Mother of all Ultrawide shootouts - Nikon vs Canon vs Sigma

March 10, 2013  •  18 Comments

 

Ultrawide lens shootout. Nikon vs Canon vs Sigma

 

The ultrawide. Known as the most difficult lens to design and manufacture. Even more so the ultrawide zoom. The playing field for full frame ultrawide lenses is unsurprisingly narrow for the same reasons. Both Nikon and Canon users have been spoilt for choice in this segment, thanks to some excellent primes and zooms form the original manufactures. For those on a budget, there are a few third party choices too. All of us have, at one point or the other pondered over the "Which ultrawide to buy" question. As a Nikonian, I naturally ended up asking a variation of that question that tens of thousands have asked before. "The 14-24 or the 16-35"?

 

You see, when it cones to Nikonland, there are two contenders to the throne. Each has its own pluses and minuses and are in fact, two very different lenses. Yet, there's a certain degree of overlap between them and I don't just mean focal lengths. Like many before me, I spent many hours searching on websites, forums and blogs to get a definitive answer on which one's "The one". Most of them unfortunately, were filled with very subjective opinions from people who may or may not have even seen these lenses in first person. Frustrated with the lack of some definitive information, I decide to do what I should have done in the first place. Review them both myself and reach my own conclusions. I managed to borrow a mint copy of the 14-24 thanks to a generous friend and decided to rent the 16-35.

 

I mentioned this to my buddy Srikeerthi, who proposed the interesting idea of making this a Nikon vs Canon shootout. He owned the highly regarded Canon 16-35 f/2.8 L II, which made the task easier. As with most things we do, this too expanded its scope very quickly. We decided to add two more lenses that him and I owned into the mix. The Sigma 12-24 and the Sigma 20mm respectively. Our sentiments towards both lenses were similar. We both bought them at a time when wallets were tighter, knowing fully well that these aren't the best options one could get. We've used them in a variety of situations, yet never really fell in love with them. We threw them into the mix just to see how much the Canon and Nikon lenses were better than them. i.e. to see how much more performance the extra dough buys us.

 

Just to make things interesting, we decide to rent the Canon 8-15 f/4 Fisheye too. It wasn't a serious inclusion, but more of an "It's available, so why not" thing. That said, there are guys who look into using fisheyes for ultrawide shots and manually undistorting the images in post to recreate the image as if a rectilinear lens shot it. In case you do not know the difference between a rectilinear ultrawide and a fisheye, the former keeps (Or tries to keep) vertical lines straight while the latter bends them like pretzels. For this reason, fisheyes are easier to make and generally cheaper.

 

Anyway, here's a formal introduction of all the competitors:

 

Nikon 14-24 f/2.8 ED

Nikon 14-24 f/2.8 ED

 

When Nikon introduced this lens, it created some major noise in the industry. Here was a lens in a  zoom range that was never attempted before. Reviewers were singing high praises on how little distortion it has and how it is as sharp as primes in this focal range. There was nothing else quite like it. A very distinctive looking lens thanks to the bulbous front element, this lens unfortunately also has a major drawback. No possibility to add front filters (Easily). While companies like Lee and Lucroit have come out with kits to add filters to this lens, it is safe to say that they are nowhere as convenient to use as a normal screw in/ square filter kit on lenses with filter threads. Also, forget using polarizers and multi-stop NDs on this lens (Even if they are available) because of the serious color shifts that may occur. This makes it quite limited for a landscape shooter. But would the superior image quality make up for this? We'll see.

 

Handling notes:

 

This lens is built like a tank. No question. It is built to Nikon's highest standards and will probably outlast you. On the camera though, it's a bi awkward. It's VERY front heavy, but not very long, which makes for an odd weight distribution on even a D800. I don't mind walking around with a 70-200 attached to my D800 which in turn is hung around my neck because of the even weight distribution, but this combo here felt odd. Maybe I needed more time to get used to it.

 

Another point to note, the lens cap comes off rather easily, so pay attention!

