Whether you just bought a new lens from a reputable vendor or snagged a deal off an auction site, you’ve got to test it. Fortunately, you don’t need a crazy, dedicated setup just to check out your gear. In this guide, I’ll run through how I’d test a newly acquired lens in about 10 minutes, all without any fancy gear.
With the prices of today's lenses, you could be unboxing thousands of dollars worth of precisely manufactured glass, which might have just gotten plopped at the door by your mailman (not to mention the abuse the packaging took all the way through the shipping chain). Suffice to say, things can go wrong despite the best packaging efforts. The quicker you find an issue, the quicker you can get it addressed with the carrier or vendor. Furthermore, on high-resolution bodies and in certain optical formulas, even a tiny misalignment might mean big issues. That means just looking at the packaging isn’t enough to catch every problem.
Now, the testing that I'd advocate isn’t going to involve pulling out MTF charts and an optical testing bench. Instead, I’d suggest taking a methodical approach to checking out the lens’s functions and visible optical performance. You don’t have to look too hard to find issues, if there are any problems, you’ll be able to spot them with this method.
Before you get to shooting the test images, take a second and look over the lens. Are the front and rear glass clean and without damage? You can also look through the lens to check for things like fungus, although that’ll typically only be an issue on older lenses. Also, check the operation of all the rings. Do they zoom and focus smoothly? Are all the side switches operational? If the lens collapses, does that motion feel natural?
The Big Picture
To get started with the optical testing, grab your new lens, camera, and tripod. If you don’t have a tripod handy, you can easily make use of a stable surface like a countertop or table — the idea is just to remove any source of shake. You’ll also want some bright lighting. Sunlight works best if it’s not fluctuating, window light is acceptable, and good continuous lighting will also work. Lastly, you’ll want a good target. A high-contrast, clearly printed, flat surface works best. For that, you can purchase a test chart, like this Siemens Star Test, this lens focus card, or print your own from a file like this EIA Resolution chart on Wikipedia.
You’ll want to set up the chart on a wall. I like brick or stucco, as it makes checking the corners easier than a flat wall finish. Try your best to position your camera so that it’s facing the wall straight on. Make sure you’re far enough away that you’re not close to the lens’s minimum focus distance. If possible, have the same wall surface in all four corners of your frame.
On your camera, there are going to be a few key settings to configure: AF mode, shutter type, and shutter release delay. These might vary based on your model, so check your manual or Google your model and the relevant setting.
For overall settings, I like to test both wide open and at f/8, at base ISO. Shutter speed isn’t particularly relevant, but I’d suggest avoiding the range between about 1/5th of a second and 1/100th of a second to reduce the risk of motion induced by the mirror or shutter.
For AF, you’ll want to be in a single shot mode, with the camera set to release only after the focus is locked. This might be AF-S, One Shot, or similar. Also, make sure you understand what you’re testing based on the focus mode. If you’re in Live View, some cameras may be using contrast-detect AF instead of phase-detect. This can conceal issues with the focus calibration compared to phase-detect, so I’d recommend testing each mode, at least wide open, to check focus performance. For more on this issue, consult the section on analyzing your results.
If you can, configure your camera’s shutter to electronic front curtain mode and enable the mirror up or exposure delay modes. All of these can help drive out that last potential bit of blur from camera movement or activity. Ideally, you’ll be left with the sharpest possible image to draw your conclusions from.
Once your camera is set up, start capturing some images. For easy comparison later, I like to start zoomed out, testing both a wide aperture and narrow but not diffraction-impacted aperture, like f/8. From there, move on to a couple of other points along the zoom range, capturing both sets of aperture values.
The idea when capturing these images is that you can check for any disparities in the images. If one side is significantly sharper, you might have a decentered element. If you’ve got a significant softness issue, greater level of chromatic aberration, vignetting, or some other issue, you can probably spot it by comparing the shots at different focal lengths or aperture settings.
There are two additional sets of images you might want to capture: a high-speed burst at a narrow aperture and a few images at infinite focus. For the narrow aperture set, configure your camera to shoot continuously at a high frame rate, with manual exposure and narrow aperture. If there are inconsistencies in the exposure shot to shot, this may indicate the aperture blades are sticking. For the infinite focus images, just head outside and make sure the camera focuses accurately at infinity. In this set, you’re really just checking that the remaining focus range works as well as the up-close range you already tested.
After you’ve got your images captured, it really just comes down to some careful checking. The big issues you’re looking for in the wall series are drastically different levels of sharpness in the corners, unusually high vignetting for the lens in question, or aberrations on the target. Those aberrations can be a loss of contrast, missed focus, colored blur, etc. If you think you see something, take the time to re-run that set of images, as it might have just been a mistake in testing. If it’s still there, however, that lens might have an issue.
Fortunately, these issues are pretty rare. I’ve only had one lens come in with issues, which the vendor quickly handled. It’s still worth it to test, however, as it’s way better to find the issue at delivery, rather than halfway through a shoot.