One of the biggest complaints the Sony full-frame mirrorless system has been receiving, is lack of good lens choices. With the launch of the Sony FE mount, Sony introduced only two high quality prime lenses in collaboration with Zeiss, the FE 35mm f/2.8 ZA and the FE 55mm f/1.8 ZA, both of which have been designed specifically for the short flange distance of the Sony A7 series cameras and have stellar optical characteristics. In September of 2014, Zeiss introduced two additional high quality primes for the Sony FE mount. Dubbed “Loxia”, these lenses are quite different from the Sony versions in a number of ways. First, they are both engineered and made by Zeiss, which means higher quality build and construction. Second, similar to many other Zeiss lenses, the Loxia line is manual focus only – and it is designed to be so. Third, they are also optimized for videographers, with a “DeClick” feature, which allows for smooth adjustment of aperture right on the lens. A number of our readers expressed interest in the Loxia lenses, so after having an opportunity to shoot with these gems, I was able to measure their optical performance in my lab. In this article, I would like to provide some information on the optical characteristics of the two Loxia lenses. Let’s start with the Loxia 35mm f/2:
Zeiss Loxia 35mm f/2 Biogon T*
The Zeiss Loxia 35mm f/2 feels nothing like the Sony 35mm f/2.8. It is noticeably heavier, has an all-metal barrel (vs plastic) and hood and feels completely different to handle. Unlike the Sony 35mm f/2.8, it has an aperture ring, so you control aperture directly through the lens (there is no way to do it through the camera). The focus ring is smooth and very responsive. The moment you start moving the focus ring, the camera goes into the zoom mode automatically, allowing you to get precise focus. Overall, it feels and acts like a much higher-grade lens than the Sony-made version. Let’s take a look at its MTF characteristics:
And here is the comparison with the Sony 35mm f/2.8:
Yup, we have a new center resolution champion – the Zeiss Loxia 35mm f/2 outperforms the Sony 35mm f/2.8 in the center at f/4. However, its mid-frame and corner performance are a bit disappointing due to field curvature, particularly when you focus at close distances. So if you focus at something flat like a wall, you might get really sharp center focus, but the mid-frame and the corners will look less sharp in comparison. As you focus towards infinity, field curvature diminishes greatly. And stopped down to f/5.6, it practically disappears – you can see that it matches the performance of the 35mm f/2.8 from there on. The second issue is focus shift, which was quite apparent when I was measuring the MTF performance of the lens. The good news is that focus shift for this lens does not matter, since you will most likely be setting aperture first, then focusing. But if you decide to focus once and change aperture, keep in mind that focus shift can result in focused area being slightly moved, resulting in unintended area getting in focus instead. Although I might sound like a Zeiss fanboy right now, but this is also not a big issue for this lens. Zeiss did not optimize the design the Loxia lenses to be ultra-sharp all across – the highlight of these lenses is the beautiful rendering of images, which they accomplish wonderfully.
The highlight of the Loxia 35mm f/2 is distortion, or lack thereof. Imatest measured pincushion distortion at 0.12%, which is even lower than on the superb Sigma 35mm f/1.4 Art! Vignetting is pretty noticeable at f/2, but stopped down it is reduced dramatically. Chromatic aberrations are well under control and Imatest measured at most over a pixel of CA. To be honest, I actually like the way this lens vignettes and I would personally choose not to remove it in post – for me, it is a characteristic of the lens rather than an optical issue.
A couple of image samples that will be published in the upcoming review:
Zeiss Loxia 50mm f/2 Planar T*
The Zeiss Loxia 50mm f/2 is another beautiful lens, which is crafted very similarly as the Loxia 35mm f/2. In fact, if you buy the 35mm f/2 for your Sony A7 series cameras, it makes sense to get this lens as well, since you will most likely fall in love with both. And here are the MTF numbers produced by Imatest:
And here is a comparison to the Sony 55mm f/1.8:
Once again, the Zeiss Loxia outperforms the 55mm f/1.8 in the center at f/4. However, just like the 35mm f/2, the mid-frame and the corner performance certainly suffer in comparison, especially at wider apertures. The Sony 55mm f/1.8 is better across the frame when stopped down though, so if you are looking for the best sharpness at most apertures, the 55mm f/1.8 is still the way to go. By f/8, both lenses are more or less the same in terms of sharpness.
The Loxia 50mm f/2 also suffers from field curvature and focus shift issues, but the problems are a bit less noticeable than on the 35mm f/2. To get the best out of this lens, I would still recommend to set the aperture first, then attempt to focus. At close distances, you are dealing with very thin depth of field, so you want to keep this in mind. As for field curvature, it is much less pronounced, but is definitely there when compared to the Sony 55mm f/1.8.