 

Nikon 16-35 f/4 VR

Nikon 16-35 f/4 VR

 

When Nikon announced this lens, a few eyebrows were raised. Many were expecting Nikon to emulate Canon and do an f/2.8 lens. Instead, they made it a stop slower and added VR. Now, a lens like this would mostly reside in the f8-f/11 area, so the loss of one stop isn't as bad as some people make it seem to be. One consequence of this is that the lens is much lighter compared to its Canon counterpart (And certainly its big brother). But the good thing is, it comes with a standard 77mm filter thread, so your existing filters will work perfectly!

 

Reviewers of this lens had some criticism about its image quality. Namely, significant vignetting (Non-factor), and distortion (Fixable, but you'll lose some image area)at the wide end. Would that be the Achilles' heel of this lens? let's find out?

 

Handling notes:

 

The outer barrel of this lens is made from polycarbonate (Much like may of Nikon's newer lenses), but it feels almost as solid as the 14-24. The only giveaway is the slightly hollow noise that arises when you tap the barrel. Although this lens takes up about the same space as the 14-24 in the bag, it feels much more balanced on the camera.

 

 

Canon 16-35 f/2.8 L II

 

Canon 16-36 f/2.8 L II

An update to a well loved original, this lens is highly regarded in the Canon community. I have extensively used its predecessor in the past and found that it was a bit susceptible to flares. Thankfully, this new version controls flares much better. At 82mm, the filter thread on this lens is one size larger than what most of us have bought filters for, but it's still better than having none. Besides, if you're a square filter system user, it's a simple matter of buying another adapter ring (I am and I did. This is the reason why I always recommend square filters to everyone).

 

Handling notes:

 

Being an L lens, this is built incredibly well. No surprises there. It's almost all metal although I suspect the filter thread might be plastic. That's not a bad thing as a plastic thread won't expand/ contract when the temperature fluctuates. Much like it's Nikon counterpart, this lens felt quite balanced on a full frame body, even though it's a bit heavier than the former.

 

 

Sigma 12-24 f/4.5-5.6 (Canon mount)

 

Sigma 12-24 f/4.5-5.6

 

This is the widest rectilinear lens available for fullframe cameras, bar none. However, it's far from being the best. Furthermore, it doesn't have a filter thread either. That said, it is certainly very wallet friendly and does come in many mount choices.

 

Handling notes:

 

Being a mostly plastic affair (With that awful finish that last gen Sigmas were (in)famous for, this lens sits comfortably on a fullframe body. I wouldn't put it through the same tortures as the Canon and Nikon lenses though.

 

 

Sigma 20 f/1.8 (Nikon mount)

 

Sigma 20 f/1.8

 

 

Another unique lens from Sigma. This is the fastest prime you can buy in this focal range. I've personally used and abused mine for years now, first on DX and now on FX. I've never been truly happy with its corner performance, but the centers have always been pretty sharp. But is it sharper than the Canon and Nikon lenses?

 

Handling notes:

 

Again, an all plastic affair with that awful finish. Mine has been peeling in many places. AF is slow on Nikon bodies as it has no built in AF motor and uses the screw drive (Not that it matters much for landscaping). Weighs nothing, feels like nothing on an FX body.

 

Canon 8-15 f/4 L Fisheye

 

Canon 8-15 f/4 L

 

 

Although this is technically an EF lens (Which means that it can be used on fullframe cameras), I personally think that this lens is more at home on APS-C cameras. As you can see below, the image circle that this lens produces is smaller than the fullframe sensor until one zooms to 14mm. On an APS-C sensor, 8mm covers all of the imaging area (And give you 12.8mm equivalent FoV, which is wider than the 14mm you get on fullframe). Plus you have the added benefit of additional pixel density.

 

Handling notes:

 

This is again, an L lens which means it's built like a tank. It's pretty petite too, which means it's pretty balanced on the camera. Make sure to always keep the hood on though, as the front element protrudes a lot.

