Carl Zeiss 50mm f/1.5 C Sonnar

Concept: 4 out of 5
Execution: 4 out of 5
Yeah, but: There's no middle ground on this one.

The Long Version: Carl Zeiss makes good lenses. In the same way that exceptional photographers can create wonderful photographs that would look like mistakes if they came from a novice, Zeiss has intentionally broken the rules. They could have created a technically perfect 50mm lens, but instead they chose to let this Sonnar be a different kind of beast.

This isn't just another lens. The M-mount C Sonnar T* 1,5/50 ZM is almost the exact opposite of the Zeiss 2/35 Biogon that I reviewed last month. The 35 is descriptive and impartial, while the 50 is stylish and opinionated. The Biogon's crisp technical abilities are normally reserved for telephotos, but the only 'mainstream' lens that I can compare the arty C-Sonnar to is Canon's 50/1.2L. It's fussy, temperamental, and demands a lot from the photographer. This isn't just another version of the typical 50mm lens any more than a fiddle is just a sexy version of a banjo. It needs to be used a different way.

But when it's wide open the ZM 50/1.5 lens looks like nothing else I own.

The key to the 1,5/50 C Sonnar is its uncorrected spherical aberration. Light from the periphery of the lens slightly misses focusing on the film plane, diffusing the peak crispness that photographers generally look for. Widely considered to be A Bad Thing, spherical aberrations are usually squashed by modern lens designs.

The spherical aberration of the "C-stands-for-Classic" Sonnar makes it behave radically differently when it's wide open versus the smaller apertures of f/5.6-16. Stopping down cuts out the light from the periphery of the optics, and turns the C-Sonnar into a typical lens with very good resolution and consistent sharpness. Wide open, or close to it, there's a gentleness to its images that becomes very apparent with high-contrast edges. The result is almost a mild soft-focus effect, giving images from the 50/1.5 a certain gentleness. That's one of two excellent reasons to use this the lens at f/1.5.

Another effect of spherical aberration is a darling trait called "focus shift". This means that the plane of sharpest focus moves farther away as the aperture is stopped down – the lens simply doesn't focus in the same place at f/1.5 as it does at f/8. So in addition to the margin of error within the rangefinder camera and user error – the typical source of focus issues – there's also a bit of willfulness in the lens to deal with. Naturally, focus shift is most significant at close distances in addition to changing its severity according to the aperture, so there's no automatic formula to compensate for it.

But in practice focus shift is easy enough to tame; I learned what I needed to know from the Photoschool review of this lens and about five minutes of practice. Simply focus at the first part of the subject that needs to be sharp(ish) and expect the DOF to fall behind that point, rather than depending on the usual 1/3-2/3 distribution to catch things in front of the focus point. While focus shift never goes away, it moves back more slowly than depth of field increases, so that trick's effective at any aperture.

My ratio of out-of-focus images is the same as with my 2/35mm. I don't use the 1,5/50mm at apertures between f1.5-5.6 unless I absolutely have to; I'm not exactly afraid of them, but within that range the lens is neither fish nor fowl. I know exactly what it looks like at f/1.5 or at 5.6 and smaller, and those two options have such distinct results that choosing between them isn't a tough decision. I also avoid using this lens at very close range, where the shift can be problematic, because the bokeh/OOF blur isn't particularly wonderful. Problem(s) solved.

But remember that focus shift isn't a defect, and there's no point complaining about it when the underlying spherical aberration is the raison d'être for the lens. There's no need for the nonsense of finding a "good copy" of a Zeiss 1,5/50 or Canon 50/1.2L. Would-be photographers who "test" these lenses by photographing internet printouts at close range – typically hand-held – simply don't understand what they're buying, and should look elsewhere for something to spend their money on. There are plenty of more suitable options: The Zeiss 2/50 Planar, Canon 50/1.4, or perhaps a nice collection of fancy goldfish.

One historical characteristic of the Sonnar design is flare resistance, and the addition of modern coatings doesn't hurt. This photo (larger) is stopped down to f/16, the C-Sonnar's minimum, while the first photo in this review (larger) was wide open. While the f/1.5 shot looks like it has veiling flare, I suspect that the effect is actually the bright sun being diffused by spherical aberration. The photo that's fully stopped down lets us count the aperture blades (ten, two lost in the bright sky) but doesn't show any ghosting. I would love to hear the thoughts of the many people who know more about this than I do, but for my purposes I won't hesitate to include a bright light within the image as long as the lens is clean.