Distortion is also very well taken care of – Imatest measured -0.97 barrel distortion, which is not an issue and won’t be visible in your images. Vignetting is controlled very well. Slight vignetting is visible at f/2 and f/2.8, but almost disappears by f/4. Chromatic aberration is even lower than on the Loxia 35mm f/2, measuring at most half a pixel.
A couple of sample images:
With the two Zeiss Loxia lenses and the three newly introduced Sony FE primes, Sony’s line of high quality primes now consists of a total of 7 lenses. I am looking forward to seeing more high quality lenses like the Zeiss Loxia series and I am definitely anxious to see how the new 35mm f/1.4 performs compared to the above two 35mm lenses!
I would like to know from a professional like you the pros and cons of using manuel lenses with cameras like the Sony r7 and others…
Nasim, are the Loxia ‘made by Zeiss’ in Germany? That would make them much more interesting to me but I thought they were made in Japan? Also, do you really think one needs both 35mm and 50mm ? What would be the advantage of both FL’s? To me they are very close together. I would love a Loxia ~105mm instead.
Jeff, rarely anything is made in Germany anymore by Zeiss, due to high costs. Most lenses are made by Cosina in Japan, just like many of Nikon’s lenses are made in Thailand nowadays. However, the glass that is used in Zeiss is German-made and there is strict control in the QA and other processes, so there is little difference between the German-made Zeiss and Japanese-made Zeiss.
As for focal lengths, 35mm is definitely not close to 50mm – that 15mm difference is very noticeable.
Nasim, can I propose that you in your graphs put a sort of “lower limit of decency” when it comes to resolution ? So that people that are less technically inclined can see what do you consider to be “decent” resolution.
Elvir, thank you for your suggestion. I have been thinking of a way to add interpretation data in graphs, but have not found a solution yet (I use Google graphs). Will work on doing this ASAP.
The first question I’d like to ask Nasim – what was the camera author used to measure MTF? Was it 24MP or 36MP?
For me, the most interesting details if Loxia 35mm is perfect lens to create high detailed landscapes. I can’t say it about Sony FE 35mm/2.8. I should say that based on my expirience Sony FE 35mm/2.8’s resolution is unpleasantly depending on distance that is also related with too noticeable and generally unacceptable its field curvature that is absolutely terrible at short distance about 20 inches. If you try to shot an unfolded pages of some magazine with small font by Sony FE 35mm/2.8 you’ll face with a terrible discovery… no any uniformity of field sharpness…
And the second unpleasant thing about Sony FE 35mm/2.8 is how it works at big distance too (let say 500 feet and more up to infinity) , you can find that some figures of apperture are not applicable because they are not able to provide good sharpness. Based on my expirience f/4.0 till f/5.6 appertures provide slightly blured and definetely non perfect pictures. To understand better – my copy of this lens provides noticeable better picture at f/2.8 than at f/4.0 when I shot some far scene, it’s incredible nonsense, but it’s so. I call such defect – restricted range of appertures. You can also look at www.imaging-resource.com/artic…-has-moved to make sure.
And in this connection it would be nice to get to know if Loxia 35mm suffers from similar ugly features or not?.
Anna, it was the Sony A7R, so 36 MP. I always measure MTF with the highest resolution cameras available for a particular mount. As for the Sony 35mm f/2.8, it seems to be a hit and miss for some people. I tested two copies of that lens and both performed admirably at both relatively close distances of 5-6 feet away from the subject and at infinity. Perhaps there are some nasty lemons out there…
Loxia definitely suffers from field curvature at short distances and has noticeable focus shift. But like I said, those are not particularly problematic if you know what you are doing. I think the Loxia 35mm is quite good at infinity – looking at landscape shots that I have with that lens, sharpness seems to be uniformly good, especially when stopped down to f/5.6 and smaller.
Nasim, many thanks for your reply. I’m owner the most of Sony E Full Frame native lenses (24-70/4, 16-35/4, 55/1.8, 35/2.8) and I own also Samyang 35mm f/1.4 with Sony E design.
And I should say that for me, the only Samyang 35mm is really perfect lens that is able produce uncompromising highly detailed landscape pictures. Two years ago you have tested Samyang 35mm and found it’s very good lens: photographylife.com/revie…5mm-f1-4/3
The only thing is terrible about Samyang 35mm – it’s unbearably hard and bulky (800g /5.5″ length) gadget even for DSLR, let alone a relatively light and compact mirrorless camera. And my dream is to find something compact one instead of this monster but with the same optical quality.
The figures of MTF for Samyang 35mm look like unachievable for as Loxia as well as for Sony 35mm f/2.8 , but I don’t know what MTF threshold has been used for Samyang test, it may be MTF50 or MTF10… maybe something else… the finest details of resolution measuring methodology can blow my mind….
