Test/ First Impression Sigma 35mm f/2 DG DN Contemporary

[Jonathan-Mac]

This was in the "Latest 'this system has it'" thread; but I've owned several of the Sigma mid-range primes for E-Mount and M4/3 (i.e. the 30/2.8 DN Art and 60/2.8 DN Art), as well as the L-Mount 45/2.8 Contemporary, and none of them have wowed me. They've all been bland to one degree or another, clinically neutral in rendering (when I prefer a lens with vivid color/contrast), and the M4/3 editions were seriously lacking in tonality (subtle things like gloss on smooth water, or tonal shifts on a rounded surface) compared to counterparts like the Panasonic 20/1.8.

The 45mm is known to be the poorest-performing of the I-series lenses and m4/3 lenses are entirely different. The other I-series lenses, at least the ones I've tried, are superb.

There is no Panasonic 20/1.8 - there are 18/1.8 and 24/1.8 for L-mount and there's a 20/1.7 for m4/3 but no 20/1.8 anywhere I think.

 

[Travis Butler]

The 45mm is known to be the poorest-performing of the I-series lenses and m4/3 lenses are entirely different. The other I-series lenses, at least the ones I've tried, are superb.

There is no Panasonic 20/1.8 - there are 18/1.8 and 24/1.8 for L-mount and there's a 20/1.7 for m4/3 but no 20/1.8 anywhere I think.

Sorry, yes, I meant the 20/1.7; been spending too much time talking about the L-mount f/1.8s lately. ^^;;

Note that it wasn’t just on M4/3, I was similarly unimpressed with them on APS-C E-mount.

 

[GeorgeHudetz]

Arooo?
The Sigma Art 135/1.8 is just amazing.
...But then, for what it costs and what it weighs, it bloody well should be.

Too bad it's been discontinued. I've been thinking of getting one used, but then I suspect (hope) that the discontinuation of the HSM version means there is a DG DN version coming.

 

[GeorgeHudetz]

Sorry, yes, I meant the 20/1.7; been spending too much time talking about the L-mount f/1.8s lately. ^^;;

Note that it wasn’t just on M4/3, I was similarly unimpressed with them on APS-C E-mount.

Of course, as soon as we start talking about subject aspects of a lens like rendering, etc., it becomes very personal.

However, every L-mount Sigma I own (105 macro, 65/2, 24-70/2.8 Art (first gen), 14-24/2.8 Art, even an older DSLR-era 12-24 Art) has wowed me. I am looking at the resultant images through a landscaper's eyes, so what you find clinical may be exactly what I want to see (i.e., lots of detail and contrast). I have also found Sigma lenses, on average, to be a bit warmer than my LUMIX lenses, which, to me, are a bit on the cool side. Nothing that a nudge to white balance can't fix, of course.

 

[Travis Butler]

Of course, as soon as we start talking about subject aspects of a lens like rendering, etc., it becomes very personal.

However, every L-mount Sigma I own (105 macro, 65/2, 24-70/2.8 Art (first gen), 14-24/2.8 Art, even an older DSLR-era 12-24 Art) has wowed me. I am looking at the resultant images through a landscaper's eyes, so what you find clinical may be exactly what I want to see (i.e., lots of detail and contrast). I have also found Sigma lenses, on average, to be a bit warmer than my LUMIX lenses, which, to me, are a bit on the cool side. Nothing that a nudge to white balance can't fix, of course.

Mmmm… not really; I normally like lots of detail, and love contrast.

What I mean by ‘clinical’ is… well, right there in the term - makes things look too much like a clinic or medical/lab setting. So things like being aggressively neutral - not just lack of a color cast, warm or cold, but weak and muted color. Sterile. Like so many other photo issues for me, tonality plays a huge role - I want to see a wide spectrum (dynamic range), capturing subtle differences (clouds, smoke), smooth rendering of tonal gradients (light falling on a curved surface, glossy surfaces, with no compression in the range of tones). A lens lacking that feels ‘clinical’. And razor-sharpness can contribute to the effect when combined with the others.

Or to put it in a more subjective way: clinical lenses feel lifeless to me. Just sort of ‘there’. Useful for documentary purposes, maybe, but not art or joy.

The Sigma lenses I named have what felt like muted color to me, and were distinctly lacking in tonality. The 45 Contemporary wasn’t as bad as the E-mount and M4/3 lenses, but was still lacking the ‘spark’ I want to see in a lens. The Panny 20/1.7 and PanLeica 15/1.7 did have that spark for me on M4/3, for a comparison, and were two of my favorites lenses on the system.

Some of the pics with the 35/2 in this thread do have that spark, so that’s hopeful. Unfortunately, doesn’t appear to be available to rent at the places I checked, and buying is out of the question until I get stable employment again. Maybe someday…

 

[GeorgeHudetz]

Travis, thanks for the explanation. I've never heard somebody define what they mean by a "clinical" rendering. I always thought it meant that the lens was too sharp and too perfect, so I typically just wrote those statements off. However, I certainly agree with you that I would not want what you describe above in my images.

