Click on images to open them in a separate tab for closer inspection

Intro

As a landscape photographer, it is important to have a wide angle lens in your tool kit. Although some people consider these types of lenses to be “landscape” lenses, that’s not strictly true, as pretty much any focal length can be used for landscapes. However, wide and ultra-wide lenses do offer the ability to create stunning landscape images, aided in some cases by the perspective distortion they create.

Let’s define “wide” and “ultra-wide” first. Generally, 24mm-35mm in full frame terms are considered wide, while anything shorter than that would be “ultra-wide”. To give you an idea about the difference, here is an image taken with my Sony 20-70mm f4 at 20mm, 23mm (close enough to 24mm), and 28mm. Some of the mismatch in the images is due to the fact I had to readjust the composition, while taking the image at 28mm. all images are unedited beyond slight exposure adjustments (to even out the image exposure).

Comparison of Basic Ultra-Wide and Wide Angles of View, Using Sony 20-70mm f4

As you can see, there is a considerable difference in the field of view between 20mm and 28mm. Interestingly, when I was still using my Sony a77, my main lens was the Sigma 18-35mm f1.8 Art. In full frame terms, this focal range translates to ~27mm-52.5mm. So, it is wide angle on the wide side, but not ultra-wide. Because of this, I often stitched several images into a panorama to obtain a wider field of view. I did have a wider lens, Rokinon 14mm f2.8 ED AS IF UMC. It’s a manual focus lens intended for a full frame, and translates to ~20mm on my Sony a77. I was never particularly happy with the quality of this lens, and used it only sporadically.

Now that I transferred to Sony a7 IV, which is a full frame e-mount camera, my main lens is Sony 20-70mm f4. It is a great lens, but there are some cases when I wish for a wider option. I did buy a manual adapter for my Rokinon, and it is able to produce good images if you take care, but, as I mentioned, I never particularly liked it. It has relatively significant distortion towards the edges of the image, which always bothers me, and the focus is not great.

Sony 20-70mm @20mm, Panorama and Focus-Stacked

With this in mind, a few weeks ago I rented two ultra-wide lenses from Lensrentals (They are great, by the way. So, if you need to try out equipment, give them a visit. I am not getting anything from them, and this is based solely on my experience with them). The two lenses I wanted to try were Venus Laowa 15mm f2 zero-D (this is a manual lens, which has been around for a while) and the new Sigma 17mm f4.

I took the lenses, together with my 20-70mm workhorse and the 14mm Rokinon to a short trip in Maine, and used them in standard landscape situations as well as some Milky Way photography.

Size, Build, and Pricing

There are, obviously, many additional ultra-wide options out there from more “standard” 16-35mm f2.8 GM or f4 G by Sony (or the Tamron and Sigma “equivalents’ of 17-28mm f2.8 and 16-28mm f2.8 options respectively), some prime lenses, such as the amazing Sony 14mm f1.8 GM and 20mm f1.8 G (or the Sigma 20mm f1.4 Art and f2 contemporary alternatives), and down to the premium zooms, such as 12-24mm GM by Sony and 14-24mm f2.8 Art by Sigma. My choice of lenses was driven by my needs and considerations (such as weight, ability to take standard filters, overlap with my existing lenses, as well as potential price). If you have a crop sensor, there are fewer options, but you should consider one of the dedicated crop-sensor lenses as they are likely to produce better results compared to full frame ones (besides the fact that they are considerably smaller and lighter).

As such, this comparison is not exhaustive by any means. It is also not particularly scientific. Granted, I tried to maintain the camera settings and composition when comparing things, but landscapes are not really controlled environments, and some differences would naturally exist. On the other hand, these comparisons are “real-life” rather than lab, which, in the end, are the ones most important to me.

Lenses Compared. From Left to Right — Sigma 17mm f4; Laowa 15mm f2 zero-D; Sony 20-70mm f4 G, Rokinon 14mm f2.8 ED AS IF UMC with Fotodoix A-to-E mount Manual Adapter

When you look at these lenses, first thing that jumps out is just how different they are in size and weight. The 17mm Sigma is just tiny. It looks like a toy on my camera. This is all the more amazing given the great quality it produces. It is also interesting that even though the Sony is a zoom lens with pretty big focal range, it is second lightest, albeit pretty close to the Laowa. Rokinon, being an older DSLR design (with an adapter added), was by far the heaviest. I don’t know if the e-mount version is lighter, but I think that the newer designs definitely are, albeit also significantly more expensive.

Despite the size differences, all the lenses are well built. The Sony G , being a zoom lens, also offers the biggest flexibility (there is a reason why it is my main lens).

In terms of price, Laowa and Rokinon are available on the second-hand market (although in case of Rokinon, you probably would look to the e-mount version rather than the A-mount, which I have), while the Sigma and the Sony are recent releases and would be more limited in this respect (I did manage to find several offerings of the Sony). Here is a comparison of weight and price for the lenses (second-hand price is taken as an approximate average from eBay completed sales). For the Rokinon, the pricing is for the e-mount version.