So, let me ask simply – can Loxia 35mm be close to Samyang 35mm f/1.4 in terms of resolutions and other aspects of optical quality?
Thanks for that interesting review. Allow me to throw in that the A7R seems to have a loss of resolution wide open and in the corners compared to the A7II. I have no explanation and can only link to www.traumflieger.de/repor…7-Modellen::903.html where the loxia 50 shows this effect and equally the Loxia 35mm in another review from Traumflieger.de. The higher resolution sensor does not show equally higher lpm’ as one would have expected, if my mentioned review is in order.
In addition the A7S has even better resolution than the A7R wide open. Overall the A7S seems to be the most efficient camera in relation to the Loxias although the A7II has higher resolution: The A7S does not show that much of difference from center to corner like the A7II, while the A7R somehow sabotages itself.
Doc, again, that is simply impossible – a lower resolution sensor cannot possibly yield higher numbers. If that’s what the reviewer says (sorry, cannot properly read the German article on my iPhone, as I’m currently traveling), then you should avoid reading his/her reviews, as it shows incompetence in proper testing with Imatest and lack of understanding basic principles of camera resolution.
Thanks Nasim for your answer. As a reader without technical knowhow about testing one just tries to interpret. The A7R shows highest resolution @ optimal apertures and leaves the A7II and A7S far behind at these apertures. My understanding was, that the A7R sensor has problems @ critical apertures and mostly in the corners wide open, like Leica glass does not perform to its optimum on the a7 line due to certain reasons mentioned on the web. Latter seem to pronounce especially on the A7R. Now your explanation bears credibility due to your arguments and your respected website. Still the reader stands a bit alone. If you could give more information about your testing sceme, like Lukasz asked, this would be interesting and helpful. Not to be meant to challenge you instead to get an idea how the reader can figure out a solution from his own inquiries and to filter out reviewers that are wrong according to your sceme. The consequence of your explanation might be, that we do not only need the best camera and glass to get the best results but also RAW converters that are mostly unknown and that the most popular RAW converters seem to be insufficient. Quite a frustrating consequence since optimal Camera/lens/RAW-converter combinations reach a complexity impossible to handle.
I read the article on traumflieger. It contains a paragraph which explains the unexpected lower resolution of Loxia 50mm with A7R at special apertures: It’s the well known shutter shock problem of A7R.
In that case the reviewer should not have pointed out that other cameras show better numbers. The shutter shock issue is completely separate from the lens capabilities…
The reviewer was surprised that it happened with a 50mm lens. It’s clearly not the fault of the lens. But the dimensions and the weight distribution of that particular lens (and stability of the used tripod) triggered the well known shutter shock at 1/125s of the A7R. So I think it’s valid to present that result. I guess Doc has read only the numbers of tables but did’t read the explanation below. (The reviewer should add a footnote to table cells with the problematic samples.)
You are right. Although I read it I did not understand it completely and still do not. I am sorry if I caused an obsolete discussion from your point of view. IMHO it is a mixed bag of something whatever and I will therefore test and compare the cameras and lens combos before buying. The theoretical explanation has to many unknown points to come to a straight conclusion: Now we have camera+lens+RAWconverter+speed+aperture to align to get optimum results from the A7R – impossible to handle outside a laboratory. The shutter shock issue is a massive failure from Sony, the delicate 36MP sensor on the other hand a well known disadvantage easily handled with the right lenses. Still I believe the A7II is the way to go with the Loxias for allround usage. The A7S seems to be the most forgiving camera if one wants to use vintage lenses, and its results are very pleasing on smaller sized pictures still large enough for most of us.
Doc, I’m sorry, but I cannot take any reviewer who uses Lightroom as RAW converter seriously, because of how much it screws up the data. I know this, because I used to do it myself and after deep analysis, I decided to abandon the process and move to DCRaw.
It is not possible that a higher resolution sensor would yield lower numbers, particularly in the center. The reviewer was most likely not doing something right, or his shutter speed was in the danger zone where the shuttershock caused lower results at particular shutter speeds, which led to erroneous results. I have personally seen this myself, so I had to find ways to properly test the setup.
Nasim, a comprehensive review of avaiable RAW processing software pieces with their strengths and weakness would be great, unless you already wrote one and I missed it?
The drawing of these lenses reminds me of the Contax G lenses (unsurprisingly I guess). Looking forward to the full review.
Nasim, Samsung has upped their game with NX1, think they are serious to give some competition in the mirrorless cropped frame segment. I would love to hane a review of NX1 from your side. Thanks…..
Sourav, the NX1 seems like a great cam to test – have heard great things about it. Will have to put it in my “to test” queue, thank you for your suggestion!