Vignetting is quite apparent wide open, but I did not find it objectionable and its easily removed in post if you wish. Stopping down to f4 sees a significant reduction in vignetting and personally I am very happy. If you are shooting snow scenes, or cloudy seascapes, you may find this lens requires some tweaking in post to get things perfectly even, but this is perhaps a small price to pay for the optical excellence elsewhere. If I can tear your attention away from the exquisitely romantic rooftop scene to the vignetting you will see how it it largely disappears one stop down. I’m usually quite happy with vignetting most of the time anyway and if its not there, I’ll often add it in post.
Sony Zeiss 35 Sonnar Complete Frame @ f2.8
Sony Zeiss 35 Sonnar Complete Frame @ f4
Bokeh is beautiful. That’s not a term I use regularly, but it really is excellent most if not all of the time. There’s no caffeinated energy to be found here. A bit like the 35 Summarit-M I have made references to (which has utterly dependable smooth bokeh), this is one you can rely upon to produce relaxed backgrounds. This is something Zeiss and Canon tend to be quite good at, even with aspherical lenses, whereas some of the mode modern asph Leica lenses have had quite unsettling Bokeh at times. I still think the 35 Summarit-M is the best 35mm lens I have ever used i this regard, but the 35 Sonnar is within spitting distance.
CA, now I have not left this till last with a view to hiding it. Rather it’s just that I don’t think it s a big deal. It is however a little disappointing. CA seems to be more of an issue with this lens that it is with the 28-70 Kit lens. This is something I just can’t get my head around that, but it seems to be a fact and not something specific to my sample. Its generally easily removed in post anyway, so I am not going to worry. Interestingly, it is a criticism levelled against some of the ZE and ZF lenses and so I wonder if it is something to do with the T* coatings that otherwise give beautiful colour, contrast and flare resistance. If so, I’ argue it’s a price worth paying for most people.
Conclusion
So there you have it: a very well made, reasonably compact, ultra light and perfect handling lens, with fantastic optical performance for about £650. While it may not survive being thrown around a room as well as a Leica M lens, it is less than half the price of a Leica Summarit-M, which is the only comparable lens in that lineup at £1365. As lovely as the Leica lens is, I know which one produces the sharpest edges and corners at wide apertures and it is the Sony Zeiss, by quite a margin. Stopped down, the Zeiss 35 Sonnar does not improve to the point that some other lenses do and it gets overtaken by a small margin. Here is the compromise and it won’t be for everyone. I’m certainly not inclined to lug the tank-like 35 CV 1.2 II around for scenic work, despite the slightly better overall performance at f8.
I think that Sony has sought to maintain the integrity of the Sony A7/A7R full-frame system, which is exceptional imaging performance in a compact and lightweight package. It’s perhaps not surprising that their first 35mm is this one, rather than something larger and more expensive that results in a more cumbersome combo. Considering the superb high ISO performance of the A7/R cameras, f2.8 is not much of a hindrance from a light gathering point of view, but it does limit subject-background separation. I suspect that Sony have pitched the A7/R cameras at travel, scenic and landscape photographers more than hard core people photographers and so regarded this as a necessary sacrifice for the greater good. That they have managed to produce an AF lens that is lighter than the lightest M lens in production (the 28mm f2.8 Elmarit asph) is quite astonishing.
Personally, I would have traded a little edge sharpness at f2.8 and f4, for a greater degree of perfection at f5.6 and f8, but everyone is different and for many the wide open performance will be where its at. While this might sound like a criticism, 1:2 images shot using the 35 Sonnar on the A7R (a 37″/94cm print) already show utter perfection at the edges at f4 and pretty well the same at f2.8, so I’m being really, really hard here. Just a bit more performance at f5.6 and f8 at the edges would have been welcome for those making super large prints in the 50″ range. I can’t believe I am saying this, because only two years ago, before the Nikon D800, such performance was way beyond full-frame and now here it is in a camera and lens combo that weighs just under 600g!!!!!
At this point I will add that I was sceptical of this lens. I had heard the complaints about its lightweight construction and price, but all I can say after consideration is ‘wow’. This is a direction worth heading in Sony and please keep going. My first sample was decentred unfortunately, with a very soft bottom right corner which never got sharp. Amazon’s replacement process was super slick and I had my second copy the next day, which happens to be the little gem on review here.
Pick up a Sony A7/A7R with the 35 Sonnar FE mounted and you’ll see why I find it so exciting. I am a self-confessed 35mm lover (on full frame, or equivalent on other formats) and this combo delivers an unmatched performance to weight ratio. I can’t think of anything that comes close and that’s before you factor in price. It is therefore highly recommended for travel and scenic shooters – an absolute must have at the core of your new Sony A7/A7R kit.
Fantastic review. I recently purchased the SonyA7r with the Zeiss 35mm and love it! It’s a fantastic combo and I was about to pick up the Zeiss 55 but holding back to check out the Loxia. Please continue with your reviews.
I was wondering if you had any opinion on the image quality differences between this lens (Zeiss 25mm f/2.8) vs the Loxia 35mm f/2. Any experience or thoughts?