 

 

Testing methodology:

 

As mentioned earlier, we wanted this test to be as objective as possible. We also wanted it to be very relevant to real world scenarios. If you had come here expecting shots of brick walls and test charts, I will have to disappoint you. You may look elsewhere for those.

 

Instead, we chose to shoot a beach scene as our test frame. This is a common location for a lot of landscapers, so hopefully, the test images will resonate well with you folks. We scouted until we found a rather nice frame and set out tripods down. The lenses were tested on a Nikon D800 and a Canon 5D Mark II, as per the mounts they came in.  Why not a 5D Mark III? Or for that matter, why not a Tokina 16-28 or a Carl Zeiss 15 f/2.8 or **Insert your favorite camera/ lens here**? Simply because we didn't have access to them. We are not endorsed by any manufacturer/ dealer and have to buy/ borrow/ rent the gear we review. This was the best we could muster for this round. If you have some equipment that you'd like to see a review of/ compared against something else, do let me know and I'd be glad to do it.

 

At the location, we set up the frame for the widest rectilinear lens, the Sigma 12-24. Both cameras were set up in manual exposure mode and settings were carried over throughout the test. The D800 was shot at ISO 100 while the 5D Mark II at ISO 160, as these are the native ISOs for both cameras respectively. This means that they would have the lowest noise and most dynamic range at these ISOs. All shots were at f/11. The canon metered at 1/250s and the Nikon at 1/125s. For the lenses with a filter thread, we used a Lee foundation kit with a Hitech 3 stop hard grad. Why? because again, this is a real world test and we wanted to test these lenses the way we'd use them in the field. All images were shot with mirror lockup and timer to minimize shake. All lenses were focused at the middle bunch of rocks, at each focal length they were tested in.

 

All images were brought into Lightroom for RAW conversion. Lens distortion correction was enabled to remove some variables (Chromatic aberration wasn't corrected as we are not testing that in this test). We applied "Camera neutral" color profile to all files. No attempt was made to color correct the images, therefore they retain their signature looks (Green/yellow for Nikon and Red/magenta for Canon files). No sharpening was applied during editing or exporting. All files were exported at 12 MP resolution as JPEG, 100% quality.

 

Now let me rant on for a bit. I've seen far too many reviewers take image files from two different cameras with two different resolution and compare them at 100%. This is stupid. You cannot in any way make an objective comparison with this technique. Think of the final output format for your images. It may either be displayed online or printed. No matter what resolution camera it was shot with, you export to certain standard dimensions. And this is how the images should be judged. At a fixed size. If you disagree with this methodology, you're free to do a test of your own with your  own testing methodology. But here's a spoiler: You'd be wrong. Anyway, 12MP gives us a good sized file (It's big enough to be printed at 12"x8" at 300 DPI and still maintains a good amount of detail).

 

All images were segregated based on focal lengths. Images up to 14mm, images at 16mm, at 20mm, 24mm and 35mm. Each group was overlaid and 700px 700px crops were taken from different areas. Namely, lower left, lower right, center, foliage (To show detail in the leaves) and horizon. The crops were arranged next to each other and will be presented below.

 

I decided to award points to the lenses based on their performances in each of these categories. Remember, these are points, not rankings, so there might even be some negative numbers! The highest any lens can get in each category is 5 points.

 

So without further ado...

 

Up to 14mm:

Canon 8-15 f/4L at 8mm

 

 

Canon 8-15 f/4L at 10mm

 

 

Canon 8-15 f/4L at 12mm

 

 

Canon 8-15 f/4L at 14mm

 

 

Sigma 12-24 at 12mm

 

 

Nikon 14-24 f/2.8 at 14mm

 

 

 

 

Right off the bat, we can see some differences. Let's look at the 100% crops (Only looking at 14mm crops for the Canon 8-15 as the wider focal lengths do not cover the full imaging area):

 

Lower left:

 

Wow! I had always known that the Sigma had terrible corners, but I didn't expect the results to be this bad. The other two lenses simply trample over it. The Canon, by virtue of being a fisheye is very strong in the corners (And weaker in the center, as we will see later).  The Nikon's corners are exceptionally good for a rectilinear lens at this focal length, but I will give the narrow win to the Canon here.

