My most popular video of 2024 was a showdown between the (then) primary standard zoom options available on the Fuji X-mount platform – the Fuji XF 18-55mm F2.8-4 R LM OIS, XF 16-55mm F2.8 R LM WR, and the Sigma 18-50mm F2.8 DN. My surprising conclusion was that the Sigma actually held up the best on the modern Fuji 40MP resolution standard, and that the two Fuji zooms were in need of an update. Fuji obviously agreed, as both lenses received an update in the latter part of 2024. I reviewed the XF 16-55mm F2.8 R LM WR II and the XF 16-50mm F2.8-4.8 R LM WR in early 2025, and decided to follow those reviews up with another showdown using the fresh Fuji options along with last year’s Sigma winner. The results? Find out in the video review here…or you can see my review notes in the article below.
The Fuji lenses are loaners from Fujifilm Canada, while the Sigma is my personal lens. As always, this is a completely independent review. *I’m doing this comparison on a 40MP Fujifilm X-H2 camera.
16-55 II – Sharpest, best corners, best contrast, most consistently good
16-50 (TIE) Slightly better corners on wide end than Sigma
Sigma (TIE) – Aperture advantage vs 16-50, better corners at telephoto than 16-55 II
COLOR
Fuji lenses perhaps slightly better color, but hard to see side by side.
BOKEH
16-55 II – Bokeh richer, no bokeh fringing, most potential (55mm)
Sigma – Decent geometry but fringing
16-50 – Slow aperture means least defocus
Reasons to Choose Sigma
Cheapest MSRP
Constant F2.8
Fairly competitive AF and IQ (balanced)
Highest magnification (0.36x vs 0.21x)
Very compact
Reasons to Choose 16-50mm
Can be had as kit lens at cheaper price
Only internally zooming lens
Weather sealed
Most lightweight
Mostly good AF
Reasons to Choose 16-55mm II
Most professional grade build
Best zoom range
Best feature set
Lowest fringing
Best overall image quality (most special)
Best autofocus
Best standard APS-C zoom out there
Conclusion
A sense of order has been restored. The first party lenses from Fuji are now mostly better than the Sigma, though the Sigma will continue to be intriguing because it is small, cheap(ish), lightweight, and constant F2.8. If you can handle a little more size (and cost) the XF 16-55mm F2.8 II is the best standard APS-C zoom you can buy right now.
Keywords: Fujinon XF 16-55mm F2.8 R LM WR II, Fujinon XF 16-50mm F2.8-4.8 R LM WR, Sigma 18-50mm F2.8 DN, 18-50mm, DN, DC, Fuji, Fujinon, 16-55mm II, R LM WR, 16-50mm, F2.8, F2.8-4.8, OIS, LM, WR, R, X-Mount, X, Fuji, Fujifilm, X-H2, weather sealing, Review, Hands On, Dustin Abbott, Real World, Comparison, Sharpness, Autofocus, Image Quality, Sample Images, Video, Photography, Astrophotography, #letthelightin, #DA, #weather sealing, #VS, #headtohead, #headtoheadbattles
DISCLAIMER: This article and description contains affiliate links, which means that if you click on one of the product links, I’ll receive a small commission. As an Amazon Associate I earn from qualifying purchases.
Sirui’s new Aurora series of full frame lenses has kicked off with the well received Aurora 85mm F1.4. I was interested in doing some direct comparisons with existing options. The most logical comparison is the Samyang AF 85mm F1.4 Series II lens, but since I don’t have that one on hand, I’m doing a comparison to what has become my standard in this class – the Sigma 85mm F1.4 DN | ART. You can find out my full thoughts by watching the video review below…or reading on.
Thanks to Sirui for sending me a review sample of this lens. As always, this is a completely independent review. *The tests and most of the photos that I share as a part of my review cycle have been done with my Sony a7RV or Sony Alpha 1 cameras.
Mostly similar feature set (click, declick, function, AF | MF)
Both weather sealed
Both come with lens hood and case
Same rated magnification (0.12x)
Reasons to Choose Sigma:
Extra feature – iris lock
Higher contrast
Sharper
Faster autofocus
Better up close performance (>magnification, contrast)
Richer color
Less fringing
Better centering
Winner if you like a higher contrast, more deeply saturated result.
Reasons to Choose Sirui
Price | MSRP $599 vs 1199 | Current $499 vs 949
Weight | 540g vs 625g
15 rounded blades vs 11
Smaller filter threads 67mm vs 77mm
Faster video focus pulls
Comes with two filters (if you buy from Sirui)
Lower distortion and vignette
Rounder specular highlights wide open
Winner if you like a lower contrast, more cinematic looking result (almost looks like Cinetone is on).
Conclusion
The Sirui Aurora 85mm F1.4 is a lot of lens for the money. I personally prefer the results from the Sigma, but I could live with the Sirui. I still think the Sigma strikes the best balance between high end performance and price, but the fact that it costs twice as much as the Sirui could be a serious factor for people.
Keywords: Sirui Aurora F1.4 GM II, Aurora, Sirui, F1.4, 85mm, Sigma, DN, Sigma 85mm F1.4, ART, DG, Comparison, Versus, VS, Full Frame, Review, Sony Alpha 1, Sony a7RV, Review, Hands On, Dustin Abbott, Real World, Comparison, Sharpness, Bokeh, Flare Resistance, Autofocus, Image Quality, Sample Images, Video, Photography, let the light in, weathersealing, #letthelightin, DA
DISCLAIMER: This article and description contains affiliate links, which means that if you click on one of the product links, I’ll receive a small commission. As an Amazon Associate I earn from qualifying purchases.
The Sigma 10-18mm F2.8 DC DN is the second of two Sigma zooms that we’ve seen rereleased on Canon’s RF mount, and they are very welcome. Earlier this year as I was doing my review of the Canon EOS R50 compact APS-C mirrorless camera I couldn’t help lament the state of the lens options for Canon’s RF-S mount. At the time of the review (two years since the release of the EOS R7, the first of Canon’s RF mount APS-C cameras), this was the sum total of the available lenses.
Yuck. I couldn’t imagine a less exciting lineup of lenses. The fastest maximum aperture in the bunch was F3.5, and that lens is F4 and smaller before you make it to 30mm. Canon clearly knew they had a problem, for between the time that I filmed my review and the time I released it, they announced that certain Sigma (and later Tamron) lenses would become available for RF-S (the APS-C version of Canon RF). Now, to be clear, this is only for APS-C lenses, but for now it seems like Canon is going to let third parties solve their APS-C lens problem. This is great news for consumers, however, because it means that Sigma’s lineup of DC DN lenses will be coming to Canon RF. The first release was the Sigma 18-50mm F2.8, a lens that I actually own and gave a positive RF mount review here. A few months later they have followed up with the second zoom release, this time the wide angle Sigma 10-18mm F2.8 DC DN. Is this also a lens worth considering for your Canon R-mount APS-C camera? You can find out my thoughts in the video review below…or read on.
Thanks to Gentec (Sigma’s Canadian Distributor), for sending me a review loaner of this lens. As always, this is a completely independent review. All opinions and conclusions are my own.*The tests and most of the photos that I share as a part of my review cycle have been done with the 45MP Canon EOS R5, which I reviewed here.
This becomes the first “fast” wide angle zoom for Canon RF-S (APS-C), giving a constant aperture of F2.8 over its zoom range. Now, to be fair, that’s not a big zoom range, but then again, Canon’s own RF-S 10-18mm not only has that same small zoom range but also has a variable aperture of F4.5-6.3! That means that the Sigma is over two stops faster at 18mm (F6.3 is 2 1/3rds stop slower than F2.8). And, while that zoom range is not large, it is enough to give you a variety of options for framing a scene, from this at 10mm:
…to this at 18mm:
This review of the 10-18mm RF is not a whole new review, but rather an update on how the 18-50mm RF translates to Canon RF. I’ve already tested the image quality at both a 26MP (Sony) and 40MP (Fuji) in 2023, so that covers the full gambit (and more) of what’s available on RF bodies. There are no optical surprises here, but I’m always curious to see how the lens changes in terms of its dimensions, but, most importantly, how well it autofocuses on a platform that it wasn’t originally designed for.
The 18-50mm RF is one of six lenses released by Sigma for Canon RF, and those six lenses represent all of the APS-C mirrorless lenses they’ve released over the past three years. You can see my review of the original releases on Sony here.
Sigma’s DC DN lenses have all fallen under the umbrella of their Contemporary lineup, and while Sigma has proven willing to experiment with new features and design elements within the ART and Sport lineups, the Contemporary lenses have all stuck to a fairly rigid design formula. That continues to be the case here, so this little zoom mostly looks and handles pretty similarly to other lenses in the lineup.
There’s nothing new here on the RF mount, but the lens is more of a standout on Canon due to having essentially zero quality competition. The 10-18mm RF is a genuine treat as an addition here. Why? Here’s a few reasons:
Has at least a weather sealing gasket
Better build quality than RF-S lenses
Includes a lens hood
Constant F2.8 aperture
Good up close performance
Great autofocus on RF
In essence, the poor competition makes this lens a star, though frankly I still found it one of the more compelling options on these other platforms as well.
The crop factor on Canon is different than Sony or Fuji, so the zoom range works a little different. Fuji/Sony have a crop factor of 1.5x, which means that the lens goes from a full frame equivalent of 15-27mm. Canon’s APS-C crop is 1.6x, and so that changes the effective focal range to 16-29mm. You’ll miss the extra bit of width at 10mm but get a little more reach. In this case, I think that most of us would prefer the extra millimeter of wide angle coverage, but that’s just not going to happen.
Marked positions in the zoom range are 10, 12, 14, 16, and 18mm, and here’s what the same scene looks like at each of these marked positions.
There’s really only one significant change to the 18-50mm on Canon, and that is that the lens mount portion of the lens has to be flared out to accommodate the much larger diameter of the RF mount.
Unlike the 18-50mm, however, this isn’t the widest part of the lens, so the exterior measurements don’t change despite this change. It is 72.2mm in diameter and 62mm in length (2.8 x 2.4″). That’s actually 2mm shorter than the Sony version, and this obvious reflects a slightly different flange difference between the two mounts. That extra bit of girth near the lens mount also adds a tiny bit of weight, with the RF version weighing 260g (9.17 oz). Other than these minor physical changes to the RF version, the build is otherwise identical to either the E-mount of X-mount versions.
