Sigma have recently released a brand new 14-24mm f/2.8 lens to their Art series of professional grade optics. When I saw this announcement I was a little surprised as their existing 12-24mm f/4 Art is only a couple of years old, and this new lens seems to compete with the existing one. Okay, so the older lens has 2mm more range on the wider end, and the new lens has an extra stop of aperture range, but other than that, they seem to be pretty much the same on paper. The only other thing which caught my eye was the claim that the new 14-24mm boasted “zero distortion”.
I wasn’t sure what this meant, as it could be interpreted in a couple of ways. Firstly it can refer to pin cushion or barrel distortion, which is something many wide angle lenses suffer from to varying degrees. However, this can be eliminated with “Lens Corrections” in Lightroom, so it’s not really a major problem (although admittedly it is better not to have distortion on a lens). The other way a wide angle lens distorts the image is when objects of a known shape (such as people or buildings for example) become stretched when they are placed close to the edges of frame. If you take a photo of a ball and place it in the centre of the frame on a very wide angle lens it looks natural and round. But as you move it away from the centre its shape starts to distort and it gets stretched into an oval. The further from the centre of the image you place such an object, the more it will distort. This is only an issue with wide angle lenses and occurs because wide angle lenses allow you to tilt or turn the camera to a greater degree away from the subject before it disappears from frame. The more a camera’s sensor is twisted away from a subject, the greater the angle of light from that subject strikes the sensor. Think of a torch which projects a round beam. When you point the torch head on to a flat surface, the light will form a perfect circle. But if you now alter the angle of the torch to that it strikes the wall at 45 degrees, the circle gets distorted and becomes an oval.
Above shots at 14mm. When using wide angle lenses, known shapes in the centre of the frame are true, but known shapes at the edge of frame become stretched and distorted.
I love shooting portraiture on wide angle lenses, but when doing so I am very limited in composition options due to the fact that I need to position my subject close to the centre of the frame, otherwise they end up looking distorted and unnatural. If a lens manufacturer could ever correct for this type of distortion it would be a revelation in photography and this is what I had really hoped Sigma meant by “zero distortion”. Sadly, this new 14-24mm still stretches things as they get towards the edges of the frame. This is not a weakness of the lens, but more a law of physics.
So now that’s put to bed, let’s look at the performance of this new super wide zoom.
The thing I am most interested in, is how sharp this new lens is when used at its shortest focal length. There’s not much point in having a super wide angle lens if it is not sharp at its widest focal setting. I’m also interested in how it performs at its widest aperture. The only advantage of the new lens over the existing one is a stop more performance in speed, but if it doesn’t perform well when wide open at f/2.8 then this advantage is lost. I will also be testing it to see how well it handles flare when shooting into the light - something done more often of super wide angles than any other focal range. Flare is an issue which many wide angle lenses suffer from and it seems to have little resemblance to how much a lens costs. Another area I’m interested to explore is the lens’ ability to deal with coma. Coma is where small points of light (such as stars) become more like flying saucers in shape as they are positioned further from the centre of the frame. This is obviously important for astrophotographers, and this new lens’ faster f/2.8 aperture will be of interest in this area. Other important factors are weight and ease of use.
Many lens tests involve charts and graphs and lots of other boring things. I am not interested in testing lenses like this, because I don’t photograph charts and graphs. Here I am simply going to use the lens in a real life situation and examine the results using my own eyes. Daring maybe, but that’s the way I roll.
Just like all the other Sigma Art lenses I have used, this one feels as though it is built like a tank and makes a satisfying clunk when coupled to the camera. Both the focus and zoom rings are firm and solid, with well ridged rubber lines on them which make them easy to use with both bare hands and with gloves. The only switch on the barrel is for AF/MF control. There is no Image Stabilisation on this lens.
The first thing you will notice when handling this or any other of the Art lenses is their weight. They aren’t going to win any awards for being the lightest lenses on the market, but personally I don’t mind this at all. When fixed to the camera they feel well balanced and give you the impression you are holding a serious piece of optical glass.
So let’s start by taking a few photos and see how we get on.
I took the new lens with me on a recent Worldwide Explorers photo holiday I was hosting in Utah. I waited until I found a super wide scene which was all at infinity (trust me, that’s more difficult than it sounds) so that I could examine the sharpness of the lens knowing that everything in the image was exactly on the plane of focus. The camera body was a Canon 5D IV, mounted on a sturdy Gitzo tripod. There was no wind, focusing was done manually for maximum accuracy and consistency and mirror-lock-up was used to reduce camera vibrations. The images here have not had any lens corrections applied and have only had minor adjustments to contrast (all exactly the same settings) applied.