 

Canon 8-15: 5 points

Nikon 14-24: 4 points

Sigma 12-24: 1 point

 

Lower right:

 

 

Same story here, although the distance between the Nikon and the Canon is much closer now. I am inclined to give them both top scores. The sigma lags behind again.

 

Canon 8-15: 5 points

Nikon 14-24: 5 points

Sigma 12-24: 1 point

 

Center crop:

 

 

I looked at this long and hard (Especially difficult as the higher dynamic range of the D800 gives its files lesser apparent contrast compared to the Canon files).Ignoring the faux-detail that the deeper shadows in the Canon files bring, I think that the 14-24 resolves a wee bit more detail, but not by much. The Sigma continues to disappoint, even in the center.

 

Canon 8-15: 4.5 points

Nikon 14-24: 5 points

Sigma 12-24: 1 point

 

Foliage crop:

 

 

 

This crop was done to see how much fine detail the lenses can resolve when faced with a non uniform, fine pattern such as foliage. There's no contest here. The Nikon 14-24 is the king of this jungle. The sheer amount of detail it resolves is astounding. The Canon has significantly more color fringing and doesn't quite manage as much detail. The sigma looks as if someone draped muslin over its front element.

 

 

Canon 8-15: 3 points

Nikon 14-24: 5 points

Sigma 12-24: 1 point

 

Horizon crop:

 

 

 

Same story as before with the Nikon emerging the winner with the Canon slightly behind and the Sigma way, way behind.

Canon 8-15: 3 points

Nikon 14-24: 5 points

Sigma 12-24: 1 point

 

Overall score up to 14mm:

 

Canon 8-15: 20.5 points

Nikon 14-24: 24 points

Sigma 12-24: 5 points

 

The winner at this focal range is the Nikon 14-24 f/2.8. The Canon fisheye put up a surprisingly good fight and if you're willing to do some post production, can replace a rectilinear lens in your bag. But always remember, it is strongest in the corners, most of which you will crop out after correcting distortion.

 

 

That said, let's not ignore the elephant in the room. The Nikon 14-24 has a flare problem. You can see it prominently in the above sample. Here's a crop in case you didn't notice it.

 

 

Even with a hood and Nano coating, the huge front element of this lens makes it susceptible to flares. Make sure you cup it with a hand while shooting near strong light sources.

 

16mm:

This is where things start to get interesting. We have more players in this round, with the 16-35s from both manufacturers entering the fray. Let's see how they stack up against the 14-24, which is relishing its victory in the last round:

 

Canon 8-15 F/4 L (At 15mm)

 

 

Canon 16-35 f/2.8 L

 

 

Nikon 16-35 f/4 VR

 

 

Nikon 14-24 f/2.8

 

 

Sigma 12-24

 

 

As mentioned earlier, a Grad ND filter was used where possible. Please take note of the flare that occurs in the same exact spot for both 16-35s. This is due to the filter (One can't use the hood while using a square filter. I will be getting a Lee hood in the future to compensate for this).

 

Lower left:

 

 

 

Competition is much closer this time around. The Nikon 14-24 is shockingly bested by both 16-35 lenses. ignoring the false sharpness created by the deeper shadows in the canon files again, I feel that the Nikon 16-35 has the slightest of edges over its Canon counterpart. The Sigma has improved its performance over the previous round, but is still behind the rest. The Canon fisheye (At 15mm) once again has good detail in the corners, but is considerably less sharp at its longest end compared to the performance in the previous round.

 

Canon 8-15: 3 points

Nikon 14-24: 4 points

Sigma 12-24: 3.5 points

Canon 16-35: 4.5 points

Nikon 16-35: 5 points

 

Lower right:

 

 

The Nikon 16-35 is still the best of the lot and the 14-24 has jumped ahead of the Canon 16-35 by a little bit. The rest remain unchanged.

