The 10-24mm gives generally good, if not outstanding results; distortion is well controlled and central sharpness is high once stopped down F5.6, especially at 10mm. However sharpness wide open is distinctly low, and corners are soft at all settings. Our test sample also showed slight decentering, which was most visible at the 24mm setting with a rather soft right-hand side to the frame.
|Sharpness||Sharpness is generally poor wide open, and examination of the checkerboard crops shows low contrast due to halation effects (quite possibly a result of the slight decentering mentioned above). The lens improves dramatically on stopping down; central sharpness is extremely high at 10mm, but diminishes progressively at longer focal lengths; however the extreme corners remain soft at all settings. As expected for this type of lens, best results are obtained at F5.6-F11, with the optimum aperture being F8.|
|Chromatic Aberration||Chromatic aberration is kept reasonably low; fringing is visible at all focal lengths, but at levels no worse than those found on typical standard zooms. The CA profile at 10mm is rather complex with significant re-correction towards the corners (resulting in fringing which changes progressively from red/cyan, through green/magenta, to blue/yellow); this may well cause problems for software correction. However at longer focal lengths the pattern changes to red/cyan fringing with a simple linear profile, and therefore straightforward to correct in software.|
|Falloff||We consider falloff to start becoming a potential problem when the corner illumination falls to more than 1 stop below the center. The 10-24mm shows about 2 stops wide open at wideangle (with a slightly asymmetric pattern), but this decreases rapidly on stopping down, settling at 1.3 stops at F8 and smaller apertures. This will rarely be noticeable in real-world use, and usually swamped by natural lighting variations across such a wide field of view. At 13mm and longer, falloff is essentially a non-issue at any aperture.|
|Distortion||Distortion is kept impressively low throughout the range for such a wide angle lens; 1.1% barrel distortion at 10mm, increasing to a maximum of 1.6% at 13mm then decreasing again at longer focal lengths. In context, this is much lower than that usually seen at the wide end of typical standard zooms.|
|A wide angle zoom is unlikely to be your first port of call for macro work, but this lens doesn't do too badly anyway. Maximum magnification is 0.2x, achieved at 24mm and a closest focus distance of 24cm, giving a rather tight working distance of just 10cm from the subject to the front of the lens.
Image quality is actually pretty good; the image is soft wide open, but stop down to F8 and both central and corner sharpness is high. There' s noticeable barrel distortion, and a little blue/yellow chromatic aberration.
|Macro - 118 x 78 mm coverage
Distortion: Moderate barrel
Corner softness: Low
Focal length: 24mm (36 mm equiv)
FX (Full Frame) Coverage
Both the Nikon and Canon mount versions of this lens will mount on full-frame DSLRs; on Nikon cameras (D3, D3X, D700) DX crop mode will be automatically selected (and the camera will therefore shoot at reduced resolution). The lens's image circle obviously doesn't fully cover the 35mm full frame format at wideangle, but as is common for this type of lens, the vignetting progressively diminishes on zooming in, and coverage is essentially complete at focal lengths of 15mm and longer. We're not going to formally test a DX lens on full-frame, but impressions are that the lens would be quite useable in an emergency or for non-critical applications.
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 Tamron 10-24mm performed perfectly well in normal use, with no serious vices.
Control of flare is a critical feature of a superwide lens; with such a broad view of the world, bright light sources will find themselves in the frame on a regular basis. Overall the 10-24mm does pretty well in this regard, retaining contrast well in side- and back-lit conditions, even in the bright, low winter sun.
At its widest settings, the lens gives a signature red arc-shaped flare pattern when bright light sources are placed either in or just outside the frame; this progressively reduces in size and intensity on stopping down. The lens also deals rather well with strongly backlit situations at the long end of the zoom range (and therefore with the sun impinging directly on the front element); the sample below loses some contrast on the right hand side, and there is one small flare spot visible, but overall it's a pretty good result.
|10mm F8, sun in corner of frame||24mm F6.3, strong backlight|
As we'd expect from a wideangle zoom, the 10-24mm suffers from its fair share of chromatic aberration, although no more than the typical DSLR kit lens. The samples and crops below give an idea of what to expect; at 16mm (around the point where CA is strongest) there's some fairly obvious red/cyan fringing. At the widest setting of 10mm, the fringing is changes in color across the frame; the selected crop shows a rather abrupt change from green/magenta to blue/yellow fringing in the extreme corners. The crops at the bottom show how effectively Nikon's in-camera CA correction on the D300 (also present on the D90) removes this fringing; this also gives an idea of how well software correction in post-processing is likely to work.
|F8, Nikon D300||F11, Nikon D300|
|100% crop, upper left||100% crop, top left corner|
|100% crop, Nikon D300 JPEG||100% crop, Nikon D300 JPEG|
Our studio tests reveal a certain propensity for soft corners, and the samples below illustrate how this looks in practice. The 100% crops come from the three regions outlined in red on the the thumbnail image.
At 10mm, central sharpness is impressively high, and detail continues to hold up well out to the location of the second crop shown below (approximately 55% along the image diagonal). Beyond this sharpness gradually drops towards the extreme corners, which are very soft indeed. The story at 24mm is slightly different; here the centre is not quite as good as at 10mm, and sharpness gradually declines across the frame, with the corners again extremely soft. (Our review sample also showed some asymmetry at 24mm, with the top right corner worst affected, showing extreme softness and ghosting as shown below).
|F8, Nikon D300||F8, Nikon D300|
|100% crop, centre||100% crop, centre|
|100% crop, upper right||100% crop, upper right|
|100% crop, top right||100% crop, top right|
Overall this lens may not be the best choice if you demand the extreme corners to be perfectly sharp, but it is also worth bearing in mind that just a slight crop will remove the most obviously soft regions.
Softness wide open
The second clear issue identified in our chart tests is softness wide open. This turns out to be a case of extremely low local contrast due to halation, with fine detail still being recorded. The crops below show this clearly, with two images taken seconds apart with the lens wide open at F3.8, then stopped down to F8. The 100% crop from the centre of the frame looks somewhat 'hazy' at F3.8, and the loss of contrast on the underside of the bridge walkways (on the left side of the crop) is very obvious. The 10-24mm is therefore best shot stopped down to at least F5.6 whenever possible; however as it's the kind of lens you'll probably be shooting at F8 anyway, this isn't really a significant problem in actual use. Afflicted images can also be 'rescued' to some extent by the application of a broad radius unsharp mask; amount = 10, radius = 50, threshold = 0 or similar. (It's also quite possible that this is to some extent a sample-specific issue, resulting from the decentering observed in our studio tests.)
|Nikon D300||Nikon D300|
|100% crop, centre||100% crop, centre|
|100% crop, left of frame||100% crop, left of frame|