This weighs about 110g more than the Canon, though the Canon is a very plasticky lens that also has a very slow aperture. The build quality between these two lenses can’t really be compared.
The 10-18mm RF is a mixture of metals and “thermal composites” (high grade engineered plastics) in construction, and the construction feels quality in the hand. It is built around a metal mount, unlike the Canon lens, which has a plastic mount.
The overall build quality just feels much higher than any of the Canon RF-S lenses that I’ve tested so far. This feels like a quality lens.
Up front we have the ability to use traditional filters in a very standard 67mm size.
There is a weather sealing gasket at the lens mount (thicker here on the RF mount than on previous versions), but, like other Contemporary lenses, there are no internal seal points.
The only thing on the barrel is the focus and zoom rings. The closest to the lens mount is is the focus ring, which is fairly narrow (10mm), ribbed, and moves smoothly. Like other lenses designed for mirrorless cameras, this is a focus-by-wire system in which input from the focus ring is routed through the focus motor. It is actually the focus motor that moves the elements, making manual focus more of an emulation than a direct movement of the elements. Some lenses pull this off better than others, and the Sigma 10-18mm DN is one of them. There is good weight/damping on the focus ring, and no visible steps during focus changes. I didn’t notice any focus lag either; the action on the ring and the movement of focus seemed directly linked. I will note that the lens will automatically alert the body to magnify the image when you are manually focusing, which is a big help in visually confirming focus. The focus weight is a little lighter than what I like, but not bad.
There is about 10mm between the focus ring and the beginning of the flared section that leads to the zoom ring. I appreciate some space there, as that was one of my complaints about the Sony 10-20mm; there was almost no room between the two rings on that lens, and it was easy to inadvertently move the wrong ring. There’s a very definite difference between two rings and position on the Sigma, though there is a different problem. At the fully retracted position (there is a slight external zoom action), the lens hood fits so tightly against the zoom ring that there isn’t a lot of space for your thumb to fit on the narrow ring. If you happen to have longer fingernails, you might actually find zooming a bit of a challenge.
The zoom action itself is very smooth. This is a “reverse zoom” in terms of the retraction; the inner barrel is fully retracted at 18mm and extends the furthest at 10mm, though the barrel extension is only about 10mm.
Sigma is experimenting with a new lens hood design for this lens. Rather than bayoneting into place, it is actually just a push on design. You still need to line up the markings on the hood and the barrel, but then you just push the lens hood into place rather than rotate it. A spring/lever mechanism holds the hood locked into place. You can release it by just doing a slight rotation to the left.
The primary “pro” of this design is it allows them to the make the lens hood a little thinner as there doesn’t need to be room in the hood design for the bayonet action. Sigma touts the fact that while the filter size of the 10-18mm is 12mm larger than the 18-50mm (67 vs 55mm), the diameter of the lens hood is only 3.8mm larger.
If you want to reverse the hood for storage, you now line up an arrow on the lens hood with the previous mark on the lens barrel and push it forward. Removing it simply requires a similar slight rotation to the left. As per usual, Sigma’s lens hoods are just a little bit nicer than the competition, including some soft-touch materials, texture variations, and lens information imprinted on them.
The 10-80mm DN has two different minimum focusing distances for the wide and telephoto ends of the zoom range. At 10mm you can focus as closely as 11.6cm (4.6″) and can get as high as a 0.25x magnification…though you have to be pretty much right on top of your subject to get it.
At 18mm that disance extends to 19.1cm (4.6″) and the magnification drops to just 0.14x…though shots at 18mm are a whole lot easier to get!
Get close enough and you can produce some creative shots like this.
The aperture iris is electromagnetic and controlled from within the camera (no aperture ring). It has seven rounded aperture blades that do a fairly good job of retaining a circular shape even with the lens stopped down, though frankly you won’t have a lot of opportunity to shoot shallow depth of field shots with a wide angle lens like this.
Stopping down will produce a decent (but not exceptional) 14 bladed sunstar effect.
There is no optical stabilization on the lens, unfortunately, which is the one advantage I can point to for the Canon RF-S 10-18mm. I happened to test on a camera that had in body image stabilization, so I didn’t miss it, but many of Canon’s smaller APS-C cameras don’t have IBIS.
On other platforms the Sigma is a rather plain lens in terms of features and build. Because of so little competition on Canon, however, it feels next level in terms of build and handling. It’s a nicely made little lens that follows a familiar, well-executed Sigma formula.
Autofocus and Video
Sigma continues to utilize a stepping focus motor (STM) in these smaller lenses as the elements are smaller and lighter than lenses and don’t require the higher torque that some of their large aperture full frame or sports oriented lenses. Autofocus is something that Canon does very, very well, and it is clear to me that Sigma has received access to the focus protocols and algorithms, as the 10-18mm RF definitely handles like a first party lens.
The STM motor provides fast, quiet, and accurate autofocus. Eye detection works fine, and I was able to grab some action photos without difficulty.
AF had no problem with a group photo of some friends and I on the golf course, though, to be fair, it isn’t difficult to have a LOT in focus with a wide angle lens like this.
Autofocus speed was excellent, near instantaneous both indoors and outdoors, and utterly reliable and accurate. My focus tests showed extremely rapid transitions from close to distant subjects. Focus breathing is extremely low, so this helps the transitions feel very smooth and rapid.
This is also a huge factor in video recording, as focus pulls are not only quick and confident, but the extremely low focus breathing helps them to feel very seamless and cinematic. No steps, no pulses, but just clean, smooth transitions that are nicely cinematic.
My “hand test” (where I block the view of my face with my hand and then move the hand to allow focus to pop back to the eye) also went well. Focus transitioned confidently from my hand to my test and vice versa.
The lens is not parfocal, however, and I did notice that there were moments of defocus before the lens refocused while zooming and do video recording. I also noticed a bit of warping as I began to move the from the wide end towards the telephoto end.
But overall this is a really solid autofocus system that delivers a fantastic performance on Canon.
Sigma 10-18mm RF Optical Performance
The Sigma 10-18mm RF sports a relatively complex optical formula for such a small lens. There are 13 elements in 10 groups, and that includes number of exotic elements like 4 aspherical elements (including a large GM aspherical concave lens as the front element), 3 FLD elements, and one SLD element. The MTF chart and comparison shows a significant improvement over the older Sigma 10-20mm F3.5 that the was equivalent lens for DSLRs. The MTF chart shows a very strong center and midframe at both ends of the zoom range with minimal falloff in the corners at 10mm. There’s a stronger falloff at 18mm in the corners.
In this section I will be reusing some of my results from the 40MP Fuji X-Trans sensor. It is higher resolution than anything Canon currently makes (current resolution points are between 24 – 33MP. Fuji’s images are also harder to sharpen than Canon, so you will have no problem getting beautiful results on Canon bodies. You can expect sharpness results to look better on Canon. I don’t currently own a Canon RF APS-C camera, so I’m doing this test on my the APS-C mode of my EOS R5, which is only 17MP. That’s fine for showing real world images as I’ve done in this review, but not really for evaluating the limits of the performance.
It has no problem producing beautiful images.
For those that are interested, Canon is not going to allow you to even experiment with this lens for full frame. Only the 1.6x crop option is available; all other options are greyed out.
I will redo the vignette and distortion tests, however, as the vignette amount does change from one mount to another. Here’s a look at the before and after of manual corrections of both at 10mm:
I didn’t expect to see any real difference in distortion (it’s the same optical formula), but I’ve noticed a trend with lenses first developed for Sony that then move other mounts that the vignette tends to be heavier.
At 10mm there is a significant amount of barrel distortion. I settled on a +23 to correct it, though there is definitely some “mustache” shape to the distortion pattern where correcting the barrel distortion across the majority of the frame tends to create some pincushion distortion in the corners. The standard profile does a cleaner job of correcting the distortion. On Sony I needed a +53 to correct the vignette, but on Fuji and on Canon I had to max out the slider (+100). That’s close to two stops higher.
So how about the “telephoto” end of the zoom range?
The distortion flips to pincushion style at 18mm but it was very linear and easy to cleanly correct for (I used a -4 on Canon). Vignette was a little lighter but still required a +74 to correct (+36 more than on Sony). Sigma gets profile support on Canon in camera for everything but distortion for some reason. There is already a correction profile available for Lightroom/ACR that will also clean things up nicely for RAW files. So, not a flawless performance, but nothing critical here, either.
How about chromatic aberrations?
I saw very little longitudinal chromatic aberrations before and after the plane of focus. You can see very little fringing in all of the shiny metal and crystal bits on this decorative reindeer.
Lateral Chromatic Aberrations usually show up along the edges of the frame as fringing on either side of high contrast areas, but that doesn’t seem to be an issue here. Winter is the most obvious time to see issues with LaCA due to all the bare branches, but I don’t see any issues here even with corrections turned off.
That’s good news, as you are far more likely to see lateral CA on a wide angle zoom in real world images than you are to see longitudinal CA.
So how about resolution and contrast? All chart tests done with a the X-H2 (40MP) using a tripod and a two second delay. Here’s a look at my test chart:
And here are the crops (at roughly 200% magnification) from the center, mid-frame, and extreme corner shot at 10mm and at F2.8:
In the center of the frame we can see that the Sigma 10-18X has no problem. It still looks fantastic. Mid-frame is a little murkier, though the corner look relatively good. Stop down to F4 and the mid-frame and the corners take a nice leap forwards:
There’s a bit more on tap at F5.6, so that means that real world landscape images at 10mm will look very nice all across the frame.
Diffraction comes early on pixel dense cameras like the X-H2, so I actually found F8 to have very slightly less contrast compared with F5.6. After F8 things go downhill fast as diffraction robs away contrast and acuity. By F22 (minimum aperture) the image looks very soft.
I would recommend sticking with F5.8-F8 for landscape work on a high resolution camera to get your best results.
I slightly preferred the overall image quality at 12mm. The center is about the same as 10mm, but the results in the midframe and corners are clearly better.
Real world 12mm images look great, though the absolute edges aren’t pin sharp on 40MP.
14mm is very similar to 12mm in contrast, though I didn’t feel like it resolved some of the fine lines quite as well. Take a look at the number 2 and the area around it on the right as compared to the left (12mm).
There’s a clear improvement at F4, however, and landscape apertures look pretty great in real world images.
16mm is mostly similar to 14mm with perhaps a little better detail and contrast…particularly along the edges of the frame.