 

Canon 8-15: 3 points

Nikon 14-24: 4.5 points

Sigma 12-24: 3.5 points

Canon 16-35: 4 points

Nikon 16-35: 5 points

 

Center crop:

 

 

 

Same results as the left crop. Very close fight between the Nikon and Canon 16-35 lenses with the former slightly inching ahead. The Sigma has that hazy look about it that never seems to go away.

 

Canon 8-15: 3 points

Nikon 14-24: 4 points

Sigma 12-24: 2 points

Canon 16-35: 4.5 points

Nikon 16-35: 5 points

 

Foliage crop:

 

 

Once again, the moment you move away from the corners, the Nikon 14-24 flexes its muscles. The Canon fisheye has delivered a surprisingly great performance and comes in at a very close second. Third is the Nikon 16-35. The Sigma surprisingly, is a tad better than the Canon 16-35 in this test.

Canon 8-15: 4.5 points

Nikon 14-24: 5 points

Sigma 12-24: 3.5 points

Canon 16-35: 3 points

Nikon 16-35: 4 points

 

Horizon crop:

 

 

 

I will have to give it to the Canon fisheye for this one. It's tack sharp. It's followed by the Nikon 14-24, the Nikon 16-35 and the Sigma 12-24. The Canon 16-35 shockingly comes in last.

 

Canon 8-15: 5 points

Nikon 14-24: 4.5 points

Sigma 12-24: 3.5 points

Canon 16-35: 3 points

Nikon 16-35: 4 points

 

Overall score 16mm:

 

Canon 8-15: 18.5 points

Nikon 14-24: 23 points

Sigma 12-24: 16.5 points

Canon 16-35: 19 points

Nikon 16-35: 23 points

 

The Nikon 14-24 and the Nikon 16-35 are the joint winners of this round. I was frankly surprised by the performance of the Canon 16-35 in the foliage and horizon tests. The Sigma clearly is in the sweet spot of its zoom range, but doesn't do enough to get out of the last place. The Canon fisheye delivered some great performance in this round and with the considerably less distortion it has at 15mm, could be an alternative to a rectilinear lens in a pinch.

Do note though, after distortion correction, the image area that you get from the Nikon 16-35 is slightly lesser than from the 14-24 or the Canon 16-35.

 

20mm:

 

The Canon fisheye bows out in this round and in comes a new challenger, the Sigma 20mm prime. As the only prime in this test, how will it fare against some of its more illustrious competitors?

 

Sigma 20 f/1.8

 

 

Nikon 14-24

 

 

Sigma 12-24

 

Canon 16-35

 

 

Nikon 16-35

 

The two Sigmas couldn't be any more different in this test. The 20 prime flamboyantly shows off what I like about this lens. Superb contrast. Let's look at the crops to see if that translates to a victory.

 

Lower left:

 

 

 

The Canon 16-35 comes off as the winner here. The Nikon 16-35 inches off the 14-24 by a bit. The Sigma 12-24 has much less contrast than the 20 prime, but its edges hold up better.

 

Sigma 20: 2 points

Nikon 14-24: 4 points

Sigma 12-24: 3 points

Canon 16-35: 5 points

Nikon 16-35: 4.5 points

 

Lower right:

 

 

 

The Nikon 16-35 has redeemed itself here and claims top spot, followed by the 14-24. The Canon 16-35 is close behind. The Sigma 12-24 after than at the 20 prime, although showing a better result than on the left, is still last.

 

Sigma 20: 2 points

Nikon 14-24: 4.5 points

Sigma 12-24: 3 points

Canon 16-35: 4 points

Nikon 16-35: 5 points

 

Center crop:

 

 

 

Color me astounded, but the Sigma 20 prime takes the top spot here! Having been using this lens for a while, I knew that it has great center sharpness, but I didn't think it would be class leading. The Nikon 16-35 comes in next and the 14-24 and Canon 16-35 are tied behind that. The 12-24 comes in last.