Weight (g) Weight (lb oz) Price New (B&H) Price Used (eBay)

Sigma 17mm f4 250g 8.8 oz 599$ n/a

Laowa 15mm f2 575g 1 lb 4.3 oz 649$ 350-500$

Sony 20-70mm G 545g 1 lb 3.2 oz 1,098$ 800-900$

Rokinon 14mm f2.8 687g 1 lb 8.2 oz 299$ 150-200$
+ adapter

Obviously, from price perspective, Rokinon is the cheapest, and if you can live with its quirks, it would do in a pinch when you are on a budget. Price for the new Laowa is rather high, but you can get it at reasonable price second-hand. Sigma is somewhat expensive right now as it is new, but will likely be available on the used market in a few months as well if you are patient.

Field of View and Image Quality in Good Light

Shooting one morning at Portland Head Light, I compared the field of view between the lenses to get a real world idea of the difference. I also added a 24mm from my Sony 20-70mm for comparison (as this might be what you have if you’ve got the “standard” 24-70mm zoom).

Field of View Comparison Between the Lenses

All images were shot with similar f16 aperture and the only editing done was to even out the exposure to similar levels for comparison (it required a bit more with the Laowa due to the sun being in the frame). As can be seen, there is a pretty big difference in real world field of view. In this case, the difference was all the more apparent when I switched to the Sigma and Laowa, since in these cases the sun was within the field of view. This presents its own set of advantages and challenges, and only you can say whether you can live with the differences. For my taste, 17mm here is probably the sweet spot, though it lacks the versatility of the Sony zoom.

All the lenses have their own distortion and vignetting profiles. Lightroom handles them ok, but in case of the Rokinon and Laowa there is no information transmitted and you need to choose the correction profile manually. Generally, Laowa, being a zero-D lens, offers the lowest distortion despite pretty wide field of view. However, the Sigma is not far behind with the correction applied.

I also compared the center (or somewhere between center and midframe) sharpness by cropping the lighthouse. Although this is no exactly the center, this area is relevant for landscape photographers as we often put our main elements somewhat off-center. If you are a portrait photographer, more centered image might be of more use to you (coming up in a bit).

In real life, looking at this crop, it is very difficult to see any difference between the lenses. However, if you really pixel-peep and look at the details around the window of the lighthouse, it looks like the Sony 20-70mm @24mm is the sharpest, followed closely behind by Sigma 17mm, and then Sony @20mm and Laowa. It should be noted that Laowa has the least amount of contrast in this image, but it might be due to some veiling due to the presence of the sun in the shot. Overall, in real life, I would be perfectly happy with either of these.

I took an additional test shot to compare the lenses. In this case, I produced two images for each of the lenses, one focusing on the tree trunk (closer to center) and the second one, focusing on part of the hedge closer to the corner. Granted, because of the differences in field of view, Laowa has an advantage here by having this area closer to the center of the image.

When examining the crops, the more central crop of the main tree showed that the Sony produced the sharpest image followed very closely behind by the Sigma. It would be very difficult to tell these apart without extreme pixel-peeping in my view. The Laowa, in this case, still looked good, but noticeably softer, though this might be due to my ability to focus manually.

“Center” and “Corner” Crops from the Second Test Image

In case of the “corner” crop I could not tell the difference (all images were focused on the same area of the hedge). This actually gives a slight advantage to the Sony, as this area is closer to the corner in that lens’ field of view. In any case, again, I would be pretty happy with either in real life.

Milky Way Photography

Milky Way photography is one of the common uses of ultra-wide lenses. 20mm and wider is generally considered as best for this type of photography. If you follow standard advice, you also want a faster lens (smaller f-number). Most people say that you need at least f/2.8 and f/4 really being not a particularly good choice. As an alternative, you can stack multiple images using software such as Sequator to overcome this issue. This is the reason why lenses like Sony 20mm f1.8 G are generally go-to choices for astrophotography.

However, with tools such as Topaz Denoise AI and the recent AI Denoise option in Lightroom, f/4 looks like a more viable option even for a single image capture (which removes the need to stack images). A couple of weeks before this test, I shot some Milky Way, using my Sony 20-70mm at Nobska Lighthouse on Cape Cod. With the applied AI Denoise in Lightroom, I was able to produce a single image pretty decent quality (I still needed to take several images to properly expose the foreground).

Milky Way at Nobska Light, Cape Cod Taken with Sony 20-70mm f4 G @20mm (Single Image Edit)

I went to the Nubble Lighthouse in York and spent a couple hours photographing the Milky Way there. I spent the first part of the evening, before the Galactic center was visible, taking multiple exposures with each lens to be used later for the foreground. These images were taken with a somewhat closed-down aperture to produce deep enough depth of field, and thus don’t offer a good comparison between the different lenses in terms of aperture.

