Fujifilm XF 55-200mm F3.5-4.8 R LM OIS review
Photographic Tests (continued)
In normal shooting, you'll see essentially no distortion from the XF 55-200mm, with straight lines towards the edge of the frame looking perfectly correct. You'll also see this with RAW files processed through RAW File Converter EX (Fujifilm's version of SilkyPix that's bundled with X system cameras), Adobe Camera Raw and Capture One at its default settings. However you can turn off distortion correction in Capture One, at which point it becomes clear that this geometric perfection is due to the X system's integrated software distortion correction.
With distortion correction disabled, images from the XF 55-200mm show a small degree of pincushion distortion at all focal lengths. It's never huge, though, and not obviously worse than you might get from a typical SLR telephoto zoom. The rollover below illustrates this in the middle of the zoom range; distortion is fractionally weaker at 55mm and stronger at 200mm, but not by much.
|95mm, Distortion Corrected||95mm, Distortion Uncorrected|
One consequence of this is approach that, because the correction algorithm slightly 'stretches' the centre of the frame, the uncorrected version looks sharper if you view it at 100%. This can be seen from the crops below. So if you're after the very sharpest results and distortion isn't a problem, it might be worth turning the correction off. Note though you need to look very closely at your images, or make huge prints, for this to really matter.
|Camera JPEG, 100% crop||Uncorrected RAW conversion, 100% crop|
The story for vignetting is essentially the same as for distortion. In normal shooting you'll never notice it, because the camera's JPEG engine removes it automatically, as do RAW File Converter EX and Adobe Camera RAW. But Capture One doesn't, which means we can see how the lens itself behaves.
The lens shows some vignetting wide open at all focal lengths, but again, it's not unduly strong. It's most pronounced when shooting wide open at the long end, where there's visible, but far from excessive shading towards the edges and corners of the frame. In the rollover below we're comparing the out-of-camera JPEG with a Capture One RAW conversion that has neither vignetting nor distortion correction applied.
|200mm F4.8, JPEG (vignetting corrected)||RAW + Capture One 7.1 (uncorrected)|
The vignetting is quite obvious in the uncorrected version here, partly because we're comparing directly with the corrected version, but also because the background is evenly-toned. The key point though is that it's not particularly objectionable if you look at the uncorrected version in isolation, because the falloff in illumination is quite gradual across the frame, rather than a precipitous dropoff in the extreme corners.
One genuinely desirable, but difficult to measure aspect of a lens's performance is the ability to deliver smoothly blurred out-of-focus regions when trying to isolate a subject from the background, generally when using a long focal length and/or a large aperture. The degree of blur isn't all that matters, though, but also its aesthetic quality (which is what the word 'bokeh' refers to). This changes with focal length, aperture, focus distance and background distance, so isn't possible to pin down in simple terms. The best we can usually do is look at lots of pictures and make some sort of judgement.
The XF 55-200mm tends to give rather mixed results - in general distant backgrounds are attractively rendered, but regions which are only slightly out-of-focus look rather 'busy' and distracting. For example the shot of the lioness above has rather unattractive blurring of the slightly out-of-focus twigs, but a shift in composition to include a more distant background gives much better results. The samples below give some flavour of the results you can get.
|200mm F4.8, X-Pro1||Background crop|
|67mm F4, X-Pro1||Background crop|
|200mm F4.8, X-Pro1||Background crop|
The XF 55-200mm uses linear stepper motors for autofocus, which means it's almost silent in operation. The AF speed is pretty much what we'd expect from the X-system - not the fastest in class, but for most purposes not unusably slow either. It's not going to compete with a similar-priced SLR setup for focus tracking on moving subjects, but we suspect even Fujifilm would admit that's not a strength of the system right now. We've also found that the lens can struggle to acquire focus more than usual when shifting from close subjects to distant or vice versa, and end up hunting back and forward, but to be fair this probably more a fault of the camera than the lens itself.
What you do get in return, though, are some pretty significant advantages: consistently accurate focusing (an inherent characteristic of contrast-detection autofocus systems), along with the ability to place your autofocus point almost anywhere in the frame, and use small focus areas when necessary for increased precision. This is helped by the fact that with their latest firmware updates, the X-Pro1 and X-E1 now offer one of the quickest interfaces around for setting the size and position of the AF point with the camera to your eye.
|200mm F4.8, 1/500 sec ISO 400, X-Pro1||100% crop|
In this example, shot with the X-Pro1, the AF point was placed over the heron's head, and set to the second-smallest of the five available sizes to discourage the camera from shifting focus to the background. This allowed multiple shots while maintaining the composition, and with the ability to quickly refocus as the bird moved around. SLRs tend to be faster at focusing, but don't offer this sort of control over AF point positioning.
Fujifilm makes bold claims about the effectiveness of the 55-200mm's in-lens OIS system, calling it 'class-leading' with a claimed 4.5 stop effectiveness using CIPA standard tests. Again we've not formally tested this, just gone out and shot in the real world, deliberately not worrying too much about shutter speeds except when it's been necessary to freeze subject motion (which with telezooms, tends to be quite often).
In practice we've found the optical stabilization to perform very well indeed, allowing the use of much slower shutter speeds (and therefore lower ISOs) than we'd be able to use without it. It also does an extremely good job of stabilizing the viewfinder image, allowing accurate composition. Below is a selection of examples to illustrate how well the OIS works. All of these were, of course, shot hand-hold.
|55mm, 1/15sec F5.6, X-Pro1||100% crop|
|105mm, 1/10sec F5.6, X-Pro1||100% crop|
|150mm, 1/45sec F8, X-Pro1||100% crop|
|200mm, 1/30sec F4.8, X-Pro1||100% crop|
These show that's it's certainly possible to shoot using shutter speeds up to four stops slower than usual without seeing troublesome image blurring due to camera shake. This is pretty much as good as any lens we've used, in terms of stabilization effectiveness.