To test a lens, put the camera on a tripod square to a flat, detailed target, focus with magnified live view, use a two second timer at base ISO, and shoot every full stop from wide open to f/11. Then compare the centre and all four corners at 100 percent. A healthy lens is sharpest in the centre, improves as you stop down, and has four corners that match each other. One side that stays soft is the warning sign.
The guide to the lens sweet spot uses a similar aperture series to find a lens’s best f-stop, and lens sharpness defines the term. This page is a diagnostic. It shows how to separate a lens fault from a technique problem, how to check for decentering, focus errors, field curvature and focus shift, how to inspect the glass and mechanics, and how to judge whether what you find is normal.
Rule out technique before you blame the lens
Most “soft lens” complaints turn out to be motion blur, missed focus or noise reduction. Control these seven things and whatever softness remains belongs to the lens.
- Support. Use a tripod on a hard floor. See how to use a tripod for a stable setup.
- Release. Use the self-timer at two seconds or a remote, so your finger is off the camera when the exposure starts.
- Shutter vibration. Use the electronic shutter or electronic first curtain if you have one. Curtain vibration blurs fine detail at moderate speeds, as explained in mechanical vs electronic shutter.
- Stabilisation. Follow your manual’s advice for tripod use. Many systems should be switched off on a tripod, as covered in image stabilization.
- ISO and light. Use base ISO and plenty of even light, so noise and noise reduction do not smear detail.
- Focus. Focus manually in live view at maximum magnification. This reads focus from the sensor itself and takes the autofocus system out of the test.
- Files. Shoot raw, or JPEG with a neutral picture style, and keep sharpening and lens corrections the same for every frame. Remove any filter from the lens, and check that front and rear glass are clean.
If handheld shots are soft but the tripod test is sharp, the lens is fine and the cause is in the camera shake or focusing habits, not the glass.
Setting up a flat target in ten minutes
You do not need a printed chart. You need a flat surface with fine, high-contrast detail right into the corners of the frame: a brick or stucco wall, a sheet of newspaper pages taped to a wall, or a large map.
Distance
Do not test at the closest focus distance. Most general-purpose lenses are designed to perform best at ordinary shooting distances, and many get weaker up close. A workable rule is to stand somewhere between 25 and 50 times the focal length from the target: about 1.25 to 2.5 metres for a 50mm lens, 5 to 10 metres for a 200mm lens. For long lenses, an outdoor wall is easier than a room.
Squaring the camera to the wall
This is the step that ruins most home tests. If the sensor is not parallel to the target, one side of the wall is farther away than the other, and at a wide aperture that side falls outside the depth of field. The result looks exactly like a faulty lens.
The mirror trick solves it. Tape a small flat mirror to the wall at the centre of your target. Look through the camera and move the tripod until you see the reflection of your own lens in the mirror, exactly in the centre of the frame. When the lens looks straight back at itself, the camera is square to the wall.
Then wait for even light. Open shade or an overcast day is ideal.
Shooting the series: centre and corners at every aperture
- Set aperture priority or manual, base ISO, and the lens wide open.
- Magnify the centre of the live view and focus carefully. Rock the focus ring back and forth through the sharpest point and settle in the middle.
- Shoot one frame at each full stop: for an f/1.8 lens that is f/1.8, f/2.8, f/4, f/5.6, f/8 and f/11.
- For a zoom, repeat the whole series at the widest, a middle and the longest focal length. Zooms often behave like three different lenses.
- Shoot the wide-open frame a second time after refocusing from scratch. If the two differ, your focusing is the variable. Repeat until they agree.
On the computer, view the files at 100 percent and look at five places in each: the centre and each corner, a little way in from the extreme edge. Lay the crops for one aperture side by side. You are answering three questions.
- Is the centre sharp wide open? A little softness and lower contrast at the widest aperture is normal, particularly on fast lenses. It should clean up within one or two stops.
- Do the corners catch up? Corners usually trail the centre wide open and improve steadily to around f/5.6 or f/8. By f/11 and beyond, diffraction softens everything slightly and evenly.
- Do the four corners match each other? This matters most. Corners may be softer than the centre, but they should be about equally soft.
Checking for a decentered lens
A lens is a stack of glass elements that must share one axis. If an element is tilted or shifted sideways, from a knock or from assembly tolerances, the lens is decentered. The image is no longer symmetrical: one side or one corner is softer than its opposite, often with smeared detail, and stopping down helps less than it should.
The flat target gives the first clue: one corner or one side clearly worse than the others at wide apertures. But a slightly crooked setup produces the same picture, so confirm it with a test that does not depend on alignment.
The distant scene test
Find a view with fine detail at a great distance right across the frame: a far treeline, a distant row of buildings, or a skyline. At a long distance everything is effectively at infinity, so there is no target to square up. Focus on the centre with magnified live view, shoot wide open and one stop down, and compare the left and right edges.