After I was done with the foreground shots, I took 5-10 images with each lens at their respective maximum aperture to capture the Milky Way. Using PhotoPills app, I determined that I could get away with 13-15 second exposure to not create star-trails. For initial comparison, I used ISO of 3,200.

The Laowa image had the widest aperture of f2, and I used it as reference. I increased the exposure by +1 in Lightroom and then tried to match the other images to a similar brightness. This resulted in an adjustment of +1.7 for the Rokinon, and +2.4 for the Sony. Interestingly, although on paper the Sigma is two stops slower than the Laowa, I only needed +2 in exposure (meaning extra stop) to even out the exposure. This implies that the Sigma has a higher T-stop than the other lenses (with Sony also having slightly higher T-stop compared to Laowa then implied by their respective f-stops).

For each of the lenses, I enlarged a crop of the lighthouse as well as a crop from the Galactic center. The images are otherwise unedited.

Milky Way Comparison Between the Lenses with Full image and Crops of the Lighthouse and the Galactic Center

It should be noted that I was unable to properly focus with the Rokinon, no matter how much I tried. I also, unfortunately, slightly misfocused with the Sony. This can be seen by examining the lighthouse crop. Putting that aside, though, the Sigma was significantly sharper than the other images, including the Laowa. On the other hand, the aperture advantage for the Laowa gave it the cleanest images by far (both in the lighthouse area and the Galactic center area).

So far, not particularly unexpected, but it gave me a real life idea of the differences. However, to see whether an f4 lens can produce good results, I ran the Sigma and the Laowa images through AI Denoise in Lightroom and cropped the Galactic center again. I didn’t play with settings too much and used a setting of 38 for the amount of denoise. For comparison, I also stacked seven images taken with the Sigma, using Sequator. I put the Sequator output image through Topaz Denoise with automatic settings for “low light”. For each of the images, I show the comparison between the native image and the one that went through deonoise.

As they say, results were very interesting…

Effects of Denoise and Image Stacking on Noise Levels of the Sigma and Laowa Images

Laowa image that went through denoise was crazy clean, with the noise levels greatly reduced. Surprisingly, the Sigma image, although not as clean, was very clean and definitely cleaner than the Laowa image without the denoise applied. The stacked Sigma images were cleaner than the single one, and the stacked images with applied denoise were cleaner than Laowa untreated image, though not as clean as the denoised Sigma single image. This was most likely due to differences between denoise implementation in Topaz vs Lightroom. It should be noted that both offer decent control of denoise and sharpening levels and, potentially, can produce better images than shown here.

Another couple of things I noticed were that denoise in Lightroom produced some artifacts that you would need to deal with in subsequent editing. The second thing was that stacking images produced a slightly softer image than denoised single image. I am not sure whether this is inherent in the process or due to some mistake I made.

Overall, Laowa definitely produced the best results in this test, but single denoised image from Sigma was not that far behind. This confirmed my previous evaluation that with modern denoise software you can likely get away with an f4 lens for shooting the Milky Way.

Conclusion

I had fun comparing these lenses and came to interesting conclusions. When making the decision, first thing you need to figure out is whether you need the extra field of view compared to the 20mm, offered by the Sony 20-70mm G. It is an excellent lens, and if you don’t need to go wider, you probably will be ok with just that lens as your ultra-wide to normal and short telephoto range. If, on the other hand, you decide that you want to go wider, the choice is not as obvious.

I would definitely exclude the Rokinon from consideration, unless you are really stressed for budget. It is not very sharp, and even in the situation like the Milky Way, where its brighter aperture should offer some advantage, it was basically unusable, and offered only 1/3 of a stop advantage over the Sigma.

Sigma is really a gem of a lens. It is so compact and light, you barely feel it in the bag. In good light situations it is sharp basically edge to edge, and will be an excellent choice. In addition, it appears that it has higher T-stop, i.e. it lets through more light than expected by its nominal aperture. As a result, in the real world, it is only about one stop slower than the Laowa. With the new denoise options, you can almost match the Laowa output in terms of cleanliness, but have a sharper image (Laowa wide open is definitely softer). Finally, it is an auto-focus lens, which has its advantages.

If you are looking for the best astrophotography lens out of the bunch, and don’t mind the manual focus for other situations (you generally use manual focus for astro, anyway), Laowa is a good choice. It is also available on the used market for considerably less money. However, keep in mind that it is larger and heavier.

In the end, the choice between the two will come down to your personal preferences and use cases. Both options are good, and if you wait a little, Sigma will likely pop up on the second-hand market in half a year as well.

Hopefully, this helped you make a decision if you were in the market for an ultra-wide angle lens.

Have a great week!

Budget Ultra-Wide Angle Lens Options for Sony FE-Mount (Full-Frame) Review

Updates

To Print or Not to Print… a Photobook?