O7) How to choose a compatible lens (an example)

Step1: Confirm that the camera and the lens meet the “photonics” requirements of your application.

Since the ultimate goal is to obtain good image quality – a sufficiently high Signal to Noise Ratio – from your imaging system, it is good to confirm whether the image sensor in the camera has sufficient responsivity at the wavelengths of interest. The QE curve of the imager is a good place to start. You can find this information via these links:

QE Curves for CMOS Imagers

QE Curves for sCMOS Imagers

Or, if you have already selected a camera, you can find the relevant QE curve on the product page for a specific camera – under the tab that is labelled “QE Curves”.

Are the other characteristics of the image sensor (e.g., optical format/imager diagonal, pixel size, read noise, dark current, frame rate) are a good match for your application?

Note that the wavelengths of interest impact the choice of lenses too. If your application operates in the Visible region of the optical spectrum, then it is relatively easy to find many families of lenses that can satisfy the requirements. If your application requires operation in the UV or the SWIR region, please contact us for assistance.

Step2: Confirm that the camera and the lens meet the “optical resolution” requirements of your application.

As detailed in the article titled “Optical Resolution of a Camera and Lens System” the limiting resolution of a camera and lens system may be impacted by the pixel size of the camera OR the Optical Resolution of the lens (or other optics used with the camera). It is useful to compare the Optical Resolution of the lens (usually specified in line pairs per mm, lp/mm) with the Nyquist limited resolution determined by the pixel size of the image sensor in the camera.

The Nyquist limited resolution of the image sensor may be estimated as being equal to 500/(pixel-size-in-μm).

The limiting optical resolution of the system is determined by when of the two components (lens or camera) is lower.

If the lens can resolve, for example, 200lp/mm but the pixel size of the camera is 4.5μm (such that it can only resolve 500/4.5 = 111 lp/mm), then the optical resolution of the lens + camera system will be limited to that of the camera, i.e. 111 lp/mm in our example. Such a system is considered “camera pixel size limited” in optical resolution. This is usually a desirable condition, since we want to make sure that the camera can discern the smallest features that the lens can resolve.

Note that lenses with lower optical resolution can be used – in which case the optical resolution of the system would be “lens limited”. This is not usually desirable: most users would prefer to use a lens that does not limit the optical resolution of the camera, usually the more expensive of the two components. However, if optical resolution is not a paramount concern, cost-effective, lower resolution lenses are quite usable in such applications.

The main takeaway here is that one should look at the pixel-size of the camera AND the optical resolution of the lens for this simple comparison. 

Please contact us with your application specific questions. Our lens specialists have access to a wide array of products, including some (for example, telecentric lenses) that are not shown on our website.

Step 3: Based on the optical format of the camera, select a lens with a focal length that provides a desired Field of View (FOV) at the typical Working Distance (WD) of your application.

Note that camera lenses usually have a specified minimum working distance. One should generally ensure that the WD of the application is greater than or equal to the minimum WD of the lens. It is possible to add “spacers” on the camera in order to have a lens produce a focused image at a working distance that is less than its minimum working distance. Please contact us for assistance, if this is necessary in your application.

Our camera pages include a tab labeled “Select a Compatible Lens”, from which you can select a lens with a suitable focal length such that the camera and lens combination provides a desired FOV at the typical WD of the application.

For example, let us consider an application for which the “photonics” of a 7MPix Pregius Gen3 imager based camera are well-matched. Let’s say that the application requires the inspection of samples that are known to be a maximum of 3″ x 2″ (75mm x 50mm) in size with the smallest possible WD. Let’s also say that it is important to resolve features in the sample that are known to be 0.01″ (0.25mm) in size. The diagonal FOV must exceed the diagonal size of the sample, which is 3.6″ or 92mm.

From the table below, we can see that an NMV-16M1.1 lens with a focal length of 16mm (minimum WD of 100mm) satisfies the above requirement. When used with this camera, would produce an FOV with a diagonal dimension of 112mm, and horizontal dimensions of 92mm x 63mm, which exceeds the FOV requirement of 75mm x 50mm. Exceeding the FOV requirement by some margin is always desirable: it is particularly helpful in light of the optical resolution considerations that are examined next.

Step 4: Confirm the Optical Resolution.

As indicated on the table, the Nyquist criteria for a 4.5μm pixel size is met by the lens, since its optical resolution is 160 lp/mm at the center of the lens. Note that the peripheral optical resolution of the lens does fall off to 100 lp/mm which means that one can expect the optical resolution of the system to be lens-limited towards the edges of the image.

The fall off in optical resolution towards the periphery of the image should not cause a major issue in this application because the FOV of the system exceeds the size of the sample by a good margin. Therefore, we can say that system is camera-pixel-limited in its optical resolution, a desirable condition.

Step 5: For a final sanity check, confirm the minimum resolvable feature size.

Note that the table provides the minimum resolvable feature size of 0.06mm for the camera + lens combination. This is considerably smaller than the requirement to resolve 0.25mm features. The table also informs us that a 1mm feature in the sample will be represented by 36 pixels in the image, for this camera + lens combination. This gives us an estimate, that the smallest (0.25mm) features in our sample will be represented by ~9pixels in the image.