As already mentioned, I am most interested to see how this lens performs at its shortest focal length and widest aperture as this is where this lens will be most useful.
Shot at 14mm, f/2.8
The first thing you will notice is the obvious vignetting in the corners. This is usual for such a wide angle lens when used at its widest aperture. This is totally removed when enabling Lens Corrections so it’s not a major issue. The vignetting becomes noticeably better at f/4, and then improves again at f/5.6 and f/8 but then seems to flatten out all the way through to f/22. The below image is at f/8 and you can clearly see the difference in the luminance of the corners.
To examine the sharpness of the lens, let’s look at an area of the scene with plenty of contrast which is close to the centre of the frame (where the lens will perform best), and also an area closer to the corners (where the lens will perform worst). I’ve circled my chosen areas for reference.
Shot at 14mm, f/8
Here are the two areas at 100% resolution, taken at 3 different aperture settings, all at 14mm.
14mm, f/2.8 centre
14mm, f/2.8 edge
14mm, f/8 centre
14mm, f/8 edge
14mm, f/22 centre
14mm, f/22 edge
When viewing them on my 27” 5K iMac screen, there is a subtle improvement in centre sharpness from f/2.8 to f/8 but not as much as I would have expected. This is good news, as it shows the lens performs almost as well wide open as it does in the middle of its range. There's a much more obvious difference between f/8 and f/22, with the smaller aperture performing far worse. This is in line with what I would expect, as diffraction is ruining the sharpness at f/22.
When comparing the edges, the differences are more pronounced, with f/8 being the clear winner as expected. Notice how much darker the f/2.8 edge is, due to the vignetting mentioned earlier. Once again, this is not a major issue as it can be corrected with the click of a button in Lightroom, although this will result in increased noise in these areas.
To work out the sweet spot of this lens, I also compared shots taken at f/5.6 but neither the centre or the edge seemed to be quite as sharp as f/8 - although I am splitting hairs to find a difference.
Now let’s change the focal length to a middle value where we would typically expect a lens to perform better. Obviously I am having to choose a different area at the edge of frame now, as the previous one is no longer part of the image.
20mm, f/2.8 centre
20mm, f/2.8 edge
20mm, f/8 centre
20mm, f/8 edge
20mm, f/22 centre
20mm, f/22 edge
Once again, the good news is that there is little difference between the centre at f/2.8 and at f/8, with f/8 just about being noticeably sharper. f/22 however is clearly nowhere near as sharp due to diffraction.
When comparing the edges there is once again a noticeable difference in the luminance of the the f/2.8 version which is being affected by vignetting, but in terms of sharpness there isn’t much difference. The f/8 still version wins, but only marginally, with less difference than at 14mm. As expected, f/22 lags well behind.
Using this lens around 18-20mm at f/8 would be where I would expect it to be performing to its maximum potential. As can be seen by both the centre and edge sections of the 20mm, f/8 images, this lens is bitingly sharp and doesn't lose much in the way of definition when opened up on both the focal zoom and the aperture.
Now lets look at the long end of the focal range, but before we do, I have discovered what some may see as a bit of an issue. If, like me, you manually focus for maximum accuracy, when you change the focal length of this lens, the focus ends up drifting ever so slightly and needs to be tuned back in. I make a habit of always checking this anyway, but it’s just something to be aware of if you end up buying this lens. Obviously if you are using autofocus then it will focus each time you take a shot anyway, so this won’t be an issue.
24mm, f/2.8 centre
24mm, f/2.8 edge
24mm, f/8 centre
24mm, f/8 edge
24mm, f/22 centre
24mm, f22 edge
Firstly looking at the centre of the lens. As expected, the f/22 version falls far short in terms of sharpness compared to f/8 and f/2.8. What’s noticeable here is that the f/2.8 version not only has the usual vignetting, but the entire image is about half a stop darker, even in the centre. I even checked I hadn’t messed up the exposure, but the two are identical and were shot within a few seconds of each other. I would say that although the f/2.8 version is darker here, it has the edge when it comes to sharpness.
Looking at the edges, I’ve chosen an area on the opposite side of the frame as the light is now better than it is on the area I was previously using. Once again the clear winner is the f/8 version, with the f/22 and f/2.8 being noticeably less sharp. There’s the usual vignetting on the f/2.8 version but in terms of sharpness, it still edges the f/22 version.