I did note a mild regression at 18mm in my Sony E-mount review, but I’m seeing it less here on Fuji. As I compare 16mm and 18mm across the frame I see some give and take. Some areas favor 16mm, others favor 18mm. The performance is mostly similar between the two focal lengths, making the overall sharpness and contrast very consistent across this zoom range.
I chose this crop because it illustrates the challenge well – there is very little to distinguish between the two focal lengths…even side by side and magnified to 200%!
As before, you can see some improved contrast a bit more detail by stopping the lens down. I found a more significant different at F5.6 than at F4. By F5.6 the improved contrast is very noticeable.
The two ends of the zoom range are the most important, as you’ll often be in situations where you want to go as wide as possible, while the 18mm end will give both the maximum reach but also will be the go to for up close work…like in the shot of these mushrooms.
You can see from the crop that while the amount of detail at F2.8 on a high resolution sensor isn’t pin-sharp, it is sharp enough to work in most situations.
You are also most likely to get visible bokeh at 18mm. This image allows us to look at both the sharpness of the subject and the bokeh beyond.
At a pixel level I can see that contrast doesn’t “wow” me, but at every other viewing magnification the image looks great. There’s no “bleeding” around the edges of the leaves due to low contrast, and the quality of the bokeh is pretty nice. Wide angle lenses are almost never going to give you amazing bokeh (that’s just not their strength), but I would be happy in getting images like the one above.
Here’s another bokeh shot of a field of frost covered sweet fern. You can see a bit of busyness in the transition zone as things move out of focus (some outlining), but beyond that the bokeh gets fairly soft and creamy.
Flare resistance was fairly good in most situations, with good contrast and only minor issues with ghosting. That performance was pretty similar to what I saw on Sony. Here’s a fresh shot that shows good flare resistance.
I did see some “flashing” with the sun right out of the frame when shooting video, though I did find it fairly artistic and not distracting. Again, that’s a subjective evaluation, and I would recommend watching the video review to draw more conclusions for yourself. Here’s a few screenshots to show the effect.
Overall my feelings were very positive about the flare resistance.
Canon has nice color science, and that means that it isn’t difficult to get images with beautiful color when using the 10-18mm RF on a Canon body.
Here’s another shot that shows nice, rich levels of color saturation.
The Sigma 10-18X is able to navigate the bump to 40MP on Fuji better than most zoom lenses I have tested on the sensor, and that means that Canon RF shooters at lower resolution levels (and easier sensors to sharpen) should be pretty satisfied with the amount of detail you can achieve with this lens.
That Fuji sensor is a monster for making lenses look softer at a pixel level, but that’s not on Sigma. Compared to other lenses, it actually holds up well. Here’s the Tamron 11-20mm F2.8 RXD (another lens that looked very good on Sony but softer on Fuji), and you can see that the Sigma definitely better.
All things considered, I feel like there is a lot of optical performance here for such a small package. You can see many more beautiful images by visiting the image gallery here.
Conclusion
I’ve been recommending the Sigma 10-18mm F2.8 DC DN to a lot of people looking for a good APS-C wide angle zoom on multiple platforms, so I’m not surprised that it performs well on Canon RF mount. Autofocus was fantastic, image quality looks great, and the lens remains very impressive despite its small size. It’s a great match for smaller APS-C bodies in size, and the performance is strong enough to delight owners.
There really isn’t much in terms of competition, so it’s an easy lens to recommend on RF. It is actually the premium option, and while it is isn’t cheap at $659 USD, neither is it ridiculously expensive.
It’s great to see Canon’s RF mount starting to open to these third party offerings, and very quickly we will see the amount of lenses available for RF-S (APS-C) tripled due to Canon’s move to open the protocols. This gives Canon shooters a seriously good little option, and the Sigma 10-18mm 2.8 DN is what I would be buying if I were in the market right now.
Pros:
Fantastic job porting to Canon
Extremely compact and lightweight
Nicely made with a weather sealing gasket
Useful focal range with constant F2.8 aperture
Great autofocus performance
Good video AF performance
Almost non-existent focus breathing
Good sharpness across zoom range
Nice colors
Good flare resistance
Nicer build and handling than RF lenses
Cons:
No optical stabilizer
Fairly strong and complex barrel distortion at 10mm
DISCLAIMER: This article and description contains affiliate links, which means that if you click on one of the product links, I’ll receive a small commission. As an Amazon Associate I earn from qualifying purchases.
People have been waiting for the Mark II version of the Sony 85mm G Master lens for a long time. There has been a lot of conjecture over whether or not it will be an 85mm F1.2 lens to match the 50mm F1.2 GM that I reviewed here and to compete with similar lenses from Canon or Nikon. But Sony has done their internal calculations and determined that the best way forward for them is a less extreme F1.4 design that will instead be a match for the 24mm F1.4, 35mm F1.4, and 50mm F1.4 GM lenses, and the new Sony FE 85mm F1.4 GM II is a definite match in terms of size, design, and aesthetic to those lenses. This also allowed them to keep the price at a more moderate level at $1799 USD, which will almost certainly result in higher sales. But the market is also pretty saturated at the 85mm focal length on Sony E-mount, so will the 85GM II (as we’ll call it for brevity here) be able to stand out amongst the competing lenses at lower price points…most notably the excellent Sigma 85mm F1.4 DN? That’s what I explore in detail in this video…or just can read my summary below.
Thanks to Sony USA for sending me a review loaner of the GM II lens. As always, this is a completely independent review. *The tests and most of the photos that I share as a part of my review cycle have been done with the Sony a7RV along with the Sony Alpha 1 that serve as my benchmark cameras for Sony lenses.
Similar feature set (click, declick, iris lock, function, AF | MF)
Both weather sealed
Both have a 77mm filter thread
Both come with lens hood and case
Reasons to Choose Sigma
Price: $1070 (normal $1200) vs $1800
Smaller – Sigma is 2mm narrower and 11mm shorter
Lighter – Sigma is 17g lighter
Has softer bokeh
Slightly better magnification
Reasons to Choose Sony
Slightly more features (extra focus hold)
Faster, more confident autofocus
Lower focus breathing
Less distortion
Less vignette
Better microcontrast and rendering of fine details
Conclusion
The Sony 85GM II has two definite strengths: better autofocus and a better microcontrast and delivery of the details. But is that enough to offset the over 50% price difference between the two lenses? I think that depends on your actual needs and whether they stray into the areas where the GM’s advantages will actually make a difference. If not, then the Sigma is probably still the better option.
Keywords: 85mm F1.4 GM II, Sony FE 85mm F1.4 GM II, 85 GM II, GM, II, F1.4, 85mm, Sigma, DN, VS, Full Frame, Review, Sony Alpha 1, Sony a7RV, Review, Hands On, Dustin Abbott, Real World, Comparison, Sharpness, Bokeh, Flare Resistance, Autofocus, Image Quality, Sample Images, Video, Photography, let the light in, weathersealing, #letthelightin, DA
DISCLAIMER: This article and description contains affiliate links, which means that if you click on one of the product links, I’ll receive a small commission. As an Amazon Associate I earn from qualifying purchases.
I’ve been in the review business for about 13 years now, and over that time I’ve heard the requests from my audience for someone to produce a 24-105mm F2.8 full frame zoom. The 24-105mm F4 zoom has been ubiquitous, but always the F2.8 counterpart had a more constrained 24-70mm zoom range. That has lead to countless deliberations between the bigger zoom range of the 24-105mm or the larger maximum aperture of the 24-70mm zooms. But what people really want is, well, everything. Canon was the first to the table with their 24-105mm F2.8 L IS USM Z lens, though that lens will set you back $3000 and weighs 1.3 kilos. Not an “everyman’s lens”, for sure. But while Sigma’s newest lens is a little less ambitious – the Sigma 28-105mm F2.8 DG DN | ART lens – the price is just half of the Canon lens at $1499 USD. Does that make the 28-105 DN the new most desirable zoom out there? We’ll explore that question in either my video review below , in the text review, or just by enjoying the photos in the galleries below.
Thanks to Sigma Canada (Gentec) for sending me a review loaner of this lens. As always, this is a completely independent review. *The tests and most of the photos that I share as a part of my review cycle have been done with the Sony a7RV along with the Sony Alpha 1 that serve as my benchmark cameras for Sony lenses.
The top competitor to the new 28-105 DN on Sony will almost certainly be the Tamron 35-150mm F2-2.8 VXD. While the focal ranges are clearly not identical, both of them are traditional zoom range blurring high performance lenses. The Tamron has been my “if I only could have one lens” lens since it’s launch, and I use my copy extensively. I will draw some comparisons to the new Sigma lens, as I feel like both lenses are largely reaching for the same audience.
Sigma’s approach to keeping this lens at a reasonable size actually takes a page from the Tamron book; they’ve started at 28mm rather than 24mm. In this case, in fact, the 28mm wide end of the Sigma still trumps the 35mm wide end of the Tamron, though obviously the Tamron also happens to have an extra 45mm on the telephoto end. Which of those is more useful will probably mostly depend on your own shooting needs. That extra 20mm beyond 85mm does offer up some interesting portrait applications, allowing you a little more compression. But while Sigma is focused on portrait and landscape photographers in their marketing for the 28-105 DN, I think the most interesting market is actually event photographers (particularly wedding photographers). The Tamron 35-150mm has been my absolute go-to lens for all kinds of events because it includes the meat of both a 24-70mm and 70-200mm zoom. 35mm is very often wide enough in most venues, and being able to switch framing between wide and telephoto perspectives is incredibly useful. The Sigma allows for the same kind of approach in a slightly different way.
The Sigma 28-105 DN also adds some interesting close focus capabilities with a 0.32x magnification on the telephoto end, allowing for some “macro-ish” shots.
Keywords: Sigma 28-105mm, DG, DN, ART, Sigma 28-105 F2.8, Sigma 28-105mm, F2.8, Sigma 28-105 Review, #SIGMAEmount, #SIGMA #SIGMA28105mmF28Art, #SIGMAArt, #SIGMADGDN, Full Frame, Review, Sony Alpha 1, Sony a7RV, Review, Hands On, Dustin Abbott, Real World, Comparison, Sharpness, Bokeh, Flare Resistance, Autofocus, Image Quality, Sample Images, Video, Photography, let the light in, weathersealing, #letthelightin, DA
DISCLAIMER: This article and description contains affiliate links, which means that if you click on one of the product links, I’ll receive a small commission. As an Amazon Associate I earn from qualifying purchases.