And, I've never felt that way about any of my L-mount Sigma glass, even when doing side-by-side equivalents with Panasonic glass. But, I have not owned the 45/2.8, and, as others have stated, Sigma intentionally engineered in some "character" to that lens, particularly when wide open, and - I think - for up close focus. This fact is documented somewhere in a blog post from a Sigma lens designer. But they abandoned that approach for the remainder of their Contemporary lenses, as I understand it.

Anyway, I do encourage you to experiment more with the L-mount Sigma glass, as you can. The 65/2, for example, is spectacular.

 

[Oíche]

Or to put it in a more subjective way: clinical lenses feel lifeless to me.

The thing about sigma, it's got lifeless eyes, black eyes, like a doll's eyes.

 

[Travis Butler]

Travis, thanks for the explanation. I've never heard somebody define what they mean by a "clinical" rendering. I always thought it meant that the lens was too sharp and too perfect, so I typically just wrote those statements off. However, I certainly agree with you that I would not want what you describe above in my images.

I’ve heard “too sharp, too perfect” as well. I think at least some of them may be trying to get at what I described without having a good way to express it. A lack of tonal range can make something look overly smooth/low-detail; clouds, for example, can look like white blobs when you compress the tonal range, but start showing volume and filaments and complexity when you add tonality. Or loss of tonal detail can be like the overly homogeneous, plastic-y look you can get with some smoothing/denoise filters. It looks more like a plastic replica (or a piece of lab equipment) than a real object - “too perfect”. And that effect gets worse when you combine it with sharp edges and detailing - “too sharp”. Flowers can be particularly bad there.

You can see the same sort of discussion regarding SF spaceships in the 70s. They were often criticized as being too sleek, too perfect, before Star Wars popularized the “lived-in” look. (To be fair, movies like Silent Running did it first.)

And, I've never felt that way about any of my L-mount Sigma glass, even when doing side-by-side equivalents with Panasonic glass. But, I have not owned the 45/2.8, and, as others have stated, Sigma intentionally engineered in some "character" to that lens, particularly when wide open, and - I think - for up close focus. This fact is documented somewhere in a blog post from a Sigma lens designer. But they abandoned that approach for the remainder of their Contemporary lenses, as I understand it.

Yeah, it’s interesting, I’ve heard that about the 45 as well. (And seen “character” used disparagingly to mean “deliberately flawed”, which seems unfair to me.) I’m not really sure what they mean, because I haven’t seen it (then again, I haven’t really used it for close-ups, or deliberately wide-open for bokeh effects). What I have noticed is that pics from it often feel dull to me; I haven’t pinned down exactly why, but I think lack of color/contrast has a big part in it.

Anyway, I do encourage you to experiment more with the L-mount Sigma glass, as you can. The 65/2, for example, is spectacular.

I’d love to if I have the chance.

 

[ermesjo]

Anyone in this forum that has experience with corner sharpness and chromatic failures (esp. night and stars, details in for instance trees against bright background). I am considering two lenses instead of a wideangle-zoom. The Lumix L 35 and 18 f/1.8 compared to Sigma DG Cont. 35 and 20 f/2.0. Are they similar or is Lumix or the Sigmas the best ones?

 

[AlanC]

Anyone in this forum that has experience with corner sharpness and chromatic failures (esp. night and stars, details in for instance trees against bright background). I am considering two lenses instead of a wideangle-zoom. The Lumix L 35 and 18 f/1.8 compared to Sigma DG Cont. 35 and 20 f/2.0. Are they similar or is Lumix or the Sigmas the best ones?

I can only comment on the Sigma 35/2, which I use on the Sigma fp L (60MP). Corner sharpness is good to the point that I don't notice any fall off unless I zoom the images to 200%. Purple fringing of tree branches against a bright sky (something I really dislike in a lens) is low to non-existent: it can be caused by a blown out sky, but there's never very much of it. All of the i-series primes I own (17/4, 24/3.5, 35/2, 45/2.8, 65/2) are very good in this respect.

 

[Jonathan-Mac]

I can't say I've ever noticed any colour fringing or sharpness problems in the corners, even wide open. I have used my Sigma 35 for startrails but that's not the same as static shots of stars.

As for the 20mm, I went with the 24mm as a more general-use wide-angle but I remember seeing reviews of the 20mm which said it was really superb (which I would say is also the case for the 24).

 

[GeorgeHudetz]

Lenstip tests include a section on coma (and other corner aberrations) for every lens they test, and while the don't yet test Panasonic L-mount lenses, they do test Sigmas, but on a Sony body. It's a good way to see how the various sigma primes would perform for astro.

Sigma C 35 mm f/2 DG DN review - Coma, astigmatism and bokeh - LensTip.com

Best digital cameras and lens reviews. If you are looking for the information about digital cameras and lenses you are in a right place. We have many professional tests of digital photography equipment.

www.lenstip.com

 

[GeorgeHudetz]

And, for the record, I have used the Sigma 65/2, the Panasonic 50/1.8, and the Sigma 20/1.4 for astro, and they have all done well. But I have no experience with the lenses you ask about. Looking at Lenstip, I'd say the 35/2 would do just fine, but the 20/2 would have a little visible coma, but that's not the end of the world unless you intend to print large.

