "high-refractive-index glass" is not a solution for broadband systems. The real considerations are: material dispersion + optical power + overall optical structure. VI. Material Selection Cannot Be Based Solely on "Transparency" The selection of optical materials is essentially a choice of wavelength. In visible light and some near-infrared applications, various types of optical glass and fused silica remain very common. As the operating wavelength extends further into longer bands, especially into mid-infrared and long-wave infrared systems, material selection changes significantly. For example, silicon (Si), germanium (Ge), and zinc selenide (ZnSe) are common infrared optical materials, but their applicable wavelengths, refractive indices, thermal properties, and mechanical properties differ. They cannot be simply categorized as ordinary "colorless optical glass." When selecting materials, at least the following should be checked: the material's transmission performance, refractive index and dispersion temperature characteristics, mechanical properties, environmental tolerance, surface processing requirements, and whether coating is needed, especially for infrared systems. Choosing a material simply because it says "infrared material" is very risky. The true determinant of a material's usability is: target wavelength + environmental conditions + system performance requirements. Seven, AR antireflection coatings: wider wavelengths are not always better. When light enters glass or other optical media from air, Fresnel reflection occurs due to changes in refractive index. If a system contains multiple lenses, there will be multiple optical interfaces. These reflections can accumulate and cause: decreased transmittance, increased ghosting, and severe stray light. Therefore, AR antireflection coatings are used to reduce surface reflection in many transmissive optical systems with requirements for transmittance, stray light, or image quality. However, in engineering selection, simply stating "AR coating required" is far from sufficient. At a minimum, it must be clearly stated: operating wavelength, center wavelength, incident angle, AOI polarization state, single-sided or double-sided coating, reflectivity requirements, whether it's for high-power or high-energy-density laser applications, and further attention needs to be paid to the laser-induced damage threshold (LIDT), along with test wavelength, pulse width, spot size, and repetition frequency. For example, the design goals of AR film systems are completely different between 532 nm single-wavelength laser and 400–700 nm broadband visible light imaging. Therefore, AR films are not simply about pursuing "wider wavelengths for higher performance." The important thing is that the film system must match the actual spectrum, angle, polarization, and power conditions. Eight, in machine vision, calculating the focal length is just the beginning. There is a very common misconception in machine vision: once the focal length is calculated correctly, it is assumed that the lens has been selected. In fact, it's far from over.

The final image quality is also affected by many factors: lens MTF, distortion, depth of field, aberration, pixel size, defocus, illumination, and setup error, especially for dimensional measurements. If there is a significant height difference in the object being measured, ordinary non-telecentric lenses may experience magnification changes and perspective errors due to variations in working distance. In this case, further judgment is needed based on the measurement accuracy: whether object-side telecentric or other telecentric imaging schemes should be adopted. One of the core values of telecentric lenses is to reduce magnification changes and perspective errors caused by variations in object distance within a certain working range. However, telecentricity does not mean that the system is inherently "free of all distortion." Ultimately, the lens's telecentricity, distortion, MTF, working distance, and actual calibration results still need to be considered. IX. A More Practical Lens Selection Order If the entire process is compressed into an engineering logic, you can first consider the following order: Application

↓

Wavelength

↓

FOV / WD

↓

Focal Length

↓

NA / F / #

↓

Aberration

↓

Material

↓

Coating

Breaking it down, it essentially boils down to: ① First, determine the application: imaging, focusing, collimation, beam expansion, or coupling? ② Then, determine the wavelength: single-wavelength laser or broadband system? ③ Determine the geometric conditions: working distance, field of view, beam diameter, sensor size, and installation space. ④ Determine the focal length and aperture: based on system requirements, determine the focal length, effective aperture, and the required NA or F/#. ⑤ Check aberrations and actual performance: resolution, MTF, distortion, chromatic aberration, wavefront, or focal spot—to what extent are these required? ⑥ Finally, determine the materials and coatings, considering wavelength, environment, laser conditions, and cost. Of course, this isn't a fixed, unchanging process. Real optical design often requires iterative adjustments between these parameters. A change in one parameter can easily lead to changes in others. Lens selection is essentially about system matching—whether a lens is "good" or not is meaningless without considering the specific system. A low-NA ordinary plano-convex lens may be sufficient for a simple collimation system. However, when moving into high-NA laser focusing, high-resolution imaging, or precision measurement systems, aberrations, wavefronts, materials, coatings, assembly, and even temperature conditions can all become limiting factors. Therefore, truly effective selection isn't about finding the lens with the "highest parameters," but rather finding a suitable balance between wavelength, focal length, NA, aberrations, materials, coatings, mechanical conditions, and cost. If standard lenses don't meet these requirements, then the next step involves aspherical lenses, achromatic lenses, multi-element combinations, and even complete custom optical designs. For projects with clearly defined operating wavelengths, dimensions, materials, drawings, or coating specifications, GIAI Optoelectronics can evaluate processing and customization solutions for lenses and related optical components based on actual optical requirements.