The Sigma 20mm F2 DG DN is the newest prime lens in Sigma's series of relatively fast, compact primes for E-mount and L-mount. Join Chris and Jordan as they test it in the scenic town of Kimberley, British Columbia, to show us how it performs.
I could test it at the weekend. The sharpness is great but it is no 20mm lens. Sigma says that the angle of view is 94,5° but that's only on paper. With the big distortion correction in the profile you have only a 22-23mm lens. I checked it against a 21mm with 92° M-lens on a SL 601. If you switch off the manufacture profile in the raw-converter you get the full aov but with a distortion from nearly "a fisheye". So if you have a "really" 24mm it is a mixed package.
Nice lens and these are the types of primes I'm much more interested in vs. bulky f1.4s. I love 20mm lenses (or equivalents). Plenty wide without getting silly wide. Love my little Laowa 10/f2 on M43. These new Sigma DG DN series are making E-mount look even more attractive for this Canon shooter.
I'm planning to ride my FF DSLRs into the sunset for a few more years before jumping onto the FF mirrorless train. Canon...if you're listening...bring a lens like this to RF mount please! Canon hasn't made a decent 20mm prime since...FD days? The ancient EF 20/2.8 is not good. I'm not switching yet but the competition has my attention.
This just reminds me how awesome the 22/f2 is for EF-M. It may not be the sharpest at f2/ but great bokeh and just nice rendering makes this still one of the best walk around lenses ever ever.
Can you do a comparison of Sigma 20mm F1. 4 and this lens. The Sigma 20mm F1. 4 is a stunner for some superb perspective with Bokeh. Want to see how these two compare.
Another lens built for Sony mount by a third party manufacturer. Come on Nikon, why do you have to make it so hard? Many don't understand the logic. Having choices is best for consumers. Sure, one can debate that a Z lens may be better, but at what cost? For many consumers, it's not about the absolute sharpness and color rendition, but about cost and compromise.
I suspect Nikon or any brand for that matter would much prefer you use their own lenses :-) I found that the FTZ worked for all my [ admittedly newer F mount glass } . Also for those not shooting fast action the TZE-01/ 2 allows for very good S-AF to the extent I see Sony FE mount lenses as just another option. For long tele action shooters it is a different ball game
Looks like a decentered copy. Mid to lower right side is smeary. Look at Alpine Garden restaurant shot and his lens chart at f/4 (why doesn't he have any detailed bills or bullseyes on the sides of his chart?). The landscape shots (or housescape shots #2 and #39) that he's taken in landscape mode are also not sharp on the right side. It's almost like Chris is trying to avoid showing the lens' performance in landscape mode. Most distant shots (with equidistant detail) are vertical.
I am not sure that close up test scenes are the best case example for such wide lenses . Their test shots are pretty random so hard to compare across . There are some other reviews around https://www.youtube.com/watch?v=PsqPwNRkT3A
Decentering smears detail on the periphery. It would be hard to see this in closeup shots like #1 because the sides and corners of the image are out of focus. It's easiest to see decentering with a shot of distant trees or distant buildings. Scenes where detail is uniformly in focus across the image plane. If you try to shoot a detailed flat plane up close it's difficult to align the camera and flat plane to maintain perfect edge to edge sharpness. Curvature of field can show up as peripheral softness but that occurs uniformly around the edges and usually disappears on stopping down. Decentering will show up on only one edge and/or corner. It also improves somewhat on stopping down, but that edge/corner never achieves the sharpness of the other three.
Come to canon RF land RF 16mm 2.8 cost less than half of this lens, sigma is 2.5 times expensive to be exact, and much smaller it fit in your pocket while also much wider
R. Cicala really sang the praises of the Sony 20/1.8g in the Lens Rentals blog based on tests with his multi-million dollar testing set-up. Chris is saying that he felt the Sigma and the Sony were optically very similar. Who to go with?
After two cranky/unhelpful/arrogant responses to this legit question, I say it's worth much more inquiry than meh. Sony is probably still the winner at almost the same price point, and a little brighter doesn't hurt either.
If I were them I would have not given them to comparison reviewers. There are already too many comparisons out there with the Sony 20mm which doesn't make the lens look great. [From a marketing perspective]
For us, it is of course awesome. And a company with as much integrity as Sigma, would really hurt their image with such shady tactics
I think they're too honest for their own good and/or the reviewers working off of E mount just have the largest audience anyway, or a combination of both. :P Gordon and Dustin are pretty much my go-tos for non-lab reviews and they (thankfully) love to compare stuff so... Rock meet hard place?
That's on multiple levels wrong. I am talking about geometric distortion not perspective distortion.
-geometric distortion does never change with the subject distance -geometric distortion does sometimes change with focus distance. And this lens seems to exhibit quite a lot of it
Ever wondered why different review sites come up with different distortion measurements? Quite often because their focus distance is different
And they often test at wildly different distances obvs, specially when you get into the lab tests (Lenstip/OptiLimits) vs "real world" tests at infinity (Gordon/Dustin), with all sorts of in betweens (Marc's bookshelf, Dustin's own test scene / dollar/chart).
