Sigma 14mm f/1.8 DG HSM Art Review

Sigma 14mm f/1.8 Art vs Nikon 14-24mm f/2.8G

If you are wondering how the Sigma 14mm f/1.8 Art compares to the legendary Nikon 14-24mm f/2.8G, take a look at the sharpness results from Imatest:

Based on the above results, it is very clear that the Sigma 14mm f/1.8 Art is a much sharper lens. Even wide open, it beats the Nikon 14-24mm f/2.8G at f/2.8 in the center. And when stopped down to f/2.8, there is simply no comparison, which is kind of what you get when comparing a modern prime to an aging zoom lens. The Nikon 14-24mm f/2.8G is a superb lens when stopped down to f/4, but not necessarily when shooting extreme details wide open – that’s where the Sigma 14mm f/1.8 Art excels.

What about astrophotography? The Nikon 14-24mm f/2.8G has always been regarded as one of the best lenses for astrophotography needs, thanks to its zoom versatility and very little coma at f/2.8. The Sigma 14mm f/1.8 Art is sharper than its Nikon counterpart as seen above, but then it has noticeable coma at f/1.8 and needs to be stopped down to f/2.5 – f/2.8 to get similar levels of coma in the extreme corners. The ability to change focal length on the 14-24mm f/2.8G can be quite useful, but if you find yourself shooting mostly on the wide side anyway, then the Sigma 14mm f/1.8 Art would definitely be the way to go. On the other hand, if you like the versatility of the zoom, then the 14-24mm f/2.8G is still a great astrophotography lens to use. Personally, I find myself wanting to go as wide as possible when shooting astrophotography, so my personal choice would be the Sigma 14mm f/1.8 Art. A little bit of coma does not bother me, as long as it does not look extreme (which it does not on the Sigma), but I can imagine the possibilities of the Sigma 14mm f/1.8 Art combined with something like the Sky-Watcher Star Adventurer. Long exposures with no star trails, stacked together to reduce noise, could produce pretty phenomenal results…

Here is an image that I captured with the Sigma 14mm f/1.8 Art, shot at f/1.8 with an exposure length of 20 seconds (single image, post-processed in Photoshop):

Thanks to the large aperture of f/1.8, I was able to shoot the Nikon D850 at ISO 800 and still get a great shot. While coma is clearly visible in the corners, it is not too distracting for my taste. Ideally, stopping down to f/2.5 and stacking images, or using an astro tracker would be the way to go with this lens.

Sigma 14mm f/1.8 Art vs Samyang 14mm f/2.8

I tested two copies of the Samyang 14mm f/2.8. The first sample was severely decentered and showed poor overall performance, so I had to get a hold of another copy. The second copy was certainly better optically and did not have any serious decentering issues, but it was still not as good as I was hoping it would be. No matter what I tried, I could not get both samples to produce sharp results at f/2.8, so I suspect that it is just a weak optical design that results in soft wide open performance. Let’s take a look at how the two lenses compare:

As you can see, the Samyang 14mm f/2.8 is far from being a sharp lens at f/2.8. Once stopped down to f/4, it gets much better optically in the center, but still far worse than what the Sigma 14mm f/1.8 Art is capable of. Personally, I would not use the Samyang 14mm f/2.8 for astrophotography, but if you need a pretty decent ultra wide-angle lens for landscape photography, it is of good value. If you want much better optical performance though, the Samyang / Rokinon 14mm f/2.4 is a much better lens in comparison.

Sigma 14mm f/1.8 Art vs Laowa 12mm f/2.8

The first copy of the Laowa 12mm f/2.8 I received was soft – the lens could not resolve much wide open, barely getting to acceptable levels at f/4. The lens was heavily decentered, showing very soft corner details on the right side of the frame, and sharper corner details on the left. I reached out to Venus Optics and they sent me another sample (claiming that the earlier sample was worn out due to heavy use by reviewers, which it certainly showed some signs of). I tested the second sample of the Laowa 12mm f/2.8 and it had the same issue as the first in terms of decentering, but this time, the left side of the frame was sharp and the right side was blurry. After failing to achieve good enough performance numbers, I was forced to tilt the lens towards the same angle of decentering, which made all four corners look acceptable. Take a look at the below image of the chart:

Yes, that’s how much the lens had to be tilted to equalize all four corners. Only after this, I was able to get a good enough representation of what this lens can do. Let’s take a look at the Imatest scores and compare the two lenses:

As you can see, the Laowa can produce decent center sharpness wide open, but its corners look quite a bit blurry when compared to the Sigma 14mm f/1.8 Art. If one stops the Sigma 14mm f/1.8 Art down to f/2.8, it absolutely crushes the Laowa 12mm f/2.8 in terms of sharpness and that’s across the frame! The Laowa f/2.8 showed some field curvature at closer distances and when I tried to make the corners look better, the sharpness in the center of the frame deteriorated considerably.

In fact, the Laowa f/2.8 has a pretty bad “wavy” field curvature, where the lens gets worse in the middle of the frame than the corners – this is evident when looking at its performance stopped down to f/8, where the corners outresolve the mid-frame. When shooting at wider apertures and looking at the whole image, the nasty effects of the wavy field curvature were very clear, where some parts of the frame looked much worse compared to others. When I saw pretty average numbers for center resolution at f/5.6, I thought that the lens suffered from focus shift, but even after focusing at f/5.6, I could never get any numbers that approached 3,000 in Imatest. This means that the lens overall is simply incapable of resolving enough detail on a 36 MP sensor…

Unfortunately, build quality of third party lenses, especially from smaller brands like Samyang and Venus Optics can be pretty disappointing – there is a lot of sample to sample variation in them based on my experience. Sigma is much better in this regard, but it is not free from variation either (and no manufacturer really is).