 

Sigma 20: 5 points

Nikon 14-24: 3.5 points

Sigma 12-24: 2 points

Canon 16-35: 3.5 points

Nikon 16-35: 4 points

 

Foliage crop:

 

 

I'm having a hard time trying to see any detail differences between the Nikon 16-35 and the Nikon 14-24. I have to name them joint winners here. The Sigma 20 prime is a whisker behind. The Cano 16-35 behind that and the Sigma 12-24 behind that.

 

Sigma 20: 4.5 points

Nikon 14-24: 5 points

Sigma 12-24: 2 points

Canon 16-35: 3 points

Nikon 16-35: 5 points

 

Horizon crop:

 

 

Again, I think the Sigma 20 prime might just take the top spot here, closely followed by the Nikon 16-35 and then the Nikon 14-24. The Canon 16-35 is behind that and last is the Sigma 12-24.

 

Sigma 20: 5 points

Nikon 14-24: 4 points

Sigma 12-24: 2 points

Canon 16-35: 3 points

Nikon 16-35: 4.5 points

 

Overall score 20mm:

 

Sigma 20: 18.5 points

Nikon 14-24: 22 points

Sigma 12-24: 12 points

Canon 16-35: 18.5 points

Nikon 16-35: 23 points

 

The Nikon 16-35 wins this round, but the real star is the Sigma 20 prime. It put on a surprisingly good performance, only to be let down by the poor corner performance. I honestly feel that if Sigma could re-design this lens with stronger corners (And maybe wider at 18mm), they have a world beater in their hands.

 

24mm:

 

The Sigma 20 prime bows out on virtue of being ummm.. a prime, so the remaining 4 duke it out this round. The Nikon 14024 and the Sigma 12-24 are at their longest end, traditionally not a strong point of ultrawide zooms. How will they fare?

 

Nikon 14-24

 

 

Sigma 12-24

 

 

Canon 16-35

 

 

Nikon 16-35

 

 

...and here are the crops:

 

 

There one clear winner here and that's the Nikon 16-35. This is followed by the Canon 16-35. The 14-24 and Sigma 12-24 follow in that order. It's quite clear that the two lenses at the long end of their zoom range aren't gonna put their best foot forward in this round.

 

Nikon 14-24: 3 points

Sigma 12-24: 2 points

Canon 16-35: 4 points

Nikon 16-35: 5 points

 

Lower right:

 

 

The two 16-35s trade places here with the Canon coming up on top of the Nikon. The remaining two lenses retain their places from the last crop.

 

Nikon 14-24: 3 points

Sigma 12-24: 2 points

Canon 16-35: 5 points

Nikon 16-35: 4 points

 

Center crop:

 

 

The two 16-35s look tied here in terms of sharpness with the other two following in that order.

 

Nikon 14-24: 3 points

Sigma 12-24: 2 points

Canon 16-35: 5 points

Nikon 16-35: 5 points

 

Foliage crop:

 

 

The Nikon 16-35 wins, followed by the 14-24 and the Canon 16-35. The 12-24 comes in last.

 

Nikon 14-24: 4 points

Sigma 12-24: 2 points

Canon 16-35: 3 points

Nikon 16-35: 5 points

 

Horizon crop:

 

 

Once again, the Nikon 16-35 wins, narrowly beating out the Canon 16-35. The 14-24 follows and then the 12-24.

 

Nikon 14-24: 4 points

Sigma 12-24: 2 points

Canon 16-35: 4.5 points

Nikon 16-35: 5 points

 

Overall score 24mm:

 

Nikon 14-24: 17 points

Sigma 12-24: 10 points

Canon 16-35: 21.5 points

Nikon 16-35: 24 points

 

Another strong performance by the Nikon 16-35, which takes the win here. The Canon 16-35 really stepped up its performance, but the disadvantages of the lower resolution sensor that it was on shows its effects, even after the downsampling process.

 

 

35mm:

 

Only two remain to fight this final battle. The two 16-35s. It's truly Nikon vs Canon now. Let's see who wins this one!

