Equipment

Sharpness Tests of the Sigma 14mm f/1.8 DG HSM Art

We’ve had a number of fun, new lenses to test this summer and one I was pretty eager to get to was the Sigma 14mm f/1.8 Art, for a couple of reasons. First of all, it’s a 14mm lens that has a wider aperture than f/2.8, and that’s certainly interesting. Second, it’s a new Sigma Art prime lens, and those have been spectacular. So I begged and threatened and got the first ten copies for some bench testing before they went in stock.

As always, these are optical bench tests, so take them for what they’re worth. It is not a lens review because I don’t review lenses. That’s what photographers do. I test them, because, well, I’m a tester. Test results should tell you if the lens is worth consideration and further investigation, not that you should run out and buy it. I don’t make any suggestions about what you should run out and buy because I have no idea how you shoot or what’s important to you. But if the resolution is important to you, then read on.

As always, these are the results of 10 tested copies; each tested at four rotations with 84 data points.  For those who don’t speak MTF, the easy version is higher is better, and dotted and solid lines of the same color close together are better.

MTF Tests

I was curious about how well the Sigma 14mm f/1.8 Art lens would perform at f/1.8. Sigma has an excellent track record with the Art Primes, but there are good reasons other manufacturers are limiting themselves to f/2.8 apertures at wide angles. But the performance wide open is impressive — this is sharp even wide open, even at high resolutions (blue and purple lines), and even at the edges.

Olaf Optical Testing, 2017

 

As most of you know, I usually don’t test stopped down, but I did this time. Partly because I am aware, despite my begging and pleading, that Fanboys LOVE to take the wider aperture MTF and compare it to the narrower aperture MTF of a competitor’s lens. That’s especially true in this case where the competitors don’t reach f/1.8. So here is a set of Sigma’s at f/2.8. (I only tested the first seven at f/2.8, because, as William Gibson said, time be mos’ precious, mon).

Olaf Optical Testing, 2017

 

Two things you’ll notice. First, this lens is getting scary good in the center at f/2.8.

The second is Roger’s 12th Law: Stopping down doesn’t make everything better everywhere. (OK, actually this is because some aberrations improve a lot, some a little, and some not at all when you stop down. But since most of y’all hate math, I just decided to skip the math stuff.) This is true of all lenses; you just don’t see it often because I don’t do stop-down tests often.

As long as we’re testing things, I probably should show you the field of focus, since that gets, well, interesting on wide-angle lenses sometimes. This isn’t sometimes; this is a very flat field with just a slight curve. A lovely, fairly flat, boring field of focus curve.

Olaf Optical Testing, 2017

 

Some Comparisons

I’m making the comparisons at f/2.8 to try to even the playing field as much as possible.

Sigma 14mm f/1.8 Art vs Canon 14mm f/2.8 L

The obvious comparison is to the Canon 14mm f/2.8 L lens since they’re both 14mm primes and all. The Canon is a significantly older design, and that shows, the Sigma is clearly better everywhere.

 

Olaf Optical Testing, 2017

 

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

The Nikon has been the gold standard of wide-angle zooms for a long time, so this is a good comparison. We’d expect the Sigma to be better, it’s a much newer design and a prime lens, and that is indeed the case. The Nikon is still a very impressive zoom, though.

Olaf Optical Testing, 2017

 

Sigma 14mm f/1.8 Art vs Canon 16-35 f/2.8 Mk III at 16mm

I won’t bore you with lots of zoom comparisons, and this one isn’t really fair; 14mm is a lot wider than 16mm and zooms aren’t supposed to be as good as primes. But the Canon is the current wide-angle zoom champion, as good as it gets at 16mm, so I thought it was worth a look. I think this is so close that sample variation would be larger than this average difference everywhere except right in the center where the Sigma is better at high frequencies. So, Canon shooters can choose between the zoom that is amazing at f/2.8. Or, the prime that is a wider angle and wider aperture f/1.8. Horses for courses. Choices are good. All that stuff.

 

Olaf Optical Testing, 2017

 

So This Changes Everything, Right?