Using the Sonnar as a low-light lens would make excellent use of its flare resistance, but all of its wide-open aperture considerations still apply. Sometimes the extra 'sonnar glow' is helpful, and can have wonderful results, but the fast aperture isn't enough to make it an automatic choice for available-light photography. I'm just as likely to choose the 35/2 Biogon, which can be hand-held in the same light, but starts with the presumption of sharpness. As with every occasion to deploy the 50/1.5, it's important to use it for the things that it will be good at. Just because Carl Zeiss's marketing department calls this a "Fast and Compact Photojournalist" lens doesn't make it true. They also say that the "C" designation simultaneously stands for Classic and Compact, and by rangefinder standards only one of those is true.

And because I can't let an entire lens review go by without mentioning it, I have to say that I can see barrel distortion in these street photos. Look at the building column on the right side, the one that has the 'don't walk' hand in front of it: the effect is subtle this far from the picture edge, but the top and base of the column aren't going in the same direction. Most reviewers would say this is "not field relevant" while I'd usually care quite a lot about it. But even for me, it seems churlish to love the 1,5/50 because of a huge uncorrected aberration but then fault it for another minor one. This is not a lens for technicians, it's for artists – both the lens and I are outside of our usual realms when we get together, so some accommodations need to be made.

Physically the 1,5/50 is similar enough to the 2/35 that I need to look at the numbers to know which lens is on the camera when I pick it up. The smaller frame through the viewfinder means that the 50 doesn't intrude much into the scene, but its slightly greater girth means that it takes a different bayonet-mount hood. Zeiss's lens caps remain their own corporate humility block, so I use a micro 'hood hat' to cover my lens when it's not in use. Following Lee's suggestion in a comment on my 2/35 review, I've also found that a Nikon 62mm lens cap clips onto the front of the hood quite nicely.

It's worth noting that the 50/1.5 has a Leica-friendly (and Rockwell-approved) 46mm filter thread, which also matches my Panasonic 20/1.7, rather than the 43mm that the 2/35 and 4/85 use. I don't use 'protective' filters – those who do shouldn't be switching them from lens to lens anyway – but it would have been nice to be able to use one set of colour contrast filters across all of my Zeiss lenses. Still, buying a 43mm yellow-orange for my architectural 2/35 and a 46mm red-orange for my pictorial 1,5/50 isn't a huge sacrifice in exchange for having the lens be exactly the size it needs to be.

As with all Zeiss lenses, aperture control is in thirds-stops, with the exception that there's only one intermediate stop between 1.5 and 2.0. The unusual f/1.5 aperture is really just a marketing echo of the 1930's design; the 2.4mm difference between the theoretical f/1.4 and f/1.5 is my opportunity to finally use the classic reviewers' phrase "not field relevant". The actual reduction in light transmission is less than what lenses with more elements or inferior coatings can lose, and only cinematographers really need to care about that.

I confess that I'm smitten by the 1,5/50, and I should be the last person to fall for such an intentionally 'artistic' lens. I'm clinical in my photos and cynical in my opinions, with no patience for mysticism and mystique with my dozen cameras. I don't do portraits. But I'm also not obsessive about sharpness and bought my film rangefinder specifically for its timeless character and exceptional lenses. With an optical design that's inspired by the best from eighty years ago, the Sonnar is the closest expression of that goal that doesn't need to be bought on eBay and then sent for a CLA.

While I'm generally opposed to the idea of needing a special effect validate a photo, using the Sonnar wide open can create an image worth taking. The effect isn't Flickr-obvious, and looks nothing like something smeared or stretched in front of the lens, but the gentle rounding of the photos adds a certain something in colour or monochrome. It's not suitable for every subject, of course, but stopping down to f/5.6 gives a lens that's as crisp as anyone could need. I'm guilty of using a three-stop ND filter in daylight, which lets me use the entire aperture range and choose which temperament I want the lens to display.

If the idea of buying a lens that's soft wide open and hard to focus accurately doesn't make much sense, don't do it. The Classic Sonnar is a lens that most photographers shouldn't buy, but there may be a few people whose eyes will be sparkling at the thought of having one of these to play with. Many of those people will already have rangefinder cameras, or at least I hope they do, because this is a rare and special gem that we're lucky to have. It's absolutely not a commodity lens and it's nowhere near the marketing dream of the minimum acceptable standard. Instead it's an example of the artistry within lens design, and must have been designed simply for the sake of creating something wonderful.

If the 2/35 Biogon was the reason why I bought the Zeiss Ikon rangefinder, then the 1,5/50 C-Sonnar is its ultimate justification. Like the 2/35, it gives me qualities and capabilities that simply aren't available elsewhere. Completely opposite capabilities, at that – how often can two similar-seeming lenses, that handle perfectly as a set, and have such well-suited focal lengths, provide such a range? The Carl Zeiss 50mm C-Sonnar is far from a perfect lens, and I wouldn't want it any other way.

last updated 8 nov 2011