But to be fair that is less than I expected based on the footage vom DPreviewTV. Maybe it was just very noticeable in that particular scene. There should be lenses that do worse
Panther fan: “I am talking about geometric distortion not perspective distortion.”
I’m also talking about geometric distortion.
“-geometric distortion does never change with the subject distance”
It is you who are wrong despite the great force with which you state the contrary. You can see this to your own satisfaction by stopping down a a wide lens with, say, visible moustache distortion to f/22 and taking a picture with the edge of a close ruler along the bottom of the frame and the sea horizon (or other distant straight line) along the top edge. The two straight lines will exhibit different distortion in the same photograph.
Panther fan: “Ever wondered why different review sites come up with different distortion measurements?”
Chiefly because most reviewers have a rudimentary grasp of optics and are sloppy workers too.
“Quite often because their focus distance is different”
No, but possibly because their targets are at different distances. That’s what counts. Try it.
"The two straight lines will exhibit different distortion in the same photograph." Do you have any example that proves that? Should be physically impossible. As you would be able to look around things. If you hold a ruler perfectly aligned with the horizon, and the camera would distort both differently you would be able to look "through" a non-existent gap.
It is due to a change in focus distance. Distortion profiles are focus distance dependant, not subject distance dependant. (Would lead to some really awkward corrections)
I like your diagram. Can you do another one to explain why it’s possible for the camera to see into the “impossible” region by refocusing the lens? Since you don’t seem to have a problem with that impossibility.
The camera cannot see stuff it is not supposed to with the refocus method. Where should it?
And regarding your image that impressive that you found this. But without knowing the ruler distance I would argue that the ruler is too close. So close, that the lens can see around it. In my diagram the green wall is higher than the midline+aperture on purpose.
Old Sigma 12-24mm DG 4.5-5.6 from the first DSLR era has also distortion variance through the focus range. So it's rare but possible. Don't know, how it would be (or not) corrected with this 2/20mm automatically though.
Panther fan: “The camera cannot see stuff it is not supposed to with the refocus method. Where should it?”
You told me the distortion varies with focusing distance. If it does, the same “impossibility” arises, just after a focus adjustment instead of in a single photo.
I will let you consider why your impossibility is in fact possible.
Nikolai Vassiliev: “Old Sigma 12-24mm DG 4.5-5.6 from the first DSLR era has also distortion variance through the focus range. So it's rare but possible.”
And what makes you say it is the focus adjustment and not the subject-distance change that causes this distortion change?
I think it is because we almost always focus on the subject that this misconception has arisen. But it is the change of distance that matters, not where the lens is focused.
I made you a little diagram. Show me the areas on that, that isn't possible.
Maybe think of it from a different perspective. Think about the camera as a point in space (aperture of 12mm F22 is 0.5mm so that's close), and then ray trace from there to all objects in the scene. You will end up with one image, and one image only. That is perspective. No ray can bend, no ray can deform. There is one ground truth.
Now this whole image, the ground truth, can be distorted by your lens. The lens cannot distort individual objects in the scene differently. It can distort either everything or nothing.
How much it distorts can change if you change the focus (technically it's another optical design after a focus change). But again the lens can then only distort the whole image again.
If this wouldn't be the case, distortion profiles wouldn't work, and you would break physics. As shown in the diagram.
More on it - mentioned lens (1224 DG) has also different colour cast over frame (corners are more bluish) - and it visible via optical finder of DSLR. And it changes by focusing. Why is so? Because on of thickest lens has 'brownish' colour cast and it moves when you focus lens. Of course, Sigma optical engineering nowadays is miles ahead of early 2000s Sigma but some drawbacks is always part of design process. Distortion and Vignetting is usually perceived as less important than sharpness.
@Samuel Dilworth You have a photo of a ruler, so close to the lens that the lens would be incapable of focussing on it. It's also wildly dark. Why? Because it's basically blocking part of the lens causing vignetting.
I wanted to make some samples with my 12mm Sony, only to find out it isn't prone to distortion changes with focus distance changes. So the only thing these images prove is that objects at different distances suffer from the same distortion. [disproving the object distance theory] But it's not a great example of the focus distance distortion phenomenon, so I will have to search for another lens in my collection which exhibits this.
Nikola’s video is convincing. Thanks for that. I understand that distortion would theoretically change with focusing distance if the power distribution around the stop changes, but since that doesn’t noticeably happen with unit-focusing lenses and all others that I have tested, I thought it wasn’t worth mentioning. In the lenses I have tested, subject distance dominated the distortion characteristics. But I see there are exceptions. Possibly an ultra-wide zoom is as extreme as these exceptions get.