Nope. Let’s face it, there hasn’t been an impressive 14mm prime for SLRs for a while, and I haven’t heard the natives banging on the castle doors demanding one. But for some people this is going to be a really fun lens. I consider it reasonably priced for what it is; a unique and excellent lens that hasn’t been made before. I shot real-estate for a brief moment in time, and I would have killed for this lens then. And I like playing with ultra-wide images in general, so I’m kind of excited about it. It won’t be a huge fraction of my images, but it will be a tool I didn’t really have before.

A lot of people will never shoot 14mm and will never notice this lens exists. But for those who do, the Sigma 14mm f/1.8 Art Lens is probably going to be a fascinating lens indeed.

 

Roger Cicala and Aaron Closz

Lensrentals.com

July, 2017

 

Addendum: A couple of requested comparisons

Sigma 14mm f/2.8 Art  to Canon 11-24mm f/4 L at 16mm

Olaf Optical Testing, 2017

 

Sigma 14mm f/2.8 Art to Zeiss 15mm 

Olaf Optical Testing, 2017

Author: Roger Cicala

I’m Roger and I am the founder of Lensrentals.com. Hailed as one of the optic nerds here, I enjoy shooting collimated light through 30X microscope objectives in my spare time. When I do take real pictures I like using something different: a Medium format, or Pentax K1, or a Sony RX1R.

Posted in Equipment
  • Maureen

    Thanks for sharing the test results, Roger. I always enjoy reading your analysis. One thought: I use Canon’s 14mm L-lens for architectural and interior design photography, as is certainly a common use for this lens. This means that 99% of the time, I am shooting at a very high depth of field, slow shutter speed, with a high resolution body like the 5DSR and always on the most stable tripod. How would the new Sigma compare to the Canon 14mm under these common conditions, say at f/8 or above only? Does Sigma appear any crisper than Canon under these circumstances?

  • Rob Dickinson

    Great testing as usual.
    Its very very sought after by astro shooters who are keen to see the coma values

  • Sean, it’s work we’re doing anyway, and it’s not a huge investment to make some of it public. But that does explain why we often say we can’t test this or that. We have to use our resources to support ourselves first.

  • Maxim Podtopelny

    f/2.8 would be quite enough! Thanks!

  • Arthur Meursault

    Ding Dong.

  • What about Canon’s 11-14mm at 14? Apertures are quite different, but it’s still an interesting comparison.

  • I’ll compare the Sigma at f/2.8 to the Canon at f/4, but I’m not running the Sigma at f/4. Y’all do understand that I put this stuff out for fun and I’m trying to make a living using this equipment for gainful purposes, right?

  • Maxim Podtopelny

    Please, Roger, could you be so kind to add 11-24/4 comparison later?

  • Dave Hachey

    WOW… You guys keep busy. Before I retired I ran an academic analytical facility; we typically acquired about 2 TB/day. Storage became a serious cost (and headache) for us. BTW, this looks like a great lens.

  • Sean T

    Wow Brandon, that’s great! I appreciate you guys sharing so much, with no apparent direct revenue generation from it. That seems altruistic, not something I find on the internet.

  • Well, maybe some day. The dance card is already pretty full for July and August.

  • Because I was trying to not test the Sigma at f/4. Because (as mentioned above) I was tired. 🙂

  • Damn! I knew I was tired.

  • Brandon Dube

    I gave a talk at the International Optical Design Conference yesterday on our test methods and indexed the database for some statistics. The size of it I won’t tell you in GB, but we’ve passed 4,000,000 MTF measurements.

  • Brandon Dube

    Sure, it (combined with astigmatism) is often the limiting aberration in wide aperture, wide angle systems. I can’t simulate a point spread function for it because I haven’t typed out its formula (yet).

  • Omesh Singh

    You once referred to Oblique Sagittal Spherical Aberration (OSSA) as the problem in corner rendering for astro purposes.

  • Sean T

    How neat Roger, thank you. And thank you for showing us something you don’t usually show – stopped down MTF. I’m curious how many GB of data you have that you just don’t have time to inclination to display for the photographic public. Anyway, what a neat lens! I don’t know what I’d do with it, and I’m curious how it’ll do for astrophotography (how’s that coma?), but I’m glad it exists all the same.

  • Uri Raz

    Why compare the Sigma 14mm f/1.8 to the Canon EF 16-35mm f/2.8 mkIII, rather than the Canon EF 11-24mm f/4?