 

Canon 16-35

 

 

Nikon 16-35

 

 

 

Lower left:

 

 

I can hardly tell the difference between the two here.

 

Canon 16-35: 5 points

Nikon 16-35: 5 points

 

Lower right:

 

 

The Canon is slightly better in the extreme corner.

 

Canon 16-35: 5 points

Nikon 16-35: 4 points

 

Center crop:

 

 

Neck to neck, but I think the Canon is very slightly better.

 

Canon 16-35: 5 points

Nikon 16-35: 4.5 points

 

Foliage crop:

 

 

Slightly more fine detail from the Nikon. Very slight.

 

Canon 16-35: 4.5 points

Nikon 16-35: 5 points

 

Horizon crop:

 

 

Definitely more detail in the Nikon file here.

 

Canon 16-35: 4 points

Nikon 16-35: 5 points

 

Overall score 35mm:

 

Canon 16-35: 23.5 points

Nikon 16-35: 23.5 points

 

The two behemoths are tied here and that's not surprising. The differences between them at this focal length is microscopic at best.

 

 

Conclusion:

 

So, we come to the end of this highly comprehensive (And exhausting!) test. I've successfully managed to answer my original question (14-24 or the 16-35), but before we get to that, I think I should talk about the test in general and about the other lenses involved.

 

 

 

 

This test has taught me that there's no replacement to resolution. We tried to equalize the playing field by downsampling the files to 12MP, but it was clear that where there was fine detail involved (Foliage, horizon), the D800 simply "sees" more detail than the 5D Mark II. And this reflects in the final files. Canon's lenses are certainly up to the task, but they need better sensors to compete. Not just for resolution, but also in terms of dynamic range (Look at the clumped up shadows in the Canon samples).  That said, here are some final thoughts on the different lenses:

 

Sigma 12-24 f/4.5-5.6 (Canon mount)

 

Somebody has to come last in every comparo and that dubious honor falls upon the Sigma 12-24. It consistently came last in almost all tests and was visibly worse than the rest. Face it, there are only two reasons why you'd want this lens:

 

1) You want to go the absolute widest on a fullframe body or

2) You're broke.

 

If it's reason 1, I'd say forget it. It's terrible at 12mm. I wouldn't even call its performance at this focal length "Usable". It's decent at 16-20mm, but at that focal range, you have better options.

 

If you're broke, I'd say save up a bit more and get a used Canon 17-40 or Nikon 17-35. Both lenses are not as good as the Canons and Nikons tested here, but will perform better than the Sigma.

 

Canon 8-15 f/4 L Fisheye

 

This lens was included in the test for a novelty factor, but it did surprise us with some strong performances. That said, if I were a Canon user, I wouldn't buy this lens. I'd rent it each time I need it. It's far too limited to be a regular part of your kit. It's like Chachi from Happy Days. Fun in small doses, easily gets on the nerves once made part of the regular cast.

 

Oh, and definitely use it with a crop sensor camera.

 

Canon 16-35 f/2.8 L II

 

This lens has some quality optics in it, no doubt. The slightly disappointing performance at the wider end notwithstanding, this lens is a must have for Canon landscapers. As mentioned before, it's let down by inferior sensor technology. Once Canon's next gen, high res cameras are launched, this lens would fulfill its potential (And I suspect might even win such a comparo).

 

Face to face with the Nikon 16-35, I'd say this lens has less distortion, better consistency from center to corner and slightly better at the long end. The Nikon is overall a bit sharper and better at the wide end.

 

Sigma 20 f/1.8 (Nikon mount)

 

Before this test, I was ready to sell this lens off. Now, I am not so sure. As mentioned earlier, if Sigma launches a V2 of this lens with better AF, build and corner performance, I just might get one. I will most definitely get one if they do an 18 f/1.8. Judging by how great their recent lenses are, I am hopeful that this might happen.