 

[ermesjo]

So the Sigma 20/1.4 do better than Sigma 20/2.0 regarding chromatic aberration? Maybe someday I get the chance to photograph northern lights and stars (living nearby Oslo but the best chances is in Tromsø, in the Northern Norway). The background for my question is that I hate chromatic errors, and that I make videos indoors from time to time. That is why fast lenses are preffered. At the other hand, I read a test that the Sigma has quite severe barrel distortion, not good for architecture out/indoors and light fall off in the corners. Must evaluate the best compromise. After all, my impression is that the Lumix lenses After all is very good.

 

[AlanC]

At the other hand, I read a test that the Sigma has quite severe barrel distortion, not good for architecture out/indoors and light fall off in the corners. Must evaluate the best compromise. After all, my impression is that the Lumix lenses After all is very good.

This is very common in mirrorless lens design: distortion and vignetting are easily correctable in software, and allowing a certain amount in the lens design apparently allows the lens to be made sharper and/or other defects to be better corrected. It doesn't show up in the final images because each lens has a correction profile that's applied during processing of the RAW data, either in-camera or in software such as Lightroom.

Panasonic's lenses are no different to Sigma's in this respect: if you have a RAW developer that allows you to turn off the correction profile then you'll see that they also have distortion and vignetting that can be quite significant. A couple of other examples of this are the Leica Q series 28mm lens, which in reality is close to 24mm but requires sufficient cropping to correct the barrel distortion that the final result is equivalent to 28mm, and the Sony 20-70mm f/4 which has such severe distortion at 20mm that you couldn't even call it rectilinear.

 

[GeorgeHudetz]

So the Sigma 20/1.4 do better than Sigma 20/2.0 regarding chromatic aberration? Maybe someday I get the chance to photograph northern lights and stars (living nearby Oslo but the best chances is in Tromsø, in the Northern Norway). The background for my question is that I hate chromatic errors, and that I make videos indoors from time to time. That is why fast lenses are preffered. At the other hand, I read a test that the Sigma has quite severe barrel distortion, not good for architecture out/indoors and light fall off in the corners. Must evaluate the best compromise. After all, my impression is that the Lumix lenses After all is very good.

Yes, both Sigma 20mm lenses have distortion; the f/1.4 lens has less than the f/2 lens (again, check Lenstip).

As Alan points out, most lenses today have distortion. Historically, distortion was bad, since it's hard to correct distortion in a darkroom. It's much more common now, since it's easier to correct digitally, and as Alan also points out, allowing some distortion into the design gives the lens designer one more "lever" to adjust when attempting to reach a given target. Having said that, some people still do not want to buy lenses with distortion. I'm not sure where I am on that spectrum. I do tend to think that correcting distortion should have some kind of negative effect when viewing at 100%. This is why I like CaptureOne as I can control how much distortion correction is used. And also why I am skeptical of lenses that have LOTS of distortion, particularly when it's complex (i.e., moustache distortion).

Back to corner aberrations: I was talking about coma as opposed to chromatic aberrations. Coma causes points of light (i.e. stars) to become larger discs. Also, some lenses produce "wings" around point light sources, and although this is technically not coma, it is often reported as such. Does not matter what you call it - the point is to understand a given lens's tendency to make stars look like, well, something else. Small interesting tidbit: often lenses that have soft corners also have bad coma.

Another tidbit: like vignetting, coma tends, to varying degrees, to go down (i.e., get better) as you stop the lens down. Combining that with the fact that sharpness often goes up as you stop the lens down, you will find that savvy astrophotographers who have those giant, fast UW lenses often shoot them at F2 or F2.2 or some such, instead of wide open, to optimize the optical characteristics of the lens. Less vignetting = more light. Less coma = sharper stars. Stated another way, if you compare the Sigma 20/1.4 stopped down to F2 to the 20/2 wide-open, the faster lens will give better results, even though it is being shot at F2. And, to optimize the 20/2, you should probably shoot at F2.8.

Getting back to distortion, the 20/2 does have a LOT of distortion. The 20/1.4 seems much more reasonable to me. For whatever that's worth.

 

[ermesjo]

Thank you Alan and George, for insightful explainations. Due to economical considerations, I may tend to go for the Sigma 20/2 since it is cheaper than the Lumix 18, and also have good reviews, its small and handy. Pricewise, the Lumix and Sigma 35 (1.8 and 2.0) have more similar price. But I need the 20mm first for indoor shots/video. My old Canon EF 17-40 f/4.0 may be weak in murky light indoors. A project in May-June. It’s my first serious video-project. I have made two YouTube videos last August, but are working on 6 new episodes. This time with my new S5iiX, an Atomos Ninja, Røde and Sennheiser mics and stuff. More on that later in this forum.