  • Christian Stella

    This sounds like the lens I’ve been waiting for for indie filmmaking. I had amazing success years ago as cinematographer of a very small movie called The Battery. I shot (full frame on the 5D) primarily on a Zeiss 21mm I rented from you guys. Half the movie was confined to a car, so I needed the wide lens on full frame. But, then we started using the lens on most of the shots outside the car as well and I developed this super wide style that has been extremely hard to replicate on Super35 cinema camera sensors without losing all bokeh. That it is a Sigma Art is icing on the cake, as my super indie kit is currently the ART 24-35 (a really great range for Super35) and ART 50 1.4.

  • Petri Puurunen

    How about comparing Sigma to Samyang XP 14/2.4 ?

  • Yu

    So what is called Coma by astrophotographers is actually astigmatism, mostly loss of low-frequency details in tangential direction, is it right?

  • User Colin

    “Partly because I am aware, despite my begging and pleading, that Fanboys LOVE to take the wider aperture MTF and compare it to the narrower aperture MTF of a competitor’s lens”

    The only website where I see MTF charts of dissimilar apertures placed mirrored side-by-side is…. lensrentals. Along with some hand-waving and asking the reader to imagine the lines on the 1.4 lens are a little higher. This site generally values science, not just shooting some photos handheld of one’s bookcase and declaring the lens was “tack sharp wide open”. So in the interests of science:

    “I’m going to additionally test at f/2.8 because further down I compare the MTF charts of three other max- f/2.8 lenses and the only valid scientific way to compare with those lenses is at the same aperture.”

    For photographers who mainly shoot 14mm stopped down considerably for DoF, these charts still don’t really answer the resolution-comparison questions. Would I notice much improvement in sharpness, on a 36MP FF sensor, switching from an ultra-wide zoom at f/8 or f/11 to this prime? Apparently, “Stopping down doesn’t make everything better everywhere” 🙂

    Thanks for comparing apples with apples. Well, except for the 14mm vs 16mm, of course, which you admit isn’t fair.

  • Brandon Dube

    Coma is a wavefront aberration that modulates the phase of the pupil by rho^3*cos(phi). You can find this in many places, say https://wp.optics.arizona.edu/jcwyant/wp-content/uploads/sites/13/2016/08/03-BasicAberrations_and_Optical_Testing.pdf
    or Optical Shop Testing (Malacara), Principles of Optics (Born & Wolf), Aberrations of Optical Systems (Welford), etc.

    Here’s a picture of coma, generated using my (in development) Python fourier optics module, code6. I have also made a more matured equivalent with a less pleasant API in Matlab, “ASAK.”
    https://github.com/brandondube/code6
    https://github.com/brandondube/AberrationSwissArmyKnife
    https://uploads.disquscdn.com/images/d32731c6bb85fb5c483fc4f2f34d5f102b2c17f65c09db1b4047857f54f58039.png

  • mohammad mehrzad

    Congratulations on this great achievement.

  • mohammad mehrzad

    that is very good to know Brandon, thanks. so based on what you said, this lens should perform very well for astro landscapes, according to 10lpm and 20lpm lines. can you provide a refrence on what coma is, that is actually correct and scientific?

  • Brandon Dube

    What astrophotographers refer to as Coma is not coma. What they really mean is “is the image quality good in the corner as well as the center.” MTF is a metric for image quality, and what astrophotographers are looking for is things on the scale of maybe 5-10px in size. 1px = maybe 5um, so 5 is 25um, 20lp/mm, and 10 is 50um, 10lp/mm.

  • Brandon Dube

    Wide angle lens = very small pupil = spherical aberration and axial color (the only aberrations on axis) are very easy to correct. Barrel distortion only moves the image, it does not degrade or enhance it.

  • mohammad mehrzad

    The most important question to answer is this: how bad is the coma? because all astro landscapers are interested in this lens. another important comparison could have been with the Samyang 14 f/2.8 (I heard of an f/2.4, don’t know if it is released yet), currently considered the gold standard of coma less lenses and very popular for night sky..

  • Omesh Singh

    With wider angles I’m noticing pretty high center values and very steep drop off towards corners. Does barrel distortion increase the center resolution?

  • Ze Dong

    first time to comment first

Follow on Feedly