The 18-200mm performs much as we would expect from a superzoom; it's not outstanding, but should be quite usable within its limitations. In almost stereotypical Canon fashion, a weak performance at wideangle is balanced by a rather more convincing showing at telephoto; when compared to the Nikon AF-S 18-200mm F3.5-5.6 VR, the overall result therefore has to be declared as a score draw. Against either the EF-S 17-85mm F3.5-5.6 IS USM or EF-S 18-55mm F3.5-5.6 IS USM, the 18-200mm again looks poor at wideangle but better at longer focal lengths.
|Sharpness||At 18mm F3.5 sharpness is high in the centre but drops progressively towards the corners, however cross-frame performance improves markedly by F5.6. At longer focal lengths, this pattern of decent central sharpness but soft corners wide open continues; generally though the central sharp region generally covers a larger area of the frame. Overall results improve through to around 50mm, then dip towards the mid-tele range with the weakest focal length being 135mm, before rallying again to a respectable showing at 200mm. Generally, best results are obtained a stop or two down from wide open.|
|Chromatic Aberration||Not this lens's strongest suit, chromatic aberration is very high at wideangle with strong red/cyan fringing. In the middle of the zoom range (~50mm) CA becomes much better controlled, before rising again sharply towards telephoto; indeed the value of 0.26% in the blue channel at 200mm F11 represents a new record, and forces a recalibration of the scale in our lens review widget. Ouch!|
|Falloff||We consider falloff to become perceptible when the corner illumination falls to more than 1 stop below the centre. Here we see moderate vignetting wide open at both ends of the zoom range, however this never exceeds 1.3 stops and disappears rapidly on stopping down, so shouldn't really be an issue in normal use.|
|Distortion||Setting a brand new record for distortion, the 18-200mm weighs in with a decidedly disconcerting 3.4% barrel at wideangle; this is also a complex, wave-type distortion with re-correction towards the corners. The pattern shifts rapidly to pincushion, peaking at 50mm and -1.9% (again likely to be readily visible in real-world shots) before settling back to more reasonable levels at telephoto.|
Internal focus zooms have generally performed somewhat poorly in our macro tests, and the 18-200mm doesn't break any new ground here. The big surprise though is to find our sample focusing much closer than Canon's specification; closest focus is 34.5cm (with a 14cm working distance from the front of the lens to the subject), giving a maximum magnification of 0.31x (compared to the stated 0.45m and 0.24x magnification).
Image quality isn't too hot though; the centre is never properly sharp at any aperture, and curvature of field also appears rather high, with the corners of our test chart only coming into focus at F22. The image also shows mild barrel distortion, soft corners and significant chromatic aberration (mainly blue/yellow); not a great result at all.
|Macro - 73 x 49 mm coverage
Distortion: Moderate barrel
Corner softness: High
Focal length: 200mm
Specific image quality issues
As always, our studio tests are backed up by taking hundreds of photographs with the lens across a range of subjects, and examining them in detail. This allows us to confirm our studio observations, and identify any other issues which don't show up in the tests.
The 18-200mm IS's record-breaking barrel distortion at 18mm couldn't possibly pass without comment, and if you shoot at wideangle frequently you'll come up against its effects time and again. In any aspect of photography where straight lines really should be rendered as straight (perhaps most notably architecture and landscapes with clear horizons) they'll often end up strongly bowed, giving an unnervingly odd look to such shots.
This can be corrected in software, of course, particularly if you're prepared to shoot in RAW and process using Canon's free with camera Digital Photo Pro software (DPP), which makes it a very simple process indeed. However the complex, wave-type geometry of the distortion will make it more difficult to deal with using less-sophisticated tools such as Adobe's lens correction filter in Photoshop. This type of software correction will also cause additional blurring of the corners, as these already none-too-sharp regions are stretched further in pursuit of geometric perfection.
In the dull but functional 'brick wall' example below, the barrel distortion can be seen quite clearly, roll your mouse over the labels to see the effectiveness of the correction tool in DPP (we've also taken the opportunity to include an illustration of DPP's peripheral illumination function, as an intermediate step from the out-of-camera result).
Canon EOS 50D, 18mm F3.5; DPP Corrections
|Original||Peripheral Illumination||PI + Distortion|
We'd be amazed to find any zoom with an 18mm wide end that didn't show significant lateral chromatic aberration, and not surprisingly Canon hasn't been able to perform any miracles with the 18-200mm IS. At wideangle we see strong red/cyan fringing, and users hoping for a huge improvement in this regard over either the EF-S 18-55mm F3.5-5.6 IS or the EF-S 17-85mm F4-5.6 IS USM will be disappointed (although the green/magenta fringing produced by the 17-85mm at wideangle is arguably more unsightly). At the telephoto end, blue/yellow fringing is the order of the day, which is as broad as the record-breaking numbers obtained in our studio tests suggest; in context, we'd expect this to be visible to many viewers on an A4/12" x 8" print.