Having ascertained that f/8 is the clear winner with this lens, lets now compare the centre and edges of each shot at f/8, using focal length as the variable. This is a more difficult comparison to make, as the image areas are slightly different.
There’s not a huge amount between these centre three, but the 24mm version is slightly softer than the other two. If I had to choose a winner then I would say it is the 20mm version but there’s really nothing in it.
There’s slightly more difference at the edges, with the 24mm version lagging behind on sharpness. There’s not a great deal between the 14mm and 20mm versions here, but the latter is more obviously sharper than it was on the centre images when I was struggling to choose a winner.
So we know that this lens performs at its weakest at 24mm and f/22, but I would never use any lens at its minimum aperture setting because you are not going to get great results by doing so. Even so, the sharpness and definition at these settings is still impressive, just not as impressive as at other settings. When used around 18-20mm between f/5.6 and f/11 it is amazingly sharp right into the corners of the frame. You don’t buy a 14-24mm lens though unless you plan to use it at 14mm (otherwise you might as well buy a 16-35mm), so the performance at 14mm is critically important. And this is where the lens really does hold up well, as the difference between 14mm shots and those in the sweet spot (18-20mm) are marginal to say the least.
Don’t misunderstand what I am saying when I refer to this lens not performing at its best at 24mm. It performs excellently, just not as good as at wider settings. This says a lot about how good those other settings are.
I am always keen to test how lenses handle flare, as this can be something which ruins a shot if not dealt with during exposure. The best lens I know of for dealing with flare is the Canon 16-35mm f/4 L IS. Although at certain settings in certain situations it is still possible to get flare with it, generally speaking it is not something you need to worry about too much when using it. Flare only occurs when a bright light source is present in front of the lens. Usually this light source is the sun, but if it is not included in your frame, then it is relatively easy to eliminate the problem by casting a shadow over the front element of the lens with your hand. This is good practice even if using a lens which handles flare very well, as it will always be present even if it’s not that noticeable. Shielding the front element from bright light will always improve contrast. Using a lens hood does this to some degree, but with super wide angle lenses, the hoods are of less use in this regard.
At 14mm there is major flare occuring on the extreme left hand edge of the frame.
As you zoom in, the flare positon remains fixed and gets cropped out. This is at 24mm.
When the light source is placed in the centre of the frame, flare is handled much better, regardless of the focal length.
The further from the centre the light source is placed, the worse the flare becomes. This was 14mm.
The next area I am interested in testing this lens is for astrophotography. I already have the king of astrophotography lenses - Sigma’s 14mm f/1.8 Art which is a dream to use. Obviously that is a fixed focal length and has its limitations over the 14-24mm f/2.8, but both lenses should appeal to photographers wanting to capture the night sky and aurora. There certainly wouldn’t be much point in having both though. The main problem when using wide angle lenses for astrophotography is how they render stars towards the edges and corners of the frame. Rather than fine points of light, stars can take on weird shapes which can look more like UFOs. This effect is known as coma. The worst lens I have ever used for astro is the Canon 14mm f/2.8 L II. The coma is so bad that night sky images taken on it are virtually unusable. The Sigma blows the Canon out of the water, not only because it handles coma much better (trust me, it couldn’t be worse), but it also has over a stop more aperture to play with. Add to that the street price is significantly cheaper for the Sigma and you’ll soon realise that if you have a Canon 14mm f/2.8 L II then you’ve got the wrong lens.
Okay, back to the test. This was taken at 14mm at f/2.8, which are the obvious settings to use for astrophotography.
A full frame astro shot with Milky Way above a Utah landscape. The foreground is lit by the setting moon and is clearly out of focus due to the limited depth of field at f/2.8.
This is a 100% crop from the top left corner. As you can see, the stars are creeping due to the 25 second exposure but there is little sign of coma.
A quick comparison with the 14mm f/1.8 Art lens....
A full frame image taken on the Sigma 14mm f/1.8 Art - the king of Astro lenses.
This is a 100% crop from the top left corner. As you can see, the stars are creeping due to the 25 second exposure and coma is noticable but well managed.
To give you an idea of how bad the Canon is for coma....
A full frame image taken with the Canon 14mm f/2.8 on a Canon 5D Mk II.
A 100% crop from the top left corner shows serious coma on the Canon. The brightest stars look more like birds!