I’ve been in the review business for about 13 years now, and over that time I’ve heard the requests from my audience for someone to produce a 24-105mm F2.8 full frame zoom. The 24-105mm F4 zoom has been ubiquitous, but always the F2.8 counterpart had a more constrained 24-70mm zoom range. That has lead to countless deliberations between the bigger zoom range of the 24-105mm or the larger maximum aperture of the 24-70mm zooms. But what people really want is, well, everything. Canon was the first to the table with their 24-105mm F2.8 L IS USM Z lens, though that lens will set you back $3000 and weighs 1.3 kilos. Not an “everyman’s lens”, for sure. But while Sigma’s newest lens is a little less ambitious – the Sigma 28-105mm F2.8 DG DN | ART lens – the price is just half of the Canon lens at $1499 USD. Does that make the 28-105 DN the new most desirable zoom out there? We’ll explore that question in either my video review below or in the text review that follows.
Thanks to Sigma Canada (Gentec) for sending me a review loaner of this lens. As always, this is a completely independent review. *The tests and most of the photos that I share as a part of my review cycle have been done with the Sony a7RV along with the Sony Alpha 1 that serve as my benchmark cameras for Sony lenses.
The top competitor to the new 28-105 DN on Sony will almost certainly be the Tamron 35-150mm F2-2.8 VXD. While the focal ranges are clearly not identical, both of them are traditional zoom range blurring high performance lenses. The Tamron has been my “if I only could have one lens” lens since it’s launch, and I use my copy extensively. I will draw some comparisons to the new Sigma lens, as I feel like both lenses are largely reaching for the same audience.
Sigma’s approach to keeping this lens at a reasonable size actually takes a page from the Tamron book; they’ve started at 28mm rather than 24mm. In this case, in fact, the 28mm wide end of the Sigma still trumps the 35mm wide end of the Tamron, though obviously the Tamron also happens to have an extra 45mm on the telephoto end. Which of those is more useful will probably mostly depend on your own shooting needs. That extra 20mm beyond 85mm does offer up some interesting portrait applications, allowing you a little more compression. But while Sigma is focused on portrait and landscape photographers in their marketing for the 28-105 DN, I think the most interesting market is actually event photographers (particularly wedding photographers). The Tamron 35-150mm has been my absolute go-to lens for all kinds of events because it includes the meat of both a 24-70mm and 70-200mm zoom. 35mm is very often wide enough in most venues, and being able to switch framing between wide and telephoto perspectives is incredibly useful. The Sigma allows for the same kind of approach in a slightly different way.
The Sigma 28-105 DN also adds some interesting close focus capabilities with a 0.32x magnification on the telephoto end, allowing for some “macro-ish” shots.
So does Sigma have a hit on their hands? We’ll dive in a little further to find out.
Build and Handling
Sigma’s build standards are continually evolving, and they have been very proactive about adding new features and improvements. Many of those feature improvements have been designed around video capture, which obviously serves this new lens well. The 28-105 DN has basically all of the features currently available on modern lenses save lens-based image stabilization.
That starts with the aperture ring (something relatively new for zoom lenses, period). The aperture ring works great, giving you precise control at 1/3rd stops when in clicked mode, and gliding smoothly through the whole aperture range when declicked.
On the lower left side of the lens is the Click/Declick switch which allows you to choose between those two options.
On the right side of the lens is the iris lock. The iris lock will allow you to either lock into or out of the aperture ring. For those that want to use the aperture ring and not inadvertently switch into camera based aperture control, the lock can be engaged while you are in the aperture ring and keep you there. If you’re not an aperture ring person you can switch the lens into the A position and control it from within the camera. Engaging the iris lock while in A assures that you won’t inadvertently bump into manual aperture control.
The left side of the lens also has an AF/MF switch (always appreciated!) along with one of the two custom/function buttons.
There’s a second function button on the top of the lens for access when shooting in a vertical position.
The function of those buttons (they are redundant, not two separate functions) can be set from within the camera.
Lower down the lens barrel is a lock to prevent zoom creep. This is an externally zooming lens, though the precision in the zoom mechanism and extending barrel means that there is little chance of zoom creep unless you have something specifically rubbing on the zoom ring (like when hiking, for example). The lock will only engage at the wide (28mm) setting.
The inner barrel extends roughly 43mm when zoomed out to the 105mm position, and I’ll note that the inner barrel feels very well engineered and smooth – no wobble.
It is worth noting that Sigma lenses zoom in the opposite direction of Sony lenses (and pretty much every other brand on E-mount). It’s a little annoying, actually.
The zoom ring is the closer of the two rings, with the manual focus ring further out. The manual focus ring is well executed, with wide ribbing and good damping. It makes for a good focus emulation, though as with all mirrorless autofocusing lenses this is focus by wire. Input to the focus ring is routed through the focus motor, and there are no hard stops at minimum focus or infinity.
The overall look of the lens is very much modern Sigma, with a lot of vertical lines and a few different textures.
This flows into the lens hood, which has a lock, a rubberized section for soft touch and a ribbed section for more grip when mounting or removing. Sigma’s lens hoods are nicer than just about any other brand. A padded nylon case is also included.
While I haven’t seen a cutout diagram of the weather sealing points, I do know that Sigma touts a professional grade of weather sealing on this lens. There is a gasket at the lens mount, a water and oil repellant coating on the front element, and internal seals throughout the lens. Sigma calls this a “dust and splash resistant structure”.
The basic dimensions of the lens (in Sony E-mount, as tested) are 87.8mm (3.5″) in diameter and 159.9 (6.3″) in length. It weighs in at 990g (34.9oz) and has an 82mm front filter thread. In many ways this lens has a similar physical size to the recent 28-45mm F1.8, though that lens is an internally zooming design.
To put this in context, the Tamron 35-150mm pictured here is 89.2 x 158 mm (very similar in size) but weighs 175g more.
The Canon 24-105mm F2.8L is the behemoth, however, at 88.5 x 199 mm and weighing 310g more. Bottom line is that Sigma has done a good job of managing the size and weight of the lens, though, to be fair, it doesn’t have quite as big of zoom range as either the Canon or the Tamron, and the Canon also has image stabilization while the Tamron sports a larger maximum aperture (F2-2.8) across its zoom range.
The aperture iris is unique for a Sigma lens, in that it has has 12 blades, which helps keep the aperture iris shape circular even with the lens stopped down.
This also produces a more defined 12 bladed sunstar, as while odd bladed apertures actually produce double the sun rays (an 11-bladed aperture has a sunstar with 22 points, but a 12-bladed aperture has 12 points. I personally favor the cleaner look of these sunbursts, though Sigma reports that this is more of a one-off with this lens and they have no plans to move to aperture irises with even blade counts.
The minimum focus distance 40cm, which allows for a very useful 0.32x (1:3:1) magnification at 105mm:
At 28mm, that 40cm produces this result:
But then you can zoom into 105mm and get a very different level of magnification.
This combined with that F2.8 aperture allows for very nicely blurred backgrounds, and up close performance is quite good.
Because I review a lot of Sigma lenses, there’s a lot that is familiar here, which makes it easy to take their excellence in design for granted. Sigma is packing as many features into their top lenses as anyone out there, and that’s fantastic in a lens that is equally good for both video and stills. This is a big lens, but as we’ve seen, it is actually a bit smaller and lighter than the competition. And, at a price of $1499 USD ($2099 here in Canada), it is not exorbitantly expensive (the Tamron 35-150 VXD runs $300-400 USD more).
Autofocus and Video
This is another area where Sigma has made some key strides. Their initial offerings on Sony (and L-mount) featured STM focus motors, but Sigma has since released their more powerful, smoother HLA focus motor (High speed Linear Actuator). The HLA motor has much more torque/thrust than lenses equipped with the stepping motor, which is very important when you are talking about a lens with larger, heavier optical glass like this one. The increased thrust of the focus motor means that it is capable of moving much faster and is thus much more reactive. Real world focus changes are essentially instantaneous, making this lens capable even of keeping up with some action. I used it with success at a rodeo, and while the lighting conditions meant that I didn’t always get as fast of shutter speeds as I would like, I was able to successfully track and capture the action.
This further increases the credentials of the 28-105 DN as a great choice for things like events or weddings, as even if there is some action, the autofocus can keep up.
Focus sound is basically nonexistent. The HLA motor is essentially silent even if I put my ear next to the lens. It is quiet enough that I literally looked up at the screen to make sure the lens was actually focusing.
Focus accuracy is good, too. I don’t really recall any shots during my review having been missed due to poor focus in any reasonable situation (I had a few times during the rodeo that focus grabbed the rail in front of me rather than the subject beyond, but that’s understandable).
Eye AF was excellent, delivering very accurate results whether photographing an animal…
…or a human subject.
I also saw mostly good results for video work. Autofocus pulls were very fast and confident. No hunting or settling. Focus is fast enough that you’ll probably want to slow it down in camera if you want more cinematic focus transitions. There is some focus breathing, though not extreme.
My hand test where I alternately blocked the camera’s view of my face with my hand and then remove it went quite well, with confident adjustments to and from my face.
I also found that in real world shots that focus changes tended to be more abrupt than what I would prefer. There’s a lot of thrust in this focus motor, but it probably would have been better to detune it a bit in video mode…particularly in a lens that will probably be highly desired for video work.
On a positive note, I used the 28-105 DN for about a half dozen of my YouTube video episodes, and focus was rock solid during those episodes.
Sigma’s HLA focus motors are great, though they are slightly limited on Sony by the fact that Sony limits bursts to 15FPS with third party lenses, so this one area where the 28-105 DN is not going to be as effective as the G Master zooms in certain applications.
Image Quality Breakdown
As I noted in my intro, this is a pretty special lens optically. The optical formula is fairly complex (18 elements in 13 groups), and that includes 2 FLD elements, 2 SLD elements, and 5 aspherical elements. A look at the MTF charts for the various lenses shows that the 28-105 DN holds up well when compared with either the 24-70mm F2.8 DN ART or older DSLR era 24-105mm F4 DG HSM lens.
At it’s best, the 28-105 DN is able to produce images with a lot of detail, contrast, and pop.
That’s not to say that the lens is flawless, however, as it does exhibit a few optical weaknesses. This is a very complex undertaking, and it becomes clear where the compromises were made. We not only maintain a constant F2.8 aperture (for full frame) in this zoom, but we go from moderately wide at 28mm to moderate telephoto at 105mm. Here’s a look at the same scene at all five marked positions on the lens (28, 35, 50, 70, and 105mm):
First, some of the good. There’s not much evidence of longitudinal chromatic aberrations (LoCA) here, with clean results before and after the plane of focus. There’s basically no evidence of fringing anywhere.