The upside-down test
Now turn the camera upside down on the tripod, or as close to 180 degrees as your tripod head allows, frame the same view and shoot again. Open both files and rotate the second so the scene is upright.
- If the soft area stays on the same part of the scene in both shots, the scene is the cause: haze, heat shimmer, or something nearer or farther than the rest.
- If the soft area swaps to the opposite side of the scene, it is travelling with the camera and lens. That points to the equipment.
To separate lens from camera, try another lens on the same body. If that one is symmetrical, the first lens is the problem. If every lens is soft on the same side, the mount or sensor alignment is suspect and the body needs attention.
The defocused point of light
At night, put a single distant point of light, such as a far street lamp, in the exact centre of the frame with the lens wide open. Defocus slightly in each direction. The blurred disc should be round, with the same brightness on all sides. A disc at the centre of the frame that is lopsided, brighter on one edge, or flared to one side suggests an element out of line. Only judge the centre: toward the edges of the frame, squashed “cat’s eye” discs are normal.
No lens is perfectly centred, and zooms vary more than primes. A small difference visible only when flicking between corner crops at 100 percent is ordinary manufacturing variation. A fault is a difference you can see at normal viewing size, or one corner still smeared at f/8.
Testing focus accuracy: front and back focus
A lens can be optically perfect and still give soft pictures if the autofocus puts the plane of focus slightly in front of or behind the subject. Test this separately from sharpness, wide open, at a typical working distance.
- Use the flat target and the tripod. Take one frame focused manually in magnified live view. This is your reference: the sharpest the lens can be.
- Switch to the autofocus mode you normally use, with a single central point on a clear, contrasty part of the target.
- Turn the focus ring to throw the lens well out of focus, let the camera refocus, and shoot. Do this five times from the near side and five times from the far side.
- Compare the ten autofocus frames with the reference.
If all ten match the reference, autofocus is accurate. If they are all slightly soft by the same amount, there is a consistent offset. If they vary from frame to frame, the issue is consistency: dim light, a low-contrast target, or a worn focus mechanism.
To find the direction of a consistent offset, lay a ruler or a strip of printed text on a table sloping away from the camera at about 45 degrees, and stand a small flat card upright beside one marked line. Autofocus on the card and see which part of the ruler is sharpest. Sharpest beyond the card is back focus, and sharpest nearer than the card is front focus. The cures, including autofocus fine-tune, are in how to fix common autofocus problems.
Field curvature and focus shift: two things that look like faults
Field curvature
Ideally a lens focused on a flat wall renders the whole wall sharp. In many lenses the surface of best focus is slightly bowl-shaped instead, so when the centre is in focus the corners are focused a little nearer or farther. On a flat target that reads as soft corners, even though the lens can resolve fine detail there.
The check is simple. After your centre-focused frame, move the magnified live view to a corner, refocus there, and shoot again. If the corner is now sharp and the centre has gone soft, you are looking at field curvature, not a weak corner. This is a design trait, common in wide and fast lenses. It matters most for flat subjects such as copy work, and stopping down hides it.
Focus shift
In some lenses, mostly fast primes, the plane of sharpest focus moves slightly as the aperture closes. If the camera focuses with the lens wide open and then stops down to take the picture, a subject that was focused perfectly at f/1.4 can be a little off at f/2.8, usually with the focus moved backward. The effect is strongest a stop or two down from wide open. By f/5.6 or so, depth of field normally covers it.
Test it with the sloping ruler. Focus once, wide open, on a marked line, then shoot at each aperture without touching the focus. If the sharpest zone creeps along the ruler as you stop down, the lens has focus shift. The practical cure is to focus at the aperture you will shoot at, which many mirrorless cameras do automatically, or to focus with magnified live view while the lens is stopped down. Understanding lens optics explains the aberrations behind both behaviours.
Quick checks for fringing, flare, vignetting and distortion
These four take a few minutes outdoors. Turn off lens corrections in the camera and the raw editor first, so you see what the glass is doing.
- Colour fringing. Photograph dark branches or a roof edge against bright sky, with some of that edge near the corners. Look for coloured outlines at 100 percent. Some fringing in the corners is normal and corrects easily; see chromatic aberration for the types and the fixes.
- Flare. Shoot with the sun or a bright lamp inside the frame and again with it just outside. Look for coloured ghost shapes and for a grey veil over the shadows. Some of both is normal. A veil that washes out the whole picture suggests haze inside the lens or a dirty element, and the page on lens flare covers how to control the rest.
- Vignetting. Photograph a clear patch of sky or an evenly lit plain wall wide open, then at f/5.6. Dark corners wide open that fade by two or three stops down are normal vignetting. Darkening on one side only is not.
- Distortion. Put a straight line, such as a door frame or the horizon, near the long edge of the frame. Bowing outward is barrel distortion and bowing inward is pincushion, both normal and correctable, as shown in lens distortion.
Physical inspection: glass, blades and mechanics
Do this in a dim room with a small torch. Open the aperture fully, shine the light through the lens from one end, and look in from the other at an angle. Then swap ends.