The above Canon image seems sharper (and it is) but this is due to a much shorter exposure time than on the Sigma examples above. The purpose is to compare coma, not sharpness.
To compare the true sharpness of both Sigma lenses, I thought I would do an identical shot in daylight with both set to 14mm to see the difference at f/2.8…..
The 14-24mm at 14mm f/2.8
The 14mm prime at f/2.8
The 14-24mm at 100%
The 14mm prime at 100%
As you can hopefully see (and as expected), the prime wins on sharpness, but there's not a huge amount in it.
Other features which may or may not be important to know about this lens is that the lens cap simply pushes over the end of the lens barrel and is held in place with friction. I quite like this style as it is less fiddly than trying to fix a traditional lens cap onto a lens where you need to get your fingers inside the lens hood, especially if wearing gloves. The downside of this type of lens cap is that if you put the lens inside a tight fitting compartment in your bag face down, when you pull the lens out of your bag, the cap often gets left behind. This can actually be a good thing, as you don’t then end up losing it in a pocket somewhere (although it would need to be a big pocket to lose this lens cap). If you place the lens face up in your bag, it is not easy to pull the lens out as the only thing you can get hold of is the cap itself, and this then comes off as you lift it out of the bag, leaving the lens behind. Overall, I prefer this type over the traditional lens cap, and they double up as a breakfast bowl for your cornflakes when camping.
There is no screw thread on the front of this lens, as the lens hood is fixed and the front element is bulbous. This makes it impractical to fit filters over the front, unless you opt for a custom made holder and double glazing sized filters. There is no rear gel holder as standard, but I believe Sigma can fit this on request, which may be useful for some users who like to use heavy ND filters.
The Sigma 14-24mm f/2.8 Art is without doubt a very fine lens capable of amazing definition when used correctly. Personally I would only use this lens between 14-22mm and never less than f/11 as it’s just not necessary. f/11 gives ample depth of field at such short focal lengths, and by going any smaller with the aperture you end up sacrificing important areas in order to try to get less important areas sharper - which makes no sense at all. For anything more than 22mm I would change lens to a 16-35mm for example. This is not a criticism of the lens, as it is perfectly usable at all focal lengths.
If you are looking for a super wide zoom with a fast aperture then I don’t know of an alternative option which would perform better. The Nikon 14-24mm f/2.8 is a great lens, and although I haven’t had the chance to compare the two, I would be very confident that the Sigma would beat it in a head to head. If you’re a Canon user then this lens’ rivals would be the Canon 11-24mm f/4 L or one of the 16-35mm L series lenses (f/2.8 or f/4 IS). The Sigma has a clear advantage over the Canon 11-24mm in terms of a stop more aperture and better sharpness. I wouldn’t want to say if it had a sharpness advantage over the two 16-35mm options, but it has the obvious 2mm wider focal range which is too attractive to ignore. 2mm may not sound like much, and at longer focal lengths (more than 35mm) it would be insignificant, but the difference in percentage between 14mm and 16mm is 14%, which means you get 14% more of that amazing view in your frame. The other option you have is the Sigma 12-24mm f/4 Art, and this is where I am still a little confused as to why Sigma have released two lenses which seem to compete with one another so closely. I already own the 12-24mm and I can’t see any reason to have both (especially as I also have the 14mm f/1.8 Art for astro) so I need to decide for myself which one is the best all round lens. If you are looking for a pro class high performing wide angle zoom with a fast aperture then this lens is going to take some beating. After all, who wouldn't want to have a lens in their bag capable of taking shots like this?
Sigma 14-24mm f/2.8 Art, at 14mm, f/8
If you use a crop sensor body, then this lens may be a very sensible option for you. As we have seen in the example images to compare sharpness, this lens (like all lenses) performs least well at the edges and corners of the frame. When used on a crop sensor however, you aren’t using these weaker areas, so your entire frame from corner to corner will be bitingly sharp. On a x1.5 sensor this lens will give an equivalent range of 21-36mm which is a very useful area for landscape photography. Another advantage is that if in the future you decide to upgrade to a full frame sensor, then you already have a great lens compatible with your new camera.
Weight (when attached to the camera)
Weight (when in my bag and travelling uphill)
Prone to flare issues in certain situations
Needs refocusing when focal length is changed.
Street price in the UK is around £1,400. Comes with a soft case, but I’ve yet to see anyone use such a thing for a wide angle lens.