Sony 35mm F2.8 - asymmetrical sharpness

Started 10 months ago | Discussions thread
Flat view
Marc Heijligers Contributing Member • Posts: 589
Sony 35mm F2.8 - asymmetrical sharpness

I have tried 5 different samples of the Sony 35mm F2.8 Sonnar FE lens over a period of 2 years, and up till now I didn't manage to find a lens with edge-to-edge sharpness. The goal of this thread is to share my experiences, and to ask other people about their experiences and tests. I hear and found out myself that Sony lenses do suffer from asymmetrical sharpness issues.

The purpose of my 35mm f2.8 lens is that in combination with an A7<zyz> it creates a nice compact walk-around system, with the potential for an extreme good photographic quality. I use this setup for instance durig holidays, for portraits, paintings in museums, cityscapes, and landscapes. Because of the 42Mpix on my A7RIII, I have the opportunity to crop to ~10 MPix, and still have good full-screen resolution without having to carry along bigger tele lenses. Both the cropping capability as well as the city-landscape photography imply that corner-to-corner sharpness is an important aspect for me (in general for <= 55mm >= f1.8 lenses).

I've tested lots of lens types and samples under different conditions (some are reported here, test conditions here), and was always able to repeatable re-produce the issue. I will mention one of my tests here.

The idea of this test is to photograph a cityscape. The objects (buildings and trees) are "far" away (say 100m), which means for a 35mm lens the DoF (depth of field) @ f2.8 is in the range of 12,5 meter to infinity (f5.6 - 7m+, f8.0 - 5m+, f11.0 - 3,5m+). This excludes problems of not having a camera fully perpendicular to the objects. A screen-dump of the full example is shown below. I've adapted similarly the histogram and sharpened the pictures (+50%) in Lightroom, to make the sharpness issues clearly visible:

The sunshades, curtains, trees, and small metallic objects in these scenes make it easy to judge overall sharpness. As the late afternoon winter sun was out, I could also benefit from fast shutter speeds.

In this case, with the same lens I take 2 photos. One normal, and one with the camera turned upside-down. This reverses the bad side of the lens (normally visible on the left side of the frame) to the other side of the frame, making comparisons under the same conditions easy.

In the following screen-dumps, the left photo is with the camera in normal position, the right photo with the camera upside down (the fun is that it fools the exposure setting of the camera, so the one upside down is a bit faster, and therefore as a bonus also easy to recognise).

The first screen-dump shows the scene in the middle, at the point of focus. You can see both pictures are very (impressive!) sharp. If you start pixel peeping the right one is a tiny bit less sharp (see the Ikea lamp inside the house). This is due to focus variations, I made many pictures, and these minor variations can be seen throughout all pictures, but not impacting the analysis below.

Comparing the left side of the scene for both sides of the lens shows a significant difference in sharpness.

Turning the lens upside down, transfers the unsharp region to the other side of the frame, hence we can conclude the lens-camera setup suffers from asymmetrical sharpness (due to skewed/tilted lens elements or bajonet:

I tried many samples of the 35mm, and depending on the sample, the unsharp region is at the left or right side (2 were on the left, 3 on the right side), in different magnitude, the one shown here is the best. I had similar issues with the 28mm F2.0, 55mm F1.8 and 90mm F2.8 macro, but for those lenses I could (with help from my dealer) find good samples with variations that are so small that they become irrelevant for practical photography.

I will save you all the screen dumps, but one example below shows the difference between 2 samples (crop on the left side of a photo from a bookshelf, camera tripod mounted, the only difference is the exchange of the lens). The left side of my lens shows blurriness, the other lens is sharp, but has a similar (and larger) issue at the right side.

Back to my own lens, stopping down the lens does improve the sharpness a bit. The screen-dump below shows on the left side the lens at f8.0 on the bad side of the lens, and at f2.8 at the sharp side of the lens. As you can see, the f2.8 side is still better, the f8.0 becomes "reasonable" (e.g. look at the texture of the alu foil).

The problem does reveal itself at practical situation, e.g. look at the photo I made from Castelsardo in Sardinia, where the buildings on the left side become blurry as compared to the right side of the frame.

Full photo here:

Honestly, I think for a €799 lens this is below the level of expectation, and my recommendation is to always buy at a dealer with a good return or exchange policy. Unfortunately, I never managed to get a good copy of the 35mm, so if anybody has a tip, I would be happy!

 Marc Heijligers's gear list:Marc Heijligers's gear list
Sony RX100 III Sony a7R III Sony FE 35mm F2.8 Sony FE 55mm F1.8 Sony FE 28mm F2 +5 more
Flat view
Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow