Comparison of Zeiss, Nikon and Leica 50 mm on Sony A7ii

[Ed_Ingold]

<p>I took some time this afternoon to make some detailed comparisons between various 50mm lenses using a Sony A7ii. The images are full sized so you can see pixels on a 1:1 basis, but too large to view directly in this forum. The lenses were a Zeiss Loxia 50/2, a Nikon AIS 50/1.4 and a Summicron 50/2 v2 (1964). I used Novoflex adapters for the Leica and Nikon lenses. Each lens was set at f/2, f/2.8, f/4 and f/5.6. I tested the Nikon at f/1.4, but did not include it in this presentation. At this time, there was no point stopping down further than f/5.6, which stands at about peak performance for each lens.</p>

<p>Except for cropping and selection, no processing was done aside from Lightroom defaults and format conversions, starting with the raw images.</p>

<p>http://www.photo.net/photodb/folder?folder_id=1081662</p>

[AlanKlein]

<p>What are your interpretation of the results?</p>

[Ed_Ingold]

<p>In terms of the overall effect, the Zeiss Loxia 50/2 leads by a wide margin, followed by the Nikon AIS 50/1.4 (at f/2.8 of smaller), then the Summicron 50/2 v2 (1964).</p>

<p>The Loxia 50/2 is very uniform from corner to corner, falling off only slightly at f/2. There is almost no vignetting, but the camera firmware may be correcting the raw file (I did not remember to turn it off). Only the Loxia communicates with the A7ii. Image stabilization was turned off, and I used a remote release and electronic first shutter (which may introduce some smearing in bright light, according to Sony). I allowed all of the metadata to pass through.</p>

<p>The Nikon 50/1.4 is reasonably sharp over the entire frame at f/4 or smaller. There is some edge smearing wider than that, and f/1.4 (not published) looks almost surrealistic - very soft even in the center.</p>

<p>The Summicron 50/2 shows smearing in the corners and edges at f/2, which doesn't disappear until f/4 or smaller. The center is very sharp, comparable to the Loxia. I was pleased with the Summicron until performing this test and looking at the edges. Keep in mind this is the second version of the Summicron, which I purchased in 1964. That said, Steve Huff published a comparison between the Loxia ($1200) and the Summicron APO ASPH ($7500) without a conclusive winner. The tree line is about 250 yards distant. I can't see individual leaves, yet even the Summicron resolves them wide open. This was a very demanding test.</p>

<p>In conclusion, Zeiss makes good on their claim that their lenses are designed to deal with the short flange distance and thick cover glass of the sensor. If you like manual focus lenses, the Loxia delivers (also the 35/2) about the best the A7ii has to offer. I don't have one, but the Sony-Zeiss 55/1.8 AF lens is supposed to be sharper yet from corner to corner.</p>

<p>However, you would not be at a significant disadvantage if you use your existing Nikon or Canon lenses. Although I have not tested them as extensively as the 50/1.4, I'm generally pleased with results using Nikon 20/2.8 and 24/2.8 primes, as well as a 300/4 AFS. Sony has excellent focusing aids, both peaking and magnification, which makes even AF lenses easy to focus manually, despite their short throw and loose feel. Unless you are shooting landscapes, the Summicron is very sharp and crisp in the center, almost equal to the Loxia. I wouldn't buy Leica lenses just to use with the Sony, however.</p>

[Gus Lazzari]

<p>These results are similar to my own testing between Leica 35mm f/2 Summicron lenses (Ran a test to decide which lens to take on vacation). The modern (1999) ASPH vs. the 8 element first version (1958).</p>

<p>The ASPH was just as "clinical and sanitary" as this <strong><em>Zeiss Loxia</em></strong>; basically you only have exposure control when adjusting the aperture on them.<br /> On the other hand with the early 8 element, you instead have a very predictable <strong>picture and sharpness control,</strong> along with exposure adjust.</p>

<p>I decided against "clinical" & instead on the much better "Bokeh", color and versatility of the 8 element.<br /> Finally, especially because it's such a bargain, the Nikkor results were very impressive. Thanks for the testing <strong><em>Edward</em></strong>!</p>

[Ed_Ingold]

<p>How do you judge bokeh when everything is at infinity? At best, bokeh is a subjective quality.</p>

<p>The subject in these test images was about 200 yards distant - essentially infinity. (Novoflex adapters are slightly short (< 0.01 mm) to accommodate thermal expansion, so "infinity" is just short of the stop for the Summicron and Nikon. The difference can be easily seen in the magnified focusing view.)</p>

<p>Bokeh for the Loxia 50mm Planar is somewhat busy (to my eye, "geometric"), but not objectionably so. However it is always easy to get a pleasing contrast between the subject and background. This is a lens, much like Leica lenses, intended to be used wide open, giving a shallow depth of field, isolating the subject. Secondly, you can focus close (Loxia and the Summicron DR 50 to less than 0.5 m). Third, focus is very precise at any range with the A7ii, making the subject pop. Finally, the in-body image stabilization works will all lenses, Sony or otherwise. Camera shake is almost non-existent at 1/15 sec for a 50 mm lens.</p>

<p>The redbud photo was taken with a Loxia 50/2 at f/11, and the second photo with a Summicron 90 of the same mid 60's vintage as the 50/2, at f/2</p>

<p><img src="http://d6d2h4gfvy8t8.cloudfront.net/18045451-md.jpg" alt="" /><br /> <img src="http://d6d2h4gfvy8t8.cloudfront.net/18045449-md.jpg" alt="" /></p>

[david_smith110]

<p>Test less. Create more. :)</p>

[Ed_Ingold]

<p>Thanks, David, for the insight. However phototography has an unavoidable technical aspect which one must master in order for the creative side to shine through. Good equipment is also expensive, and most of us must spend in a way that yields the best results.</p>

<p>If, on the basis of these tests, someone buys a new Sony-Zeiss lens for half the cost of an used Leica lens and gets better results, they're better off artistically and financially. Likewise if they "recycle" good SLR lenses they've accumulated, and find the results are good enough to live with.</p>

[david_smith110]

<p>Edward, I completely agree with everything you said. I myself shoot exclusively with 'recycled' SLR lenses from the film era.</p>

<p>My comment stems from a personal feeling I have toward lens test in general and was not directed at you specifically. I believe lens comparisons to be a complete and utter waste of time. But they seem to proliferate on forums dedicated to photography and in many ways they do far more harm then good. There are three main reasons I dislike them.</p>

<p>1. <strong>They are unscientific</strong>- For the most part these comparisons are done in completely uncontrolled environments with little to no control over all the variable that can effect the results for each lens. From lighting conditions that change every few seconds, to changing wind speed and directions, from mistakes made in camera settings, to crappy tirpods, from how much alchohol was consumed at the beginning of the test vs the end....the list goes on forever. It could be that when the tester shot a photo on the first lens he nailed focus on some leaves where the lighting was good and there was no wind. But on the next lens several minutes later a cloud had covered the sun a bit and the wind had picked up. Maybe focus wasnt exactly perfect on this shot so now we might have some 'blurring' in certain parts of the photo they may have <em>absolutely nothing</em> to do with the lens. These are silly and anecdotal examples I know but in uncontrolled situations it is almost impossible to make a claim either way.</p>

<p>2. <strong>Sample variation</strong>- If both of us own the exact same old lens then the simple fact is that the results I get on mine may not be anywhere near the results you get on yours. We have no idea about what happened to each of our copies in the last twenty years. Uncentered elements, fungus, dirt, grease, internal damage we cant see. Again, to many uncontrolled variables can effect the outcome of a lens comparison and skew the results one way or the other. This may convince someone to not buy a particular lens which has the potential to become their favorite piece of glass ever and inspire their photography in ways we cant imagine.</p>

<p>3. <strong>It promotes gear centric photography over creativity</strong>- I agree completely with your statement concerning our craft having a unavoidable technical aspect. And the more I shoot thru the years the more I try to downplay that aspect every chance I get, especially with new or young shooters. Photography is one of the Arts that is, for better or worse, chained to technology in a way in which Dance or Painting is not. It is both a strength and a horrible weakness. Lens comparisons tend to focus so much on the elements of a photograph that are <em>not really important</em>. Corner to corner sharpness, center sharpness, bokeh, chromatic aberations, etc. The absolute worst thing we can expose a new photographer to are all the trivial little technical considerations that mean so little and can be overcome easily by creativity and vision. And the proof of this?</p>

<p><em>A good photographer can take an amazing photograph with a crappy lens.</em></p>

<p>If this is the case then we need to preach instead the value in learning photography...not specs. Instead of providing a row of images shot at every aperture with minute differences (that are scientifically uncontrolled anyway) we should instead show powerful and moving imagery shot with these lenses which will inspire the viewer to do likewise. In other words..."Hey, check out these images I shot with this old lens." "Holy wow man, those are great!". This is the dialogue we should be having in the community.</p>

<p>Again, this is not directed at you Edward. I see lens comparisons and tests on photographic sites everywhere across the web. And the lot of them do little or nothing to advance the creative or artistic portions of our Art.</p>

<p>I would like to thank you for the time you did invest in getting the test together however. At least you are putting effort into advancing our craft. :)</p>

[Ed_Ingold]

<p>I took time from other projects on this particular day because it was cloudy with almost no wind. There was ample light, and all of the apertures were fairly open, keeping the shutter speed at 1/1000 or shorter. There was no leaf motion of any consequence. The string of images was obtained in about 15 minutes. I prefer natural subjects because I'm partial to landscapes, and there is an infinite depth of detail in the subject. IMO, the test is suitable for comparing these three lenses in a reasonably objective manner.</p>

<p>One aspect of this test, not cited, is the effect of dark objects against a bright sky. It is interesting to note that there was very little veiling flare in any of these lenses, except for the Nikon at f/2.8 or wider. The Zeiss Loxia has almost no flare, even directly into the sun, and the Summicron is nearly as good.</p>

<p>Corner detail is important, IMO, for landscapes but much less so for portraits and portrait-like subjects (e.g., statuary). All of the test lenses are sharp in the center of focus, but the Summicron is exceptional in its ability to make the subject pop. Leica lenses are designed so that the circle of confusion expands more rapidly fore and aft of the plane of focus than most lenses.</p>

<p>I include a shot taken directly into the sun with a Loxia 35/2 and the A7ii, without any special processing, showing absence of flare and good blacks against a bright backdrop. I had intended to make a bracketed HDR of this scene, but the first exposure was good enough on its own. In the original, you can easily see the rigging lines in the sail boat.</p>

<p><img src="http://d6d2h4gfvy8t8.cloudfront.net/18046657-lg.jpg" alt="" width="700" height="467" /></p>

[mark_amos]

<p>Edward, I appreciate that you did the test. I want to add that so much of what David said makes sense, but the fact is that I like testing lenses. I always take the results with a grain of sand, but I like doing it just to see what I might discover. Sometimes I have to conclude that my test was completely inconclusive. I like comparing things and analyzing qualities of things. We can do that through photography and we can also do that as technicians realizing that there are factors of error.<br>

About your test more specifically, I believe is more worthwhile than many lens tests because we are still in a period of unknown compatibility for the Sony e mount. Like many of you, I have lots of Nikon and Leica lenses. Knowing I can mount a lens with an adapter is one thing; knowing that the lens design is at least more or less compatible with the camera design is another. I'm still not sure what to think about my Leica lenses on Sony. Bottom line is I don't want to sell my lenses, and I'm using them on a 5n. I am very tempted to buy an A7, but I'm not going to waste my time if my current 50 cron smears on the edges with it even at 2.8, so how am I supposed to find out without a lens test from somebody? I have typically heard everybody say that the 35 and up are fine on the A7. Of course "fine" is relative. Edward, you previously said you were happy with your 64 50 cron before the test. Do you think it is "good enough" on the A7? To me, good enough means you feel good about using it and would not be wishing you had the loxia (but I guess you do)?<br>

And Gus mentions the 35 cron aspherical having a clinical look. Doesn't it smear too? I've read you even get smearing of that lens on the a6000, no? I have that lens, but again, I don't know for sure if I would be disappointed with it on an a7 or not?<br>

I must further add, I occasionally use my Nikon 20 2.8 ais on my Sony 5n, but I felt it was too big on the tiny 5n, so I finally bought the Sigma 19 2.8 for e mount. I tested the Nikon ( a first class lens) against the $200 Sigma. The Nikon only sometimes seems just maybe a tad more sharp in the middle, but the Sigma is clearly better than the Nikon moving outward and way before the extreme edges. It is no contest. My point is that with the 20 at 2.8 and 4 smearing like this on my 5n while cropped, I can't imagine how it has been pretty usable on the a7II for you?<br>

