Canon EF 14mm f/2.8L II USM

Canon EF 14mm f/2.8L II USM Review

February 17, 2008
by Andrew Alexander

Introduction
The Canon EF 14mm ƒ/2.8 II USM updates its predecessor of the same name, adding a 'II' designation to reflect the overhaul of the lens. Specifically, Canon indicates that the optical layout of the lenses has been completely redesigned. Where the original EF 14mm had one aspherical element, version two now contains two aspherical elements and two ultra-low dispersion glass elements. The new lens also adds an aperture blade, going from 5 to 6.

The 14mm ƒ/2.8 features an integrated lens hood without a filter ring; filters may be added to a rear element gelatin filter holder. As an EF lens, the image circle is full-frame 35mm. The lens is available now with a MSRP of $2,199.

Sharpness
The type II 14mm ƒ/2.8 shows a dramatic improvement in sharpness over the previous model. On a sub-frame sensor the original 14mm exhibited corner softness at every aperture setting; for the type II, even wide open at ƒ/2.8 there is very little corner softness (1 blur unit in the center, and 3 units in the corners). By ƒ/5.6 corner softness is negligible, and by ƒ/8 it's non-existant. Diffraction limiting begins to set in at ƒ/16, and ƒ/22 produces uniformly softer images - however, this softness doesn't even reach 3 units on the blur scale. Overall, excellent performance.

The performance full-frame is even better. On a full-frame lens, corner softness of the type I 14mm ƒ/2.8 was extreme; for the type II version, it's significant, but nowhere near as bad (the type I approached 11 blur units on our scale out of 12; the type II approaches 5). Similar to the sub-frame performance, the type II 14mm is essentially tack-sharp by ƒ/5.6.

Chromatic Aberration
Wide angle lenses typically show problems with chromatic aberration, presenting a special challenge to lens designers. The original 14mm ƒ/2.8 showed particularly awful results for CA - at ƒ/8 and above its results in the corners were off our chart. The type II 14mm ƒ/2.8 has addressed this issue. It's not perfect: at ƒ/2.8 and ƒ/4 there are some noticeable traces of chromatic aberration in both the corners and the center. By ƒ/5.6 these are reigned in to below 6/100ths of a percent of frame height, where they remain pretty much constant.

There is about the same amount of chromatic aberration between sub-frame and full-frame testing, though full-frame testing shows less CA evident in the central region of the frame (probably because the frame is larger). Compared with the type I version of the lens, full-frame performance by the type II is also dramatically improved.

Shading (''Vignetting'')
Version II of the 14mm ƒ/2.8 has also improved on its resistance to corner shading. This wasn't a huge problem with the original version, with only a half-stop light difference in the corners when used wide open at ƒ/2.8. Version II has reduced this light difference to just over a quarter stop. At all other apertures, corner darkness isn't an issue with either lens.

Mounting an ultra-wide angle lens on a full-frame body is asking for issues with corner darkness, and there's not much the lens designer seems to be able to do about it. Corner shading with the lens wide open (ƒ/2.8) on a full-frame body is right off the top of our vignetting chart, clocking in at 1.9 stops darker in the corners than in the center. This effect decreases in intensity as you stop down, but never declines below a half-stop difference by ƒ/8. This performance is on par with the performance of the type I 14mm mounted on a full-frame lens.

Distortion
The new element layout has had a positive effect on reducing image distortion. Being a rectilinear wide-angle lens, the 14mm is optimized to keep straight lines straight instead of warped like a fisheye lens: this is done extremely well, with only just over 0.25% barrel distortion in the corners, and less than that overall. This wasn't really a problem in the previous version of the lens, with just under 0.5% barrel distortion in the corners, but it is a testament to the improvements Canon has made with this lens.

On a full-frame body distortion is slightly accentuated, never exceeding 0.5% barrel distortion in the corners. In this manner the type II lens is very similar to the type I.

Macro
You don't really think of macro when you think of wide angle lenses, and the 14mm ƒ/2.8 is no exception. It can produce a magnification ratio of 0.15x, with a close focus distance of 20 cm (almost 8 inches). It should be noted that this is an improvement over the version I of the lens, which had a magnfication ratio of 0.10x, and a close focus distance of 25 cm.