I shot photos of sunlight dancing through the morning dew, and I found all of those bright specular highlights clean of any fringing.
I also saw zero issues with LaCA (lateral chromatic aberrations), which shows up along the edges of the frame. All of the transitions here are very neutral.
While many people will certainly wish that this lens went as wide as 24mm, thus enabling photographers to more realistically use it for wide angle shots. The problem with going to 24mm (as Sigma has dealt with on their 24-70mm lenses) is that it is very difficult to control distortion, and those lenses have dealt with quite heavy barrel distortion at 24mm. The choice to only go as wide as 28mm should (in theory) allow for an easier optical design. Unfortunately the big zoom ratio of this lens has resulted in both a lot of distortion and a mildly complex distortion pattern.
I needed a very significant +26 to manually correct the distortion, though as you can see the correction isn’t linear. Getting the lines straight near the center results in the edges being pulled to a pincushion style distortion. Vignette is also very heavy, requiring a +90 to correct, which is just shy of maxing out the correction slider in Lightroom.
Sigma did provide me with a correction profile, and you can see that it essentially splits the difference on the distortion, leaving a bit of barrel distortion to avoid creating the pincushion distortion. It does clean up the vignette nicely.
There is unfortunately still a fairly significant amount of distortion (now pincushion style) on the telephoto end. Vignette is about a stop less (+61 to correct), and I needed a -10 to correct the distortion.
That’s enough distortion that I did notice it in some of my real world shots. If you look at the lines along the left side of this image you will see the curve due to the distortion. I saw this in a number of shots.
Again, the correction profile will take care of this, but it does show the area that had to give in the engineering compromises to make this lens happen.
So how about resolution and contrast? My formal tests are done on the 61MP Sony a7RV.
The MTF suggests that this is a very sharp lens in the center and mid-frame areas, with a typical corner drop-off at wide apertures. Is that what we see? Here’s a look at the test chart from which the crops come.
The MTF shows an accurate reflection of real world performance, as I find that, as suggested, the center looks amazing, mid-frame is very good, but the corners show a significant reduction in sharpness and contrast. Here are crops from across the frame at F2.8 and shown at a 200% magnification.
In most situations this probably won’t be a problem. Shooting at night or in a dimly lit venue it will be rare for this to affect things too much. I found that real world corners didn’t look terrible (this shot is 1/6th of a second, F2.8, ISO 3200).
This shot of fireworks at F2.8, 28mm looks nice and crisp.
The corners are somewhat better by F4, and reasonably good by F5.6, though they never quite rival the performance elsewhere in the frame.
It’s enough for real world shots to look good, however. Detail doesn’t jump off the page in a 100% crop of this shot, but you can see that it looks just fine, however.
Diffraction will set in at F11, and shots with apertures smaller than that will look noticeably softer. Here’s the difference between F11 and the minimum aperture of F22:
Moving on to 35mm finds the center performance largely the same, but I see some improvement in the midframe and the corners, though the corners remain softer by comparison.
Centering seems good in the lens, with consistent results in all four corners.
There’s more sharpness potential in the corners at 35mm, with corner performance looking quite good by F5.6.
If we move on 50mm, we find our strongest performance yet. Sharpness is noticeably higher in the midframe and corners (which look great!), and even the center looks a little better.
You can see that real world shots at 50mm show great detail and contrast.
70mm still looks good, though not as intensely good as 50mm.
At the end of the zoom range I find some give and take relative to 70mm. Some spots (midframe) look a little sharper, while the corners look a little softer and the center looks about the same. Overall I would say that sharpness and contrast are slightly lower at 105mm relative to other places in the zoom range.
Overall I found that sharpness was pretty good at 105mm for real world shots, though contrast isn’t top notch.
Here’s another example.
I did a few comparisons to the Tamron 35-150mm. Comparing the wide ends (28 vs 35mm) showed some give and take between the two lenses in terms of sharpness. The Tamron is sharper in the center and midframe, while the Sigma is very slightly sharper in the corners. At 105mm, the Tamron is a bit sharper all across the frame.
At least in the copies that I compared, I found that the Tamron had a little more capacity for further sharpening as you stopped it down as well. The comparison more strongly favors the Tamron at smaller apertures. The Tamron is also considerably more expensive, though it is impressive how well it holds up in this comparison three years after its release. After years of using the Tamron, I felt like the Sigma wasn’t quite as sharp and contrasty in real world applications.
So how about the bokeh? As with any zoom, that’s going to depend on the focal length. The telephoto end has the largest capacity for blurring out the background. You can see from this morning dew shot that bokeh looks fairly good.
There is some typical geometric deformation near the edges of the frame, and I also see some hints of “onion bokeh” there (concentric rings inside the specular highlights).
But while I found a slight bit of outlining in the specular highlights, I do feel like the some of the hard edges in this image are rendered fairly well. I have no real complaints here.
This casual portrait of me shows a little more busyness than what I would like in the textures of the trees beyond me.
I found the recent 28-45mm F1.8 zoom from Sigma to have some of the “special sauce” in the rendering, but I’m a little less impressed here.
Flare resistance, on the other hand, is excellent. I found very good resistance to veiling or ghosting, and I think the sunbursts look great!
This is an area of definite win over the Tamron, as it has a few issues with flare.
This allowed me to shoot right into the spotlights at the rodeo without concern that the images would be ruined.
Another plus is color rendition. I liked the overall look of images a lot. Skintones looked good, and other colors were rich.
Here’s another:
How about my truck at golden hour?
So while we don’t have a flawless performance here, we do have a very good optical showing. This is a very competent lens across all of its zoom range, which means that Sigma has pulled off something that really only one company has previously. And don’t forget that close focus ability, which allows you to pull off some unique images.
This ability could really expand the usefulness of the 28-105 DN as a wedding photographer’s primary lens, as you could grab detail shots of flowers, rings, and more without having to pull out a secondary macro lens. If you want to see more images, check out the image gallery for the lens here.
Conclusion
Genre busting lenses like the Sigma 28-105mm F2.8 DG DN | ART are rare for one simple reason: they just aren’t easy to make. There are so many constraints on the engineers, as they need to control the size, weight, and cost while also trying to engineer for optical excellence over a larger zoom range than is typical. Sigma is demonstrating their ability as a company to successfully take on exceptional challenges yet again. They were the first on Sony to do an F1.2 prime. The first to do F1.4 on a wide angle lens. The first to release a full frame F1.8 zoom. On and on the list goes, which shows just how far removed we are from the company once known for making budget alternatives to first party lenses. The 28-105 DN is yet another case in point.
Sigma’s HLA autofocus delivers excellent speed, allowing the 28-105 DN to also excel for action packed moments that lesser lenses might miss.
I can see this lens becoming a favorite for those who want a one-lens solution for events or portrait photography, allowing them to switch from wide angle to tighter framing with a twist of the zoom ring. The $1500 price tag seems a lot more reasonable when you consider that for many people it could probably replace both a 24-70 and a 70-200mm lens, particularly with the ability to switch to an APS-C crop on Sony’s higher resolution cameras. I’m not sure that it quite matches the versatility of the Tamron 35-150mm for my personal needs, but there are also several things that the Sigma does better. What a great day to be a photographer, however, when we are spoiled with one exceptional lens options upon another.
Pros:
Genre busting zoom range with a constant F2.8 aperture
Beautifully made
Feature rich
Thorough weather sealing
Ability to declick aperture
HLA focus motor is extremely quiet and has a lot of thrust
Excellent focus accuracy
Extremely sharp lens over most of the frame and zoom range
Keywords: Sigma 28-105mm, DG, DN, ART, Sigma 28-105 F2.8, Sigma 28-105mm, F2.8, Sigma 28-105 Review, #SIGMAEmount, #SIGMA #SIGMA28105mmF28Art, #SIGMAArt, #SIGMADGDN, Full Frame, Review, Sony Alpha 1, Sony a7RV, Review, Hands On, Dustin Abbott, Real World, Comparison, Sharpness, Bokeh, Flare Resistance, Autofocus, Image Quality, Sample Images, Video, Photography, let the light in, weathersealing, #letthelightin, DA
DISCLAIMER: This article and description contains affiliate links, which means that if you click on one of the product links, I’ll receive a small commission. As an Amazon Associate I earn from qualifying purchases.
Earlier this year I was doing my review of the Canon EOS R50 compact APS-C mirrorless camera. While I liked many things about the camera itself, I couldn’t help lament the state of the lens options for Canon’s RF-S mount. At the time of the review (two years since the release of the EOS R7, the first of Canon’s RF mount APS-C cameras), this was the sum total of the available lenses.
Yuck. I couldn’t imagine a less exciting lineup of lenses. The fastest maximum aperture in the bunch was F3.5, and that lens is F4 and smaller before you make it to 30mm. Canon clearly knew they had a problem, for between the time that I filmed my review and the time I released it, they announced that certain Sigma (and later Tamron) lenses would become available for RF-S (the APS-C version of Canon RF). Now, to be clear, this is only for APS-C lenses, but for now it seems like Canon is going to let third parties solve their APS-C lens problem. This is great news for consumers, however, because it means that lenses like the Sigma 18-50mm F2.8 DC DN are now available for Canon RF. I love the Sigma 18-50mm, having first reviewed it when it released on Sony back in 2021 and then last year when it became available on Fuji X-mount. I found it useful enough on Fuji that I purchased one myself, and have happily used it since. So how well has the 18-50mm survived the transition to Canon. You can find out my thoughts in the video review below…or read on.
Thanks to Gentec (Sigma’s Canadian Distributor), for sending me a review loaner of this lens. As always, this is a completely independent review. All opinions and conclusions are my own.*The tests and most of the photos that I share as a part of my review cycle have been done with the 45MP Canon EOS R5, which I reviewed here.
The good news is that this is a great porting of the lens. It’s pretty clear here that this is no “reverse engineering” but rather a direct license of focus algorithms, as this is not only the best focus experience of the now three platforms that I’ve tested the lens on, but also one of the better focus experiences I’ve had on Canon in general amongst lenses priced under $1000.
This review of the 18-50mm RF is not a whole new review, but rather an update on how the 18-50mm RF translates to Canon RF. I’ve already tested the image quality at both a 24MP (Sony in 2021) and 40MP (Fuji in 2023)) level, so that covers the full gambit (and more) of what’s available on RF bodies. There are no optical surprises here, but I’m always curious to see how the lens changes in terms of its dimensions, but, most importantly, how well it autofocuses on a platform that it wasn’t originally designed for.
