Bokeh and Longitudinal CA

Here we're looking at longitudinal (axial) chromatic aberration (LoCA), which appears as green and magenta fringing on either side of the plane of focus. This tends to be most prevalent on fast primes and it difficult to remove so you want as little LoCA as possible if you're hoping to shoot wide-open. To perform these tests, both the Sigma 85mm and the Sony 85mm were mounted to a Sony a7R II, the former via a Metabones smart lens adapter. 


Initially, it does appear that both lenses handle LoCA fairly well, with the Sigma 85mm displaying a fair bit less LoCA at F1.4 than the Sony. The majority of the Sigma's LoCA has vanished by F2.5, but it still persists until around F4 in the Sony GM (although only slightly, the majority of it is gone by F2.8) where it is resolved completely.

However, upon closer inspection we see that in more extreme backlit situations the Sigma displays some fairly vivid purple fringing when shot wide-open at F1.4, which is only suppressed when you stop the lens down to F2.5. This was also confirmed in this real world sample image with similarly harsh lighting.


While shooting the Longitudinal CA test, we put a net of Christmas lights a few feet behind our Lens Align tool to create beautiful balls of bokeh, giving us a way to visualize differences between the two lenses' out-of-focus characteristics. Hover your mouse over any given aperture of any given lens to have the main image switch to a full-frame view of the resulting shot.

Both the Sigma and the Sony exhibit some very nice and smooth bokeh at F1.4, with little to no onion-ring or bulls-eye characteristics. The Sony offers slightly smoother bokeh at F1.4, and offers a slightly more blurred background when shot at the same f-number as the Sigma. The longitudinal chromatic aberration that can be seen on the fringes of the 'bokeh balls' when shot at F1.4 is comparable in both lenses with the Sigma performing slightly better than the Sony.

Sigma 85mm F1.4 HSM Art 1.4 2.0 2.5 2.8
Sony 85mm F1.4 GM 1.4 2.0 2.5 2.8

In the above roll-over you can compare the bokeh size shot at the identical aperture in each lens. As you can see, there really is only small differences between the two lenses. The Sigma exhibits a bit more cats-eye shape to its bokeh than the Sony, as evident in the corners and along the edges of the roll-over at F1.4. This 'cats-eye' effect disappearing by F2.0.

Bokeh and sharpness fall-off are very complex topics, it's difficult to form broad conclusions about the overall bokeh characteristics at any given plane, but this should give you a good idea of the overall performance and the characteristics of the bokeh.

Real World Bokeh Example

Subject isolation and bokeh are extremely important in portraiture, so the question is how do these lenses behave when challenged with a real model?  In the below roll-over you can get a better idea of how the bokeh and subject isolation behaves in each of these lenses.

Sigma 85mm F1.4 HSM Art 1.4 2.0 2.5 2.8
Sony 85mm F1.4 GM 1.4 2.0 2.5 2.8

As you can see there isn't a huge difference in bokeh between these two lenses. Subject isolation is also very nice with both of these lenses, but the Sigma exhibits smoother transitions between in-focus and out-of-focus regions. The Sigma is extremely sharp wide open. The left eye was used as a focus point for all of these images and the Sigma is noticeably sharper at F1.4 than the Sony. The Sony, however, offers slightly smoother out-of-focus regions at identical apertures, which replicates what we saw in the widget above. These images in this roll-over reaffirm our findings in the sections above.


As you can see the focus ring on the Sigma 85mm F1.4 Art is fairly large.

In terms of AF performance the Sony 85mm GM is generally slower to lock focus when comparing it to the Sigma 85mm mounted on a native Canon body (5DSR and 5D Mark IV), particularly while in live-view, using dual-pixel AF on the 5D Mark IV. Hunting can be a bit of an issue with both lenses, but that really becomes more of a problem in low light situations.