 

Nikon 14-24 f/2.8 ED

 

This lens, in internet folklore is untouchable. It's part of the famed Holy Trinity, therefore one must not speak ill of it. But the truth is, it IS bested by a couple of others in certain areas. While it's still no slouch (Some of the differences seen earlier are darn near invisible in print), the fact that it has no filter thread breaks the deal for me in the end. However, if you're into architecture/ low light event coverage, this is THE Nikon ultrawide to get. When we applied distortion compensation in LR, I was astonished to see how little compensation was needed for this lens across all focal lengths. Still an incredible lens and I'm sure than whenever its successor is announced, it will push the envelope further.

 

...It's just not the lens for me.

 

Nikon 16-35 f/4 VR

 

And yes, here's the overall winner of this comparison. It's not the widest, not the best built or the fastest. But what it does is deliver consistent performance across its focal range. This is the lens that a Nikon landscaper can proudly own. Paired with a camera such as the D800, it really kicks some butt. And most importantly for me, it takes my filters.

 

Here are some shots I took with this lens on the D800 that same evening.

 

 

Questions? Feedback? Do let me know in the comments.

P.S. Depending on the resolution of your monitor, you may see a downscaled version of the 100% crops. Click here to view and download all the images seen in this test.

 

- Sandeep Murali -

 

Support this site by buying your camera gear from the below link:


Comments

18.Ray Avitable(non-registered)
Thank you, Thank you, Thank you! Finally a comprehensive review of all the wide angle lenses I am considering (Nikon and Sigma). I was leaning Sigma for the 12mm but it sounds like it won't capture the image quality I want. It’s the 16-35 f4 for me. I really like the ability to use filters too.
17.jtonvan(non-registered)
Thanks for the review. its a lot of work. It would be nice if you could add some defished to rectilinear samples of the canon 8 16 fisheye at 15mm to match 12mm 14mm fov. It could be very informative.
Thanks again.
16.Duncan(non-registered)
Thanks for the review but It is not clear at which aperture settings they were being tested ? all at their widest ?
with would seem unfair in a way… I would have liked to see the 14-24 nikon at F4 vs it's younger VR brother…..
15.John Leslie(non-registered)
I'd disagree about using the Canon 8-15mm on a crop body, it's better on full-frame as then you get a full 180 circular fisheye plus the full-frame (i.e. 180 degree diagonal) fisheye too. Plus it was half the price of my 14mm f2.8 mk II (UKP 895 vs. UKP 1750). Here's something I made with it at 8mm on full-frame:
http://www.viewat.org/?i=en&id_aut=7366&id_pn=22647&sec=pn
(BTW look up and down too.)

If you want to de-fish it I think the Fisheye-Hemi Photoshop plugin is the best solution, although at the 15mm end CS6's Adaptive Wide Angle filter isn't that far behind.

I also have one other comment - I wouldn't agree that the Nikon 14-24 has low distortion, it's about 4% at the wide end vs. 1.7% for the Canon 14mm f2.8. I think 4% is pretty annoying and would need fixing in post if you had vertical lines (e.g. buildings). I did toy with getting the 14-24 to use on my 5DmkII (with the 16-9 adapter: http://www.16-9.net/nikon_g/ ) when I considering buying the 14, but the lower distortion and the 14 being much better with flare (and much smaller/lighter) won out over the extra sharpness. I do know people who use them on Canon bodies though, and can see why. An example: http://www.davidclapp.co.uk/blog/view/the-nikon-14-24-on-a-canon-1ds3-a-landscapers-report (although he is using an older version of the adapter).
14.Strawe(non-registered)
Great write-up and fantastic photos too! As a side note, I can confirm that Canon's 16-35 has metal threads.
No comments posted.
Loading...

Full disclosure:
This site is affiliated with Adorama Inc. However, I do not get any free equipment from them, nor am I required to get prior approvals from them for my posts. If you shop through the Adorama banners on this site, you will not pay anything more than you would otherwise, but I would be receiving a commission from them.


Subscribe
RSS
Archive
January February March April May June July August September October November December (1)
January (2) February (2) March (1) April May June July (1) August September October November December (1)
January February March April May June July August September October November December