- Dust. A few specks inside are present in almost every lens, including unused ones, and have no visible effect on pictures.
- Haze. A milky or foggy film on an inner surface that lights up evenly in the torch beam. It lowers contrast and worsens flare, and needs professional cleaning.
- Fungus. Fine branching threads, like frost or a spider’s web, often starting at the edge of an element. Light fungus may not show in pictures yet, but it can spread and can etch coatings. The storage advice in the camera care guide explains how to prevent it.
- Scratches and coating marks. Fine marks on the front element usually have little effect beyond slightly more flare. Marks on the rear element matter more, because they sit closer to the sensor.
- Separation. A rainbow-coloured or cloudy patch, usually near the rim, where two cemented elements are coming apart.
- Aperture blades. They should be dry, with no shiny oil film, and should form an even opening. On lenses with a mechanical aperture lever, the blades should snap shut and open instantly when you flick it. Slow or sticky blades cause overexposure, because the aperture is still closing when the exposure begins.
- Focus and zoom rings. Turn each through its full range. The feel should be even, with no gritty spots, tight patches or free play.
- Barrel and mount. Check for dents on the filter thread, which hint at a drop, and for wobble at the mount or in the extending barrel.
Reading your results: normal variation or a real fault
| What you see | Most likely cause | Verdict |
|---|---|---|
| Centre slightly soft wide open, crisp one or two stops down | Residual aberrations at full aperture | Normal |
| All four corners softer than the centre, improving by f/8 | Ordinary off-axis aberrations | Normal |
| Corners soft with centre focus, sharp when focused on the corner | Field curvature | Normal design trait |
| One side or corner soft, and it follows the lens in the upside-down test | Decentered or tilted element | Fault if visible at normal size |
| Autofocus frames evenly softer than the live view reference | Front or back focus | Calibrate or service |
| Sharp wide open, slightly off one or two stops down, fine by f/5.6 | Focus shift | Design trait; focus at shooting aperture |
| Low contrast everywhere, strong veiling against the light | Internal haze, fungus or a dirty element | Clean or service |
| Frames brighter than expected at small apertures, worse in bursts | Slow or oily aperture blades | Service |
| Results change every time you repeat the test | Technique: focus, vibration or alignment | Fix the test first |
Common mistakes
Testing handheld. Even at fast shutter speeds, small differences in focus distance and movement swamp the differences you are trying to measure. Fix: tripod, timer, live view focus, every time.
Shooting a wall at an angle. A camera a few degrees off square makes one side soft at wide apertures. Fix: use the mirror trick, then confirm any asymmetry with the distant scene and upside-down tests.
Leaving a filter on. A cheap or dirty protective filter can add flare and softness that gets blamed on the lens. Fix: remove it for the test, then repeat with it on to see what it costs.
Judging only at 100 percent. At that magnification every lens has visible flaws. Fix: after the close inspection, view the frames at the size you print or share, which is the warning in pixel peeping.
Try this
Run the two fastest checks on your most-used lens. It takes about 15 minutes. Outdoors, find a distant view with fine detail across the whole frame. On a tripod, focus on the centre with magnified live view and shoot at the widest aperture and at f/5.6 with a two second timer. Turn the camera upside down, reframe the same view, refocus, and repeat both apertures. At home, rotate the upside-down files so the scene is upright, and compare the left and right edges of all four frames at 100 percent. Write down whether the two sides match, and whether any difference stays with the scene or swaps sides. You now have a baseline for this lens that you can repeat after any knock or repair.
Frequently asked questions
How do I know if my lens is soft or if it is me?
Shoot a static, detailed subject on a tripod with a timer, focused in magnified live view. If that frame is sharp, the lens is fine and the softness in your normal pictures comes from movement, missed focus or high ISO. If the tripod frame is also soft in the centre at f/5.6, the lens or its focus needs attention.
How can I tell if a lens is decentered?
Photograph a distant, detailed scene and compare opposite sides and corners at a wide aperture. If one side is clearly softer, repeat with the camera upside down. Softness that swaps sides in the scene, and so stays with the lens, indicates decentering.
Do I need a lens test chart?
No. A flat brick wall, taped-up newspaper or a distant skyline shows everything a home test can reveal. Charts help when you want repeatable numbers, but they are harder to align and are usually shot too close.
Is dust inside a lens a problem?
Almost never. Specks inside the lens are far out of focus and too small to show in pictures. Haze, fungus and oil on the aperture blades are the internal problems that do affect results.
Related guides
- Lens sweet spot: use the same aperture series to find the sharpest f-stop.
- How to take sharp photos: the technique side of sharpness, once the lens is cleared.
- How to fix common autofocus problems: what to do about front and back focus.
- Chromatic aberration: recognising and removing colour fringing.
- Understanding lens optics: the aberrations behind what your tests show.
- Using vintage lenses on mirrorless cameras: where inspecting old glass matters most.