How to Test a Lens for Issues
Related: build quality, depth of field, MTF and Micro Contrast, optics
Look for any of the following:
- asymmetric sharpness left to right or top to bottom;
- asymmetric color fringing left to right or top to bottom;
- a double image, halo or blur in one corner but not the others
- a noticeable shift in image position when swapping two lenses of identical focal length.
Some of these will be hard to detect unless subjects are chosen that have uniform detail.
A newspaper can work, though field curvature comes into play at close distance with some lenses, and the risk of misalignment of subject to camera, so it’s best to stick with more distant targets that allow near-infinity focus.
When you first get a new lens, take some “field shots” (normal subjects)—just about anything will do, but seek out subjects with uniform detail. Shoot at various focal lengths. And shoot wide open so that depth of field doesn’t mask problems. A “slow” lens like the Canon 70-200 f/4L masks problems with its significant depth of field at f/4; an f/2 lens has half the depth of field as does an f/4 lens.
Even a blind squirrel finds an acorn once in a while. A shot like the one below is a poor choice; there is no detail to inspect at the edges and corners. There is subject motion. And even a badly misaligned lens can produce a sharp image in one area of the frame (eg the roof tiles).
Check your images; is the sharpness symmetric left to right, top to bottom?
You must account for subject-to-camera alignment; don’t be misled by shooting a building at an angle for example. Remember, the focal plane is a plane (flat surface through 3 dimensional space). The focal plane and subject plane should be as close to perfectly parallel as feasible.
Field curvature can mislead, so look for symmetry.
If your field shots reveal a possible optical problem, find a subject that will clearly reveal it. Brick buildings, mosaics, rooftops, a lens test chart, etc can all make excellent choices.
Be aware that depth of field (diminishing sharpness in front of and behind the plane of focus) is very shallow with today’s high-resolution digital cameras. A slight upward or downward camera tilt shifts the plane of the image sensor relative ; be aware of how the subject aligns with it.
Don’t come to a conclusion from just one test. Verify results with more than on subject, on more than one day!
Zoom lenses should be shot at the wide, middle, and long end. It is not uncommon to see variation as the lens is zoomed; testing at one focal length is generally not adequate. Most problems are most obvious at the long end, so start there.
Examine test images (TIF, not JPEG) at the edges and corners (you did shoot RAW, didn’t you?!). All aspects of optical performance should be symmetric about the center of the frame. If this is not the case, something is “off”, or you’ve tested incorrectly.
Don’t assume that your brand-new lens (or one you’ve banged around) is optically good. Test it and see for yourself. Today’s modern lens designs are outstanding, but real lenses must be manufactured and transported, offering many possibilities for theoretical performance to drop considerably.
If you find that your new lens is optically out of whack, you might be able to exchange it for another copy—one good reason to work with a reputable vendor. Or you can send it in for service, which nearly always resolves the problem (in the author’s experience). Both approaches require retesting the lens; there is no guarantee with either approach.
The risk in sending it in for service is exceeding the return period of the vendor. Be sure to be as specific as possible about the problem, including photographs that show it clearly. This will help ensure that the problem is fixed.