The 18-50mm RF is one of six lenses released by Sigma for Canon RF, and those six lenses represent all of the APS-C mirrorless lenses they’ve released over the past three years. You can see my review of the original releases on Sony here.
Sigma’s DC DN lenses have all fallen under the umbrella of their Contemporary lineup, and while Sigma has proven willing to experiment with new features and design elements within the ART and Sport lineups, the Contemporary lenses have all stuck to a fairly rigid design formula. That continues to be the case here, so this little zoom mostly looks and handles pretty similarly to other lenses in the lineup.
There’s nothing new here on the RF mount, but the lens is more of a standout on Canon due to having essentially zero quality competition. The 18-50mm RF is a genuine treat as an addition here. Why? Here’s a few reasons:
Has at least a weather sealing gasket
Better build quality than RF-S lenses
Includes a lens hood
Constant F2.8 aperture
Good up close performance
Great autofocus on RF
In essence, the poor competition makes this lens a star, though frankly I still found it one of the more compelling options on these other platforms as well.
The crop factor on Canon is different than Sony or Fuji, so the zoom range works a little different. Fuji/Sony have a crop factor of 1.5x, which means that the lens goes from a full frame equivalent of 27-75mm. Canon’s APS-C crop is 1.6x, and so that changes the effective focal range to 29-80mm. You’ll miss the extra bit of width at 18mm:
…but appreciate the extra bit of reach on the telephoto end.
Sigma 18-50mm RF-Mount Build and Handling
There’s really only one significant change to the 18-50mm on Canon, and that is that the lens mount portion of the lens has to be flared out to accommodate the much larger diameter of the RF mount. You can see that difference when compared to the Fuji X-mount version on the right.
As this is now the widest portion of the lens, the lens diameter changes from 61.6mm (2.4″) to about 68mm by my measurement. I found the RF version to be just a hair shorter than the X-mount version, measuring more like 75mm than the 76.8mm of the X-mount (3″). That extra bit of girth near the lens mount also adds a tiny bit of weight, with the RF version weighing 294g (10.37oz) on my scale, which is about 10g heavier. I haven’t seen the dimensions and weight for the RF mount published by Sigma, so I’m using my own measurements.
Other than these minor physical changes to the RF version, the build is otherwise identical to either the E-mount of X-mount versions.
The 18-50mm F2.8 is a mixture of metals and “thermal composites” (high grade engineered plastics) in construction, and the construction feels quality in the hand. It is built around a metal mount.
The overall build quality just feels much higher than any of the Canon RF-S lenses that I’ve tested so far. Some of them even have plastic lens mounts. This feels like a quality lens.
There is a weather sealing gasket at the lens mount (thicker here on the RF mount than on previous versions), but, like other Contemporary lenses, there are no internal seal points.
The only thing on the barrel is the focus and zoom rings. The closest to the lens mount is is the focus ring, which is fairly narrow (10mm), ribbed, and moves smoothly. It is has light damping and doesn’t have a lot of feel, and is actually slightly compromised here by the thicker lens mount flare, which clearly wasn’t part of the original design. It crowds the focus ring a bit, and I found my thumb slightly pushed off the ring.
The second ring is the wider (20mm) of the two, and it is the zoom ring. The zoom action is very smooth, with a single inner barrel that extends about 25mm. The inner barrel is well engineered and without any wobble.
Also included is a fairly shallow petal-shaped lens hood. It has a ribbed texture section to help you grip it for removal and will reverse for storage.
There are seven rounded aperture blades inside, and these do a reasonably good job of maintaining a circular shape with the lens stopped down.
A look at the front of the lens shows a 55mm front filter thread along with a significant amount of glass inside; this is the huge advantage over the Canon kits lenses with a much smaller maximum aperture value.
Like many recent zoom lenses, there is a separate minimum focus distance for wide angle (18mm) and telephoto (50mm), with interim focal lengths being on a sliding scale between them. The highest magnification (as per usual) is at the 18mm end, where you can focus as close as 121mm. Only one problem – when the lens is mounted on the camera, the distance from the sensor to the end of the lens (without the lens hood) is 95mm, leaving only 26mm of working distance between you and your subject. Here’s what that looks like:
It is basically unusable, as even without the lens hood it is almost impossible for the lens to not be shading your subject. If you add the lens hood, you actually exceed that minimum focus distance. So, in theory, you can achieve a maximum magnification of nearly 0.36x, but you are unlikely to ever achieve that in the field. Here’s what that looks like if you were able to achieve it:
You’ll see a lot of field curvature and a very small sweet spot of focus/sharpness. My advice: back up a little and get much nicer results with more like magnification in the high 20s instead.
The telephoto end doesn’t market as well, but is much more useful. You can focus as closely 30cm there, but the resulting magnification of 0.20x is much more easily attainable.
The wide open sharpness and contrast there is not off the charts, but is good enough to produce nice real-world results like this:
There is no optical stabilization on the lens, unfortunately. I happened to test on a camera that had in body image stabilization, so I didn’t miss it, but many of Canon’s smaller APS-C cameras don’t have IBIS.
As noted, on other platforms the Sigma is a rather plain lens in terms of features and build. Because of so little competition on Canon, however, it feels next level in terms of build and handling. It’s a nicely made little lens that follows a familiar, well-executed Sigma formula.
Autofocus and Video
Sigma continues to utilize a stepping focus motor (STM) in these smaller lenses as the elements are smaller and lighter than lenses and don’t require the higher torque that some of their large aperture full frame or sports oriented lenses. Autofocus is something that Canon does very, very well, and it is clear to me that Sigma has received access to the focus protocols and algorithms, as the 18-50mm RF definitely handles like a first party lens.
The STM motor provides fast, quiet, and accurate autofocus. Eye detection works well, and I was able to get nice “people” photos with quick and easy autofocus.
Autofocus speed was excellent, near instantaneous both indoors and outdoors, and utterly reliable and accurate.
I got well focused results when I took some photos of Nala (working hard).
Even without a trackable subject in frame I was able to get good focus. Here’s some foliage in the evening light.
I’ve got zero complaints about the autofocus for stills; I got a lot of well focused results without any drama. This is a seamless transition to Canon RF.
On the video front my findings were also excellent. My autofocus pulls tests were smooth and confident, with none of the settling or pulsing that I saw on Fuji.
On a positive note focus breathing is fairly low.
My “hand test” (where I block the view of my face with my hand and then move the hand to allow focus to pop back to the eye) also went well. Focus transitioned confidently from my hand to my test and vice versa.
Unlike on Fuji, I also saw smooth focus results when zooming in and out while recording video. There was none of the shattering effect and focus readjustments, which shows again how far behind Fuji’s autofocus is.
This is a great lens all around when it comes to autofocus; a fantastic performance on Canon.
Sigma 18-50X Optical Performance
Sigma has given the 18-50mm F2.8 DN an optical design of 13 elements in 10 groups, and this includes 1 SLD element and 3 Aspherical elements. The MTF charts show a good performance in the center of the frame, fairly strong mid-frame results, and a dip into the corners.
In this section I will be reusing my results from the 40MP Fuji X-Trans sensor. It is higher resolution than anything Canon currently makes (current resolution points are between 24 – 33MP. Fuji’s images are also harder to sharpen than Canon, so you will have no problem getting beautiful results on Canon bodies. It has no problem producing beautiful images.
We’ll start by taking a look at distortion and vignette. Distortion is essentially identical to what I saw on Sony, with some strong barrel distortion at 18mm.
While Sigma lenses do receive good correction profile support on Canon, I turn off those corrections and do a manual correction so that I can see what’s being compensated for. I used a +24 to correct for the barrel distortion, but you can see that there is a mild mustache pattern left over.
Expect less vignette on Canon than what I saw on Fuji, with results more in line with what I saw on Sony. I had to use a +52 to correction vignette at 18mm on Sony. On Fuji? A whopping +92! The good news is that the profile correction does a good job with both distortion and vignette. This F2.8 image has received no addition correction, but you can see that there is no visible shading on the snow.
As you progress through the zoom range the distortion starts to invert and eventually becomes a pincushion style distortion that is more strongly pronounced at 50mm.
I needed to dial in a -11 to correct the pincushion distortion, though it was nice and linear and cleaned up well. Vignette remains heavy, however, requiring a +88 to correction (again, about +40 more than on Sony).
There are longitudinal chromatic aberrations (LoCA), but they aren’t particularly strong. I mostly see them in the form of a blue-green fringing after the plane of focus:
You will see a light fringing around specular highlights as well:
Likewise there is a bit of lateral chromatic aberrations (LaCA) that show up near the edges of the frame. You can see them in the transitions from black to white.
The LaCA is easy to clear up by clicking the “Remove Chromatic Aberrations” button in software or in camera for JPEGs or video.
Chromatic aberrations are a little more visible at a pixel level on high resolution cameras for the simple reason that the fringing occupies more pixels and gets more highly magnified at a pixel level, but they aren’t any more obvious when viewing the image as a whole.
So how about the main event? High resolution cameras also expose any lack of contrast or a lens’ inability to resolve fine details because those flaws are more highly magnified. Here’s a look at our test chart.
Here is a closer look at roughly 180% crops from the center, mid-frame, and lower right corner at 18mm, F2.8:
The center looks great, the mid-frame acceptable sharp, but the corner looks quite low contrast. Stopping down to F4 provides a little more contrast and detail, and F5.6 looks better still, but the corners never get exceptionally sharp.
Real world images look great in the center and mid-frame, but the corners don’t look as crisp.
F8 is a lateral move (not really any sharper than F5.6) due to diffraction starting to soften the image (diffraction comes very early on a high resolution body). After F8 the image will show more obvious softening, and by the minimum aperture of F22 the image looks very soft in my tests.
At 24mm the image shows improvement across the frame, but in particularly in the corners, which are radically sharper.
It’s easier to get sharp landscape images at 24mm due to the more consistent sharpness profile.
28mm is largely the same as 24mm, with a very consistent sharpness profile across the frame.
At 35mm there is a dip in contrast relative to the high water mark of 24-30mm.
Sharpness and contrast do improve if you stop the lens down, however, and by landscape apertures like F5.6 the results look good.
Things rebound a bit at 50mm, and, if they aren’t the peak of sharpness like 24mm, the results look quite good even at F2.8.
When stopped down the lens can produce very nice landscape images with strong contrast.