Build Quality and Handling
The EF 14mm ƒ/2.8 lens is a member of Canon's L-series of pro lenses, and as such is built with dust and moisture seals. For its size, the lens is quite heavy at 645 grams (over 22 ounces) testifying to its mostly metal build; it's actually heavier than the previous type I model, by 85 grams. The focus ring covers a very long tracking distance - it turns a full ninety degrees - and is very smooth. Autofocus operation was very quick, well under a second to rack through the entire focus range. As a USM lens, focus action is conducted very quietly. The focus ring doesn't turn while autofocusing, and the user can override autofocus and focus manually at any point by turning the focus ring. The lens comes equipped with a distance indicator scale (with infrared index) and an autofocus disable toggle switch. Canon has moved away from aperture rings, and the 14mm ƒ/2.8 doesn't have one (and the previous version didn't either).

As previously mentioned, the lens doesn't have a front filter ring; ultrawides rarely do, as the front lens element is so bulbously huge that mounting a filter in front of it is highly impractical. However, it does take rear-mounted gelatin filter inserts. Ultrawides are also notoriously susceptible to flare, as there is just so much surface area for stray light to get in and bounce around the lens elements. Unfortunately we don't (yet) test for flare, and any commentary would be purely subjective. What I can tell you is that Canon indicates in their press literature that the EF 14mm ƒ/2.8 II USM has a special Super Spectra coating, which will ''help ensure accurate colour balance and high contrast, and suppress flare and ghosting by absorbing light.''

The lens will mount on all of Canon's bodies, but for the price, it wouldn't be very productive to mount it on all but full-frame bodies; on a 1D-series camera, the ''crop factor'' of 1.3x will give you a framing equivalent of 18mm; on a Canon APS-C sensor body, the factor of 1.6x would give you an equivalent of a just over 22mm. With a price of over two thousand dollars, you would be much better off with the designed-for-digital 10-22mm EF-S.

Alternatives

Canon EF 14mm ƒ/2.8L USM ~$1,700
As you would imagine this is the easiest alternative to consider; as we have mentioned above, the type II version of the Canon EF 14mm is well-improved in almost every respect. I would imagine the introduction of the new 14mm would see a price reduction of the old in the used market, but if you're looking to buy and you've got the money, the extra $400 is well spent.

Sigma 14mm ƒ/2.8 EX Aspherical HSM ~$800
We haven't yet tested the Sigma 14mm, but at one-third the price of the Canon version, it may be appealing. User comments suggest that the Sigma 14mm is superior to the original (type I) Canon 14mm. The Sigma focuses slightly closer (18cm, compared to 20cm) and is equipped with 7 aperture blades compared to the Canon's 6. It also takes rear gelatin filters, but does not have an aperture ring.

Tamron 14mm ƒ/2.8 Aspherical IF SP AF ~$1,000
We also haven't tested the Tamron 14mm, which seems to be structurally more similar to the type I Canon 14mm, with 5 aperture blades. It's slightly more expensive than the Sigma, but comes equipped with an aperture ring for compatibility with older bodies.

Conclusion

If you were reading this in order to decide whether an upgrade from the current Canon EF 14mm ƒ/2.8 is worth it, the answer is an unequivocal yes: Canon seems to have improved just about everything that is physically possible to improve upon with this lens. If you were looking at getting into ultra-wide angle photography and were wondering whether to buy one of the above four lenses, the choice is going to be more about how much money you're willing to spend. Thankfully, this choice only becomes truly difficult if you have a full-frame body: APS-C sensor body users are better served with the much more reasonably-priced (and excellent) Canon 10-22mm. We can't definitively say that the Canon models outperform the Sigma or the Tamron, not having yet tested those lenses, but we can say that if you do decide to buy the Canon EF 14mm ƒ/2.8 II USM, you probably won't be disappointed.

Sample Photos

The VFA target should give you a good idea of sharpness in the center and corners, as well as some idea of the extent of barrel or pincushion distortion and chromatic aberration, while the Still Life subject may help in judging contrast and color. We shoot both images using the default JPEG settings and manual white balance of our test bodies, so the images should be quite consistent from lens to lens.

Still Life shot

VFA target

As appropriate, we shoot these with both full-frame and sub-frame bodies, at a range of focal lengths, and at both maximum aperture and ƒ/8. For the ''VFA'' target (the viewfinder accuracy target from Imaging Resource), we also provide sample crops from the center and upper-left corner of each shot, so you can quickly get a sense of relative sharpness, without having to download and inspect the full-res images. To avoid space limitations with the layout of our review pages, indexes to the test shots launch in separate windows.