The silver lining is again that if you're prepared to shoot RAW and process in DPP, lateral CA can be removed to give essentially fringe-free images - literally at the click of a mouse button. It's not as elegant a solution as Nikon's in-camera CA correction (as debuted on the D300 and D3, and now also available on the D700 and D90), but it does work well, and for users of several (if not all) older camera models too. And while this processing can't cure any underlying softness of the image, it does give a visually much cleaner file if you're planning on making large prints.
The examples below illustrate the effect of chromatic aberration on real-world shots at the extremes of the zoom range, and its correction in DPP (in the wideangle shot, that barrel distortion is rearing its ugly head again, and portraying the Thames as ignoring the fundamentals of gravity).
18mm F8, Canon EOS 50D
200mm F8, Canon EOS 50D
|Camera JPEG||Camera JPEG|
|100% crop, left side of frame||100% crop, bottom left|
|RAW + CA correction in DPP, 100% crop||RAW + CA correction in DPP, 100% crop|
One area where the 18-200mm does perform well is in its handling of flare - even shooting without a hood in strong sunlight (a most-welcome rarity in the UK in October), we've had to push the lens hard to induce any flare problems at all. The worst-case scenario is with the sun in the corner of the frame at wideangle and the lens stopped right down; here we see broad areas of multi-coloured flare, but little of the diagonal patterning across the frame that we might expect from a 16-element design. When shot at more usual apertures (F5.6-8) the same image looks somewhat cleaner.
The lens also handles strongly side- or back-lit conditions rather well with, at worst, a little red-coloured veiling flare towards the edge of the frame when pointed almost directly at the sun at full telephoto. Overall quite a creditable performance here.
|18mm F22, Canon EOS 50D||200mm F8, Canon EOS 50D|
Background Blur ('bokeh')
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 large aperture. This lens can allow you to achieve quite substantially blurred backgrounds, especially at 200mm F5.6.
Bokeh is generally quite pleasing too, with attractive smoothing of relatively close out-of-focus backgrounds at macro distances, and without too much of the hard-edged effect often seen with bright highlights in distant backgrounds. Overall thumbs up here, probably one of the better consumer zooms in this regard.
|200mm F5.6, Canon EOS 50D||150mm F5.6, Canon EOS50D|
|50% crop||50% crop|
Optical Image Stabilization
The 18-200mm features Canon's latest compact image stabilization unit, which claims to allow hand-holding at shutter speeds four stops lower than usual before blur from camera shake becomes apparent. The stabilizer is almost completely silent and free of vibration in operation, and aside from the steadying effect on the viewfinder image, you’ll hardly notice it operating at all.
We've generally found the stabilization units in SLR lenses to be pretty effective in real-world use, and to quantify this, we subjected the 18-200mm to our studio image stabilization test. We used the wideangle and telephoto settings plus one intermediate focal length (50mm), with the EOS 50D as the test camera. With this combination we'd normally expect to get good results using a shutter speed of 1/50 sec at wideangle, and 1/400 sec at telephoto without image stabilization. The subject distance for these tests was approximately 2.5m.
We take 10 shots at each shutter speed and visually rate them for sharpness. Shots considered 'sharp' have no visible blur at the pixel level, and are therefore suitable for viewing or printing at the largest sizes, whereas files with 'mild blur' are only slightly soft, and perfectly usable for all but the most critical applications.
|18mm IS OFF||50mm IS OFF||200mm IS OFF|
|18mm IS ON||50mm IS ON||200mm IS ON|
We're used to seeing Canon's IS systems performing well, but the 18-200mm comprehensively wins the award for having the best in-lens stabilization we've yet seen. Indeed while we're rarely able to achieve results which match manufacturers' claims for their stabilization units, this lens comes very close. The most striking observation from this data is that with IS on (and across all focal lengths tested), we'd only consider 10% of the test shots taken at four stops slower than the expected 'safe' shutter speed as unusable, and half are essentially pixel-level sharp. This is an impressive degree of stabilization, which we've simply not experienced before.
In real-world shooting outside the studio, this performance is maintained, and indeed it's possible to achieve sharp results at even slower shutter speeds if you can brace yourself suitably against a handy wall or tree. This capability greatly increases the lens's versatility in low-light conditions, or indeed any other situation where a slow shutter speed would be desirable.