Here’s another 50mm that looks nice at F2.8, with a nice depth of field that highlights the subject. Contrast at a pixel level isn’t incredibly high, but viewed normally the image looks great.
The quality of the bokeh is mostly good. Zoom lenses like this rarely create the most creamy bokeh, and the Sigma 18-50mm is no exception. At the same time, however, I felt like images looked pretty good. The bokeh here is actually really nice, particularly in the top half of the image.
This shot shows a bit of outlining in some of the defocused textures, which does detract a bit from the bokeh.
This shot is somewhere in between, but overall I feel like it looks pretty nice.
One final example at a little further distance. In some ways this image is a little more impressive, as background isn’t quite as out of focus. The 18-50mm avoids any nervous looking textures in the transition zones, and there is a reasonable amount of 3D pop on the subject.
The look of sunbursts is nice but nothing special. The blades don’t have the distinction of straight blades, but it is nice enough to add a little extra touch to images.
Flare resistance was also fairly good, with only minimal ghosting artifacts showing up in real world images.
Contrast remained strong, however with only a bit of a flare pattern showing in some images.
I was able to test for astrophotography during my review of the E-mount version of the lens, and found a bit of coma smear in the corners along with a bit of generalized fringing on the brightest points of light.
I was able to get nice looking Milky Way images with a lot of asteroid action, too.
Before closing this section, I think it is important to put the Sigma’s overall performance into perspective. The Tamron 17-70mm F2.8 is much larger, and yet it doesn’t really hold up better on the Fuji 40MP sensor. When I compared the two on the wide end, the Sigma clearly had the sharper center of the frame along with a slightly better mid-frame, though the Tamron showed a little better in the corner.
On the telephoto end there is some give and take, with the Sigma looking slightly sharper in the center and the Tamron looking a bit sharper in the corner. The Sigma’s mid-frame is clearly better, however, with more contrast and detail visible.
The fact that the Sigma delivers slightly more sharpness on the 40MP despite being so compact is very impressive. Here’s a gallery of a few more images taken on Canon RF.
Conclusion
I was very encouraged by what I saw from the Sigma 18-50mm F2.8 DC DN on Canon RF mount. It feels like a seamless port over other than the slightly odd lens shape due to the much larger RF mount. Autofocus was fantastic, image quality looks great, and the lens remains very impressive despite its small size.
This is a great little general purpose/travel lens, as it affords enough flexibility in the focal length to shoot many scenes, and the constant maximum aperture of F2.8 gives you more flexibility in different lighting situations.
It’s great to see Canon’s RF mount starting to open to these third party offerings, and very quickly we will see the amount of lenses available for RF-S (APS-C) tripled due to Canon’s move to open the protocols. This lens isn’t cheap, per se, at $599 USD, but I’d rather spend a few hundred extra bucks to buy this lens over the current Canon options any day of the week.
Pros:
Fantastic job porting to Canon
Extremely compact and lightweight
Nicely made with a weather sealing gasket
Useful focal range with constant F2.8 aperture
Great autofocus performance
Good video AF performance
Minimal focus breathing
Good sharpness across zoom range
Nice bokeh in most situations
Fairly good coma performance
Nice build and handling than RF lenses
Cons:
No optical stabilizer
Fairly strong barrel distortion at 18mm
Some ghosting in certain lighting conditions
Flare for RF lens mount slightly interferes with MF ring
DISCLAIMER: This article and description contains affiliate links, which means that if you click on one of the product links, I’ll receive a small commission. As an Amazon Associate I earn from qualifying purchases.
Sigma’s 18-35mm F1.8 that was released nearly ten years ago became a cult favorite, particularly with the video crowd. Even full frame users often used it to help mitigate crop factors and because of the flexibility of having a zoom range AND a fast maximum aperture. Sigma followed that up with a similar full frame lens – the Sigma 24-35mm F2 – but that lens never seemed to capture the same kind of attention (perhaps due to its large size and abbreviated zoom range). But Sigma is back for the mirrorless era with a unique new offering that surely hopes to recapture the success of the 18-35mm. This is a full frame lens, but Sigma has tweaked the formula to allow for a slightly larger zoom and a little brighter a maximum aperture. We now have the 28-45mm F1.8 DG DN | ART lens…but does it have the magic of the 18-35mm? We’ll explore that question in either my video review below or in the text review that follows.
Thanks to Sigma Canada (Gentec) for sending me a review loaner of this lens. As always, this is a completely independent review. *The tests and most of the photos that I share as a part of my review cycle have been done with the Sony a7RV along with the Sony Alpha 1 that serve as my benchmark cameras for Sony lenses.
Some of the key selling features of the 28-45 DN include:
World’s first F1.8 full frame zoom
Prime-like optical performance
Full range of video features
Multiple custom buttons
Aperture ring with declick and iris lock options
Internally zooming design
Thorough weather sealing.
Sigma revealed the development of this lens to me a while back, and my immediate thought when I heard the specifications was, “a full frame version of the 18-35mm!”. I saw their marketing materials about a month later, and that was exactly how they are marketing it. The numbers aren’t exact, but since the 18-35mm was an APS-C lens, if you apply a 1.5x crop factor (the most common APS-C crop factor) to that focal length, you get 27-52.5mm. So obviously the zoom range is a little more constrained on the full frame version, though definitely a bigger zoom ratio than the 24-35mm. Sigma cheated a bit on the wide end (-4mm) to allow more on the telephoto end (+10mm) and to avoid making the lens even larger. And that size factor is going to be, in my mind, the biggest potential deterrent. This is a BIG lens – over 153mm in length and weighing 950g – 33mm longer and 140g heavier than the 18-35mm. That’s not terrible in absolute terms (considering this is a full frame lens), but it does mean that this will be slightly harder lens to gimbal.
The whole “bag of primes” in one lens applies here, though I will practically point out that the Samyang 24mm F1.8 (230g), 35mm F1.8 (210g), 45mm F1.8 (162g) weigh a combined 602g. You could throw in the 75mm F1.8 (230g) and still undercut the weight of the 28-45mm by 150g! That’s not to suggest that those lenses are of the same quality as the Sigma or that having 3-4 prime lenses are as convenient for video making, but the size and weight will be probably the single greatest deterrent for some potential buyers. The better argument here is the image quality and overall performance, which is nothing short of exceptional.
Sigma has a well deserved reputation for producing sharp lenses, but every now and then they make something exceptional – something that stands out from the ordinary lenses that I review from them and is optically special. The 28-45 DN is one of those. Here’s why…
Build and Handling
Sigma’s build standards are continually evolving, and they have been very proactive about adding new features and improvements. Many of those feature improvements have been designed around video capture, which obviously serves this new lens well. The 18-35mm was well regarded, but the truth of the matter is that the only “feature” it had was an AF/MF switch. It didn’t have weather sealing (neither did the 24-35mm F2), and it didn’t have any kind of lens-based stabilization. The latter is still true, but the difference now is that almost all of the Sony full frame mirrorless cameras have camera based stabilization, as do a number of the L-mount cameras this lens is also compatible with. Outside of that, however, the 28-45 DN has basically all of the features currently available on modern lenses.
That starts with the aperture ring (something relatively new for zoom lenses, period). The aperture ring works great, giving you precise control at 1/3rd stops when in clicked mode, and gliding smoothly through the whole aperture range when declicked.
On the lower left side of the lens is the Click/Declick switch which allows you to choose between those two options.
On the right side of the lens is the iris lock. The iris lock will allow you to either lock into or out of the aperture ring. For those that want to use the aperture ring and not inadvertently switch into camera based aperture control, the lock can be engaged while you are in the aperture ring and keep you there. If you’re not an aperture ring person you can switch the lens into the A position and control it from within the camera. Engaging the iris lock while in A assures that you won’t inadvertently bump into manual aperture control.
The left side of the lens also has an AF/MF switch (always appreciated!), and in a slight redesign for Sigma, the two custom/function buttons are moved away from the AF/MF switch to a further section of the barrel that just the two of them occupy. One is on the left side for the traditional horizontal use of the camera.
There’s a second function button on the top of the lens for access when shooting in a vertical position.
The function of those buttons (they are redundant, not two separate functions) can be set from within the camera.
This is an internally zooming lens, so there is no need for a zoom lock. Internally zooming lenses tend to be larger than externally zooming lenses for obvious reasons, but there are a number of key tradeoffs for that extra size.
The lens length remains constant, making it easier to use on a gimbal (the weight balance doesn’t change)
There is no risk of zoom creep, as there is nothing external to be affected by gravity.
The zoom action is smoother and more precise
There is less risk of getting dust or moisture in the lens because nothing is going in and out.
The handling is better
The zoom ring is the closer of the two rings, with the manual focus ring further out. The manual focus ring is well executed, with wide ribbing and good damping. It makes for a good focus emulation, though as with all mirrorless autofocusing lenses this is focus by wire. Input to the focus ring is routed through the focus motor, and there are no hard stops at minimum focus or infinity.
The overall look of the lens is very much modern Sigma, with a lot of vertical lines and a few different textures.
This flows into the lens hood, which has a lock, a rubberized section for soft touch and a ribbed section for more grip when mounting or removing. Sigma’s lens hoods are nicer than just about any other brand. A padded nylon case is also included.
While I haven’t seen a cutout diagram of the weather sealing points, I do know that Sigma touts a professional grade of weather sealing on it. There is a gasket at the lens mount, a coating on the front element, and internal seals throughout the lens. The internal zoom and focus design further helps with this.
The basic dimensions of the lens are 87.8mm (3.5″) in diameter and 153.4 (6″) in length. It weighs in at 950g (33.5oz) and has an 82mm front filter thread.
The aperture iris has 11 blades, which helps keep the aperture iris shape circular even with the lens stopped down.
The minimum focus distance of the lens is 30cm (11.9″), and the maximum magnification figure is achieved at 45mm and is a very useful 0.25x (1:4).
This combined with that F1.8 aperture allows for very nicely blurred backgrounds, and up close performance is quite good, though a little softer than if you back up another foot.
Because I review a lot of Sigma lenses, there’s a lot that is familiar here, which makes it easy to take their excellence in design for granted. Sigma is packing as many features into their top lenses as anyone out there, and that’s fantastic in a lens that is equally good for both video and stills. This is a big lens, but it’s also doing something no other full frame zoom has done before, so I feel like we can make some allowances here. And, at a price of $1349 USD ($1899 here in Canada), it is not exorbitantly expensive.
