When I tested the Sony 35mm f/1.4 Distagon, I found it to be skewed on the right side, blurry and weak there. Since then, I’ve had a number of reports of the same issue, which I deem credible, including one user who tested several samples. One has to be very careful with such things (not leaping to conclusions), since the mount flange-to-sensor parallelism could be off as another possibility (especially with high resolution digital cameras, be very, very careful to avoid placing undue stress on any lens mount, e.g. excess weight or any torque, or shocks).
Along comes the Sony 90mm f/2.8 Macro G OSS. It’s clear that the optical design is strong, but a lens has to be built to tight tolerances to perform (this is part of what makes Zeiss Otus so strong—stringent quality control). Last night I shot it on the A7R II, and every image at f/2.8 showed a blurred/weak right side. So I went and shot it again this morning on an outdoor scene (mosaic) along with the Sony 55/1.8. The 55/1.8 is sharp and symmetric wide open at f/1.8; the entire right side of the Sony 90/2.8 is blurred/smeared at f/2.8, just as with last night’s images. And the scene (very large mosaic) does not admit to photographer alignment errors. The loss of sharpness is what I call “at least an f/8” error, meaning that even f/8 cannot quite hide the loss of sharpness. But it’s more than sharpness; the effect is of smearing and with a reluctance to sharpen up with stopping down.
Three examples at close, medium and far distance are shown here, as I think this sort of thing is worth seeing documented, helpful to anyone wondering about what such issues might look like:
I’m having B&H send me a replacement 90/2.8, so I can have it in time for some field shooting in a few days. I am very glad that I checked out the lens beforehand! Which is my advice to anyone buying any lens. See How to Test A Lens in Making Sharp Images.
How can a lens be bad if it is tested properly before shipment?
Note that the Sony 90/2.8 is NOT a Sony/Zeiss collaboration. So no testing requirements would hold as per this recent post by Zeiss: Sony and ZEISS: What photographers should know about the partnership, which states:
ZEISS lenses are developed exclusively by ZEISS. ZEISS also determines the features of the lenses, such as their focal length and internal construction, in line with the company’s product strategy. Sony/ZEISS lenses are jointly developed by ZEISS and Sony. ZEISS supports Sony throughout the optical design and development process and then tests and approves the prototypes. Finally, ZEISS determines the test specifications for serial production.
Sony/ZEISS lenses are manufactured by Sony in factories across Asia. In these factories Sony uses lens testing equipment that has been developed and manufactured by ZEISS. An example is the MTF-tester K8, a versatile and compact instrument used to measure the modulation transfer function of photographic lenses, or similar systems, at an infinite object distance.
In all its brand partnerships, ZEISS sets the technical and quality standards to which partners, such as Sony, must adhere. ZEISS regularly audits the production process of Sony/ZEISS lenses in Sony’s factories. ZEISS experts examine the production processes, management systems and measuring installations. ZEISS certifies the suppliers and provides all the equipment that is needed to ensure product quality, including the equipment of third-party suppliers.
What are Sony’s standards? The Sony/Zeiss 55mm f/1.8 Sonnar showed beautifully excellent sharpness when I shot it on the mosaic a few minutes after the Sony 90/2.8. It would be subject to the above standards, being a collaboration with Zeiss.
Nick C writes:
I saw this post on the bad sample of the 90mm lens. I am absolutely appalled by Sony’s quality control; at best the inconsistency, at worst the lack thereof, as I’ve come to experience the following, all within the last week:
1. Sony FE-PZ 28-135mm lens; an alleged “pro” level cine lens, exhibiting a skew on the left side; there’s nothing sharp in the leftmost 20% of the photos, regardless of the aperture (I haven’t bothered testing darker than f/8)
2. Sony’s FE 24-70mm, exhibiting front focus from 50mm onwards; manual focus is fine.
3. Not Sony, but worth mentioning, a brand new Zeiss 135mm ZF.2 via a Novoflex adapter (which is previously known to me as having perfectly parallel flanges as tested with a different Zeiss 135mm on an A7R) - tilted focal plane running left rear to front forward. At least things are in focus at some place across the frame.
All these were observed while mounted on a sturdy RRS tripod and ball head; IBIS was disabled.
I am just getting fed up with half-baked goods being dumped on us at ever-rising prices with a race to the bottom as far as quality control is concerned. Where does this stop? At Otus level?
DIGLLOYD: I’m surprised to hear a report on the Zeiss 135/2 APO. But in this case, the use of an adapter is not a fair test; put it onto a Nikon D810. [Nick reports that the adapter worked fine on a previous sample]. One issue with all lenses is that focusing can change things. In the case of the 135/2 APO, the long focusing throw cannot always maintain perfect symmetry throughout the range—a manual focusing helicoid is after all mechanical (and so is AF), and it’s very hard to maintain micron-accuracy while moving stuff around. But this sort of skew I have found with the much more expensive Leica 50/2 APO (two samples!) and is hard to avoid. One has to get lucky. [Nick reports that his previous sample was symmetric at the same distance]. Skew of this sort should be distinguished from poor optical assembly (decentering and such); the two behaviors are very different. With skew, the focal plane is not exactly parallel but quality remains (where focused). With decentering and similar, lens performance itself is compromised, regardless of accurate focus.
But the real issue is that Sony has crap for service. Caveat emptor—if something goes wrong, good luck. I have deep reservations about purchasing Sony products (including the A7R II), but I feel forced into it as a cost of doing business.
James K (former NYC pro) writes:
Nikon reliability is king. You do a shoot that is high profile and you immediately think Nikon-no BS in that camp. That’s why it takes so long for them to bring professional products to market. Testing, testing, testing……
Sony is not a professional product line.
DIGLLOYD: Agreed. Sony makes gadgets that work like cameras, with no service or support worth crap. He continues:
I see you are testing the Sony 35mm 1.4 again. I just tested a new lens and found the same problem on the right side of the image. My original serial number is 46200*** and the new one is 46204***. Same poor quality on the right side.
I have been retesting my lenses for the A7RII. I have come to the conclusion that Sony quality control is in the dumper.
Otus 55mm first, then the Sony glass. Otus to establish that the camera body was functioning properly and then Sony optics- 35mm 2.8 FE, 35mm 1.4 FE and 55mm 1.8 FE.
Not one of the Sony optics is symmetrical. One is out on the right (35mm 1.4 FE) and the other two on the left. Varying degrees of mush.
Zeiss is most certainly not properly monitoring the Sony production line. The only reason I own Sony glass is because it was the only auto focus glass available for the E mount.
I am waiting for an Otus 35mm.
DIGLLOYD: Mirrors what I’ve heard from readers. Troubling of course. The good news is that the Zeiss Batis line covers 25mm and 85mm with more to come. The brand-new Sony 35/1.4 Distagon I received on 11-Aug-2015 shows asymmetry just like the copy I tested in mid-spring; it is skewed forward on the right side.