Hi John, I have not used the Loxia, but do have a 35mm f2 Biogon for my Leica (which is superb). Both are evidently superb performers and I would suggest making the decision based on weight (Sonar much lighter), AF and speed. While the 35mm Sonnar is not perfect, mine performs incredibly well; however, not every example performs well. Decentering is common. I’m currently shooting with the 35mm Sonnar in Iceland and not feeling remotely lacking. The 35mm Sonnar is not quite as sharp as the 55mm off centre, but its still going to push out A1 prints that look phenomenal…. I only wish there was a 24 or 25mm f2.8 or f4 Loxia and I’d snap that up! Using the Rocoh GR quite a bit and if Sony or Zeiss could produce a 25mm that performs like the little GR’s 28mm, they would sell them by the truckload.
This lens is also a great match on the A6000. Small, light, fast, sharp. A great combo.
Great article. I’ve just picked up an A7 and the Sony FE 35mm f2.8, and so far am enjoying it. (The weight and feel anyway- I’ve had it less than a day, with no real chance to use it yet).
I’d be really interested in how you determined that you had a decentered lens. You have mentioned in several articles about how good a ‘good’ one is – I’d like to check I have a good one while I still have the chance to swap it for another one. Is there any test’s you would recommend, or just go out and shoot with it and look at the results. (It’s harder to know what a good one is when I have no frame of reference with which to compare).
On a separate note- I’ve been reading trough a load of your sony articles, and am really enjoying them. Cheers…
Thanks Adam, I am glad you are enjoying the articles!
I do fairly simple test to check for obvious decentering and then, if any is evident, decide whether it really matters to me. If its unimportant, then I keep the lens, but if it bothers me in use, I return it. Examples of what would not bother me would be ever so slight decentering in a corner on a fast lens which is visible wide open, but which clears up a few stops down (because I do not shoot planar scenes at f1.4 or f2).
I have actually just returned a Canon 16-35mm f4 IS due to decentering, because although relatively minor and restricted to the bottom left corner, it was still visible at f11 in some scenes. The bottom left is the most irritating because it tends to cover the ground (where there is detail) and when you rotate the camera into portrait format it becomes the bottom right corner, which still covers land. Top right by extension is the least bothersome.
As for how to test, I find some nice flat subjects like brick walls or fences, ensure that I am as flat on as I can be and shoot some frames from wide open and stope down to about f11. I then repeat this with some new fences/walls so that any error in how I align he camera becomes randomised. If I have any suspicions, I may also find planar subject like street curbs and check the corners/edges that way. I will also shoot some real life landscape type scenes where the corners cover subject matter with detail (grass, tree branches/leaves) and check the corners. Often its very obvious that the lens is good/bad with the brick wall tests, but bear in mind not all brick walls are perfectly vertical and straight!
With the Canon 16-35mm f4 IS, the decentering was only visible from 16-27mm or so then vanished. Equally, it was little visible on close in subjects (and you have to be close to shoot a wall at 16mm), but all too visible at distance (where the bottom left corner was not close to the camera). In shooting the far street curb from the other curb, the lower left was clearly much softer from f4-f8, still softer at f11 and in all cases looked lower in contrast. Its such a shame, because the lens was amazing in every other respect.
Once you are used to checking over lenses it takes only ten minutes or so to do the shots, but if you notice a problem it can take a round or two of more shots to really pin down what is going on and at which distances and apertures it is visible.
I have never detected decentering in a Leica M lens, or Zeiss ZM, but they are the only lines where I have not. Typically, Canon seems to be very good and my worst experiences have come from Sony Zeiss lenses. I have often wondered whether mail order and poor packaging contributes to decentering, as boxes are thrown around in the delivery system.
With my first 55mm f1.8 FE, the decentering was so bad, when the right side was focused at 15m, the left focused at about 5m! The next copy is probably the most uniformly sharp optic I have ever owned.
Good luck, don’t get paranoid, but do check out the edges and corners before you shoot anything important!
Thanks for the detailed reply!
Fortunately there was about an hour or two of dry weather before the sun went down today, so I managed to shoot several brick walls (and a couple of normal scene test shots).
I’ve been staring at the results in capture one for over an hour now, and I’m starting to go cross eyed. I’m not sure if there is anything there – I think the top right corner might be a slight (and I do mean tiny) bit less sharp than the other corners at 2.8. All the real world shots (anything that isn’t a brick wall) I took look sharp, the center frighteningly so. I’m liking the A7, very impressed with the files.
I do wonder if Sony have stepped up their quality control – my lens did come with a hand signed quality control card (maybe they always had this).
I was planning to pick up the 55mm when I was in Japan in the next month or so (taking advantage of the week yen), but now I might just pay the extra £200 an get it from amazon.co.uk for ease of return – hmm…
The forecast is for slightly better weather tomorrow, so hopefully I can get out and give it another try- if nothing jumps out I think I’ll keep it. Thanks again for the help. Cheers…
Thanks for an excellent review!
I’ve tested lots of analog vintage FF lenses (on A7) including Canon, Leica, Minolta, Nikon, Konica etc.
When it comes to 35mm f/2.8 one of the very best I found is Yashica ML f/2.8 35mm. It is optically very close (95%) to the FE Sonnar f/2.8 35mm in sharpness and contrast even in the corners. Costs 1/15th the price of the Sonnar. Highly recommended if you’re on a budget and can accept manual focusing and aperture setting.