Autofocus and Video
This is another area where Sigma has made some key strides. Their initial offerings on Sony (and L-mount) featured STM focus motors, but Sigma has since released their more powerful, smoother HLA focus motor (High speed Linear Actuator). The HLA motor has much more torque/thrust than lenses equipped with the stepping motor, which is very important when you are talking about a lens with larger, heavier optical glass like this one. The increased thrust of the focus motor means that it is capable of moving much faster and is thus much more reactive. Real world focus changes are essentially instantaneous, making this lens capable even of keeping up with some action.
That makes this a great option for someone wanting to shoot events or weddings, as you have some flexibility in the focal length (unlike a prime) but have the light gathering potential of a prime, allowing you to catch those critical moments.
Focus sound is basically nonexistent. The HLA motor is essentially silent even if I put my ear next to the lens. It is quiet enough that I literally looked up at the screen to make sure the lens was actually focusing.
Focus accuracy is good, too. I don’t really recall any shots during my review having been missed due to poor focus.
I also saw mostly good results for video work. Autofocus pulls were very fast and confident. No hunting or settling. Focus is fast enough that you’ll probably want to slow it down in camera if you want more cinematic focus transitions. There is some focus breathing, though not extreme. The lens is not quite parfocal, either, but I did find that I could zoom in and out with focus turned off and the results were acceptably good.
My hand test where I alternately blocked the camera’s view of my face with my hand and then remove it went reasonably well, though there were times where focus wanted to stick on my face even when my hand blocked the view of my face. This degree of stickiness is admirable in some ways, though it wasn’t really what I was looking for.
I also found that in real world shots that focus changes tended to be more abrupt than what I would prefer. There’s a lot of thrust in this focus motor, but it probably would have been better to detune it a bit in video mode…particularly in a lens that will probably be highly desired for video work.
On a positive note, I used the 28-45 DN for several of my YouTube video episodes, and focus was rock solid during those episodes.
Sigma’s HLA focus motors are great, though they are slightly limited on Sony by the fact that Sony limits bursts to 15FPS with third party lenses, so this one area where the 28-45 DN is less competitive with Sony G-Master lenses.
Image Quality Breakdown
As I noted in my intro, this is a pretty special lens optically. The optical formula is fairly complex (18 elements in 15 groups), and that includes 5 SLD elements along with 3 aspherical elements. You know that this is a high performing lens when Sigma actually favorably compares it to two of their recent primes – the 24mm F1.4 DN | ART and the 50mm F1.4 DN | ART lenses. Now, technically, the 28-45 DN doesn’t cover either one of those focal lengths, but it does illustrate how that this zoom is sharper across a good part of the frame than either one of these prime lenses.
As I noted, however, Sigma has long been able to make sharp lenses. What stood out more to me with the 28-45 DN is that it has that special sauce where the colors, contrast, sharpness, and bokeh come together to produce a lens that makes very special images. This image was one that really stood out to me.
On its face it is a simple image. I was at a gender reveal party for our niece and wanted to capture the gift bag that my wife had put together for the baby. It’s literally just a picture of a fuzzy tote sitting on a table in a restaurant – not glamorous at all – but when I looked at the image I was surprised by how great it looked. That’s the magic of a very good lens; it makes everything look just a little more special. I’ve seen that with a few Sigma lenses along the way. Their 28mm F1.4 ART and 40mm F1.4 ART lenses were like that, and, while their 85mm F1.4 DN does have a few flaws, I instinctively just like the images that it makes.
Sigma’s recent 24-70mm F2.8 DN II zoom had a few more longitudinal chromatic aberrations (LoCA) than what I expected, but that’s not the case here. There’s basically no evidence of fringing anywhere.
I shot photos of sunlight dancing through the morning dew, and I found all of those bright specular highlights clean of any fringing.
I also saw zero issues with LaCA (lateral chromatic aberrations), which shows up along the edges of the frame. All of the transitions here are very neutral.
Sigma wisely chose to narrow the focal range to 28mm, which is much easier to engineer for than attempting to go as wide as 24mm. That means that there is much less distortion here than on, say, the 24-70mm F2.8 II, and what distortion is there is more linear and easily correctable.
I used a simple +8 to correct the distortion and a +45 to correct less than two stops of vignette in the corners.
At 45mm the distortion has inverted to a pincushion style distortion that requires just slightly more correction (-9) but again corrects in a nicely linear fashion. Vignette is very slightly heavier at a +60 to correct, so a full two stops.
That’s nothing significant, so, while the lens is big and the aperture is wide, Sigma has managed to slay some of the common dragons that affect zoom lenses.
So how about resolution and contrast? My formal tests are done on the 61MP Sony a7RV.
The MTF suggests that this is a very sharp lens in the center and mid-frame areas, with a typical corner drop-off at wide apertures. Is that what we see? Here’s a look at the test chart from which the crops come.
The MTF spoke truly, as when I examine results I see impeccable results in the center and mid-frame at 28mm, F1.8, with the corners looking obviously softer.
Other than the extreme corners, there is plenty of resolution here to produce very crisp real world results even at F1.8 (and on a high resolution camera!)
The very mild stopdown to F2 does produce a bit more sharpness and contrast, though the effect is felt more in the areas already stronger than in the corners.
The corners start to really sharpen by F4, and now sharpness is impeccable across the frame.
This lens is great at 28mm, whether shooting shallow depth of field shots at F1.8:
Or landscape style shots (here at F2.8):
The lens is sharp enough that it can handle a bit of diffraction at F11, but the image is softer at F16 (though still not bad).
Moving on to 35mm shows an even stronger performance everywhere I look in the frame. Better detail, better contrast. Look at how much crisper this midframe shot at F1.8 looks (35mm result on the right).
The corners are still the weakest spot in the frame, but they are good enough that it isn’t really a factor. Landscape images at 35mm simply “pop”!
40mm is very slightly less contrasty than 35mm (more similar to 28mm), though it is still exceptionally good. Here’s a look at another point in the frame, where you can see slightly less contrast in the 40mm result (on the right).
Remember that this is relative, however. This is still a very sharp lens at 40mm…even at F1.8:
The very important 45mm end of the zoom range shows perhaps our second best performance of the four marked positions on the zoom range. I think 35mm is still a little better across most of the frame, but 45mm is close, and is actually the best of the four positions in the corners.
The lens that I tested showed a good degree of centering, with equal performances in all four corners.
Stop down to F2.8 and the lens is absolutely brilliant across the frame. You can see a ridiculous level of contrast in the midframe.
Landscapes are beautiful. Great color and contrast, fantastic detail.
Shooting at 45mm is also going to give you the most dramatic combination of sharpness and bokeh. This is a lens capable of a very nice ratio between sharpness and softness.
In general I think the bokeh is very nice. There’s a very nice falloff from the plane of focus to defocus, and images just look great to my eye.
Even if I choose a subject a little further away, the look of the bokeh is still quite nice.
Flare resistance also seems very solid. This morning shot has the sun right out of frame, and you can see that contrast has held up well.
In this shot the sun is right out of the frame on the right, but you can see there is no glaring into the frame.
Shooting right into the sun works fine as well.
The bottom line is that is a pretty exceptional lens optically. Images just look great to me regardless of the setting I’ve used the lens in. I think people photos look great!
General purpose shots look great:
And up close photos look great as well.
The Sigma 28-45 DN is big and heavy, yes, but it is also optically special. Not only has Sigma done something no one has ever done before in creating a full frame F1.8 zoom; they have done it well! If you want to see more images, check out the image gallery for the lens here.
Conclusion
Every now and then a company surprises you. You would think the surprise here is seeing Sigma do a full frame F1.8 zoom, but frankly I’m not overly surprised by that. We are reaching a place of such saturation on Sony E-mount that it seems like many of the companies are starting to experiment with more extreme designs. They essentially have to release more unique lenses because all of the standard lenses already exist. The Sigma 28-45mm F1.8 DG DN | ART is now the most extreme zoom lens on Sony opr L-mount due to the constant F1.8 focal length, though one could argue that Canon’s 28-70mm F2L still qualifies as more extreme due to have a much larger zoom range and an only slightly smaller maximum aperture. But let’s not forget that the Canon weighs 500g more and costs well over twice as much ($3000).
But what really surprised me is how good optically the 28-45 DN was. It’s undeniably a Sigma, but an extra special Sigma. I’ve had a few of those moments in my 10+ years of covering Sigma lenses where I pleasantly surprised by my expectations being exceeded. It’s like suddenly discovering a Zeiss lens where I expected a standard Sigma offering. If Zeiss made an Otus zoom, it feels like it would be a lot like this…and that’s a very high compliment.
The size and weight of this lens means that it won’t be for everyone, and maybe it won’t achieve the cult status of the 18-35mm, but I can say with complete confidence that this is a MUCH better lens than the 18-35mm was. I can also appreciate the fact that the market is much more competitive, too, so Sigma will have a harder time standing out from the crowd.
But if you choose to spend the $1350 USD to purchase the Sigma 28-45mm F1.8 DN, know that you are getting an amazing lens for the money. It is a beautifully made, feature rich, has a high end focus system, is incredibly sharp, and has really gorgeous bokeh and rendering. Oh, and it is the world’s first F1.8 full frame zoom, so it has that going for it too.
Pros:
The first of its kind – a full frame F1.8 zoom
Beautifully made
Feature rich
Internal zoom and focus makes for great handling
Thorough weather sealing
Ability to declick aperture
HLA focus motor is extremely quiet and has a lot of thrust
Keywords: Sigma 28-45mm, DG, DN, ART, Sigma 28-45 F1.8, Sigma 28-45, F1.8, Sigma 28-45 Review, #SIGMAEmount, #SIGMA #SIGMA2845mmF18Art, #SIGMAArt, #SIGMADGDN, Full Frame, Review, Sony Alpha 1, Sony a7RV, Review, Hands On, Dustin Abbott, Real World, Comparison, Sharpness, Bokeh, Flare Resistance, Autofocus, Image Quality, Sample Images, Video, Photography, let the light in, weathersealing, #letthelightin, DA
DISCLAIMER: This article and description contains affiliate links, which means that if you click on one of the product links, I’ll receive a small commission. As an Amazon Associate I earn from qualifying purchases.
Sigma 17-40mm F1.8 DC | ART Gallery 
Sigma 17-40mm F1.8 DC | ART Review 
Viltrox Pro AF 85mm F1.4 FE Gallery 
Viltrox AF 85mm F1.4 PRO FE Review 