Panther fan: the ruler is dark because it’s underexposed compared to the sunlit building. What does that matter? If the ruler is close, that only exaggerates the distortion change with subject distance. It doesn’t invalidate the concept. Normally of course we don’t see extreme changes of distortion.
I does invalidate the result. If you can prove it so easily, why don't you simply make a shot which shows what you claim?
Also yes unit focus lenses don't exhibit this. But it is not restricted to wide-angle zooms at all. Look at this video again, the 20mm prime clearly shows this. It is also visible in some far longer FL lenses. I will search for some examples of this
Fujifilm's X-H2 is a high-resolution stills and video camera, that sits alongside the high-speed X-H2S at the pinnacle of the company's range of X-mount APS-C mirrorless cameras. We dug into what it does and what it means.
Holy Stone produces dozens of low-cost drone models aimed at consumers. We look at the HS710 and HS175D to see if they stack up to other sub-250g offerings. Are these secretly great or more like toys?
The EOS R6 II arrives in one of the most competitive parts of the market, facing off against some very capable competition. We think it rises to the challenge.
Above $2500 cameras tend to become increasingly specialized, making it difficult to select a 'best' option. We case our eye over the options costing more than $2500 but less than $4000, to find the best all-rounder.
There are a lot of photo/video cameras that have found a role as B-cameras on professional film productions or even A-cameras for amateur and independent productions. We've combed through the options and selected our two favorite cameras in this class.
What’s the best camera for around $2000? These capable cameras should be solid and well-built, have both the speed and focus to capture fast action and offer professional-level image quality. In this buying guide we’ve rounded up all the current interchangeable lens cameras costing around $2000 and recommended the best.
Family moments are precious and sometimes you want to capture that time spent with loved ones or friends in better quality than your phone can manage. We've selected a group of cameras that are easy to keep with you, and that can adapt to take photos wherever and whenever something memorable happens.
What's the best camera for shooting sports and action? Fast continuous shooting, reliable autofocus and great battery life are just three of the most important factors. In this buying guide we've rounded-up several great cameras for shooting sports and action, and recommended the best.
Fujifilm's X-H2 is a high-resolution stills and video camera, that sits alongside the high-speed X-H2S at the pinnacle of the company's range of X-mount APS-C mirrorless cameras. We dug into what it does and what it means.
A blog post from Panasonic touts the ways its organic film CMOS sensor can control color spill between pixels, giving more accurate color in challenging lighting, but doesn't propose photography as something the sensor is suited for.
Digital sensors are at the heart of digital photography, but their development sometimes gets obscured by the marketing claims made along the way. We take a look at how sensors have developed since the early days of CCD, to better understand the milestones of the past and what's really going on today.
Who doesn't love a compact prime? This full frame 26mm F2.8 Nikkor lens has a sweet price, good features...but does it take a sharp photo? See for yourself in this new sample gallery shot with a production lens.
We met up with OM Digital Solutions' senior management at CP+ in Yokohama to find out what the transition from Olympus was like and to gain insight into its focus going forward.
If you're new to the drone world, there's much more you need to know than how to unbox and launch a drone. We break down the 5 mandatory steps you need to follow to fly safely and legally.
Camera and lens rental company Lensrentals has published a list of its most popular 'point-and-shoot' cameras, only to reveal that the very concept might now be obsolete.
Holy Stone produces dozens of low-cost drone models aimed at consumers. We look at the HS710 and HS175D to see if they stack up to other sub-250g offerings. Are these secretly great or more like toys?
Leica has announced the Vario-Elmar-SL 100-400mm F5-6.3, a telephoto zoom lens for L-mount. It's also announced a 1.4x extender teleconverter to work with it.
Round One voting results are in and we're now down to 8 matches. Jump in to see who won and sneak a look at how the DPReview team debated the choices, then get ready for Round Two – voting starts now!
Mathematical correction of lens distortion is commonplace. We explain why we don't think it makes sense to ignore it or to assume it's always a bad thing.
How well do Fujifilm's film simulations match up to their film counterparts? We revisit a classic DPReviewTV episode in which Chris Niccolls and Jordan Drake shoot a few rolls of Fujifilm's Acros 100 II, and a few frames on the X-T3 in Acros film simulation, to find out.
It's March, and in America that means it's time to start arguing over which college athletics team is the best at basketball. For DPReview, it's also an opportunity for a good old-fashioned camera fight.
We take OM System's new 90mm prime F3.5 macro lens out and about around Seattle, in search of sunlight, people and very tiny things to get up close and personal with. Flip on through what we found, and see how the lens performs in the real world in our sample gallery.
After a three-year hiatus, we've been at the return of the CP+ camera show in Yokohama, Japan. In between interviews with executives of the major companies, Dale Baskin took to the show floor to bring you this report.
OM System's latest lens is a whopper of a macro, featuring optical stabilization, full weather sealing, up to 2x magnification and a whole lot more. What's it got and what's it like to use? Let's dig in.
Comments