So to David, my ongoing uncertainties are partly evidence of the problems of lens testing and relative evaluation, but what am I to do? Buy the camera and use my lenses and return the camera If I'm not happy. I say right off that if I had no equipment at all, and I bought a Sony body, I wouldn't be buying Leica and Nikon lenses...with a few exceptions, but I don't do ebay and KEH's 60% type offers don't impress me, so I'm just stuck with some of the best lenses ever made.....nice problem I guess. (The 5n has been a remarkable solution because other than the Nikon size on that tiny body, I can use all my lenses without issue on it. I prefer the Leica lenses because of their size and manual control features compared to the Sigma lenses, but I have to admit that even the current 50 cron feels heavy (too heavy almost) compared to the Sigmas.)</p>

[Ed_Ingold]

<p>Compatibility with the A7 (and siblings) depends primarily on back focus distance, hence the surprising results using Nikon SLR lenses. I, too, have a 20/2.8 AIS which does quite well. My 24/2.8 doesn't fare quite as well, but it isn't a great lens on the D3 either. I have three f/2.8 zoom lenses - 17-35, 28-70 and 70-200 VR - which do well, but are simply too huge to meet my other goal of reducing weight and bulk of my working kit.</p>

<p>If you have a Leica lens or two, don't hesitate to use them on an A7. Even with a 35mm Summaron f/2.8, problems are relegated to the extreme corners of the image, and then only if you pixel-peep as I did in this test. Overall, the results with M lenses are stunningly sharp, but especially in the center of the image, which is usually the only important part. These same lenses work nearly perfectly on an M9 (full frame, 18 MP) because the sensor is designed for these lenses, using an optimal array of micro lenses. Leica rangefinder lenses are small, unobtrusive (i.e., non-threatening) and superbly designed for manual focusing. There are also adapters for Leica R lenses (SLR), which are also manual focus. I never had an R camera, and have no interest buying their lenses just for an A7 - size and weight considered.</p>

<p>The M9 offers firmware corrections for particular lenses, but I have found it unnecessary, and a serious bother if you change lenses much. (Kits to code older lenses don't work very well nor consistently.) I made it easy to set the focal length in my A7ii for image stabilization by programming a key (waste can) for access to that menu item. When Steve Huff tested the Summicron APO on an A7ii, the images were much sharper in the corners than an earlier 50mm 'Cron, but not quite as sharp in the center. Remember, the comparison is jaded by our expectations from digital Leicas and the A7.</p>

<p>Would I still buy the Loxia 35/2 and 50/2? Most certainly, having seen the reviews and now having experienced their use for thousands of images. I like manual focus for most of what I do. I can take multiple shots (and video) without risking focus-hunting. The lenses are small (2" x 2"), only slightly larger than the 50mm 'Cron, and the focusing is smooth, with a long throw for precision. Lens settings are recorded in metadata, and the aperture is displayed in the finder. Loxia lenses are sharp from corner to corner, at a reasonable price. If Zeiss had introduced a Loxia 25 and 85, I would have jumped. Instead, they introduced the Batis autofocus 25/2 and 85/1.8, which will pretty much complete my kit. The 85 overlaps my Sony 70-200/4 zoom, but that lens is 3" longer and heavier. In my Leica days (1964-1990), I was happy with a 35/50/90 setup, which still meets my needs 80% of the time.</p>

<p>I have a Summicron 90/2 which works perfectly on the A7. However it was never as sharp as the 50 and operates a little too slowly to handle a moving subject with such a shallow DOF. It is great for video because it is manual focus and a lot sharper than you need for 1080p. 90 is also a good length for portraits and candids. I also have a 135mm f/4 Tele-Elmar, which is very small and exquisitely sharp. The 135 is a keeper for video, candids and closeups. It focuses down to a 7"x10" area - perfect for flowers and butterflies.</p>

<p>This shot was taken with a Leica M9 and a Summaron 35/2.8. I haven't taken a single shot with my D3 since this walk through the Botanic Gardens in Chicago.</p>

<p><img src="http://d6d2h4gfvy8t8.cloudfront.net/17865940-lg.jpg" alt="" /></p>

<p> </p>