3. Bokeh quality
There’s a lot of fuss made about lenses’ “bokeh quality”. Bokeh isn’t a matter of good lenses: The otherwise famously pristine Leica 50mm f/2 Summicron is infamous for its rather harsh bokeh.
But you can relax: Firstly, in general. Because in many situations, you will not be able to really tell the difference – in order for a lens to produce harsh bokeh (so that it’s eye-catching), you need to have something in the background that actually “challenges” this harshness. We will do this, no worries. Plus, even if you have a harsh-bokeh lens, the harshness is generally reduced when stopping down. In other words, if there are big differences, they usually take place in the wide aperture range.
Secondly, you can relax because the bokeh of all these lenses is very pleasing, as you will see.
So, are there no differences? Yes, there are, but the differences are in the details. We’ll go through this now:
Ok, so let’s do this step-by-step. One thing you have to know for all these comparisons is that I always compensated meticulously for the focal length difference by adjusting the distance, while the focusing point stayed the same. I don’t want to go into the details, but mathematically, this means the same depth of field at the same aperture, which is what we want here for the comparisons.
Check out this lovely hotel room scene:
As you can see, you can’t see much, which is because of the bokeh. Now, the difference between the lenses in this case is actually not so much in the character of the bokeh wide open: This is pretty much the same for all lenses, with the Nikkor 58/1.4G being indeed a tad more “creamy” than the others.
The difference is in how the bokeh “progresses” as the lens is stopped down – in other words, how fast the background “sharpens up” while stopping down. In the gallery below, you find images of f/1.4 and f/2.8. While with this motive, you can’t see a big difference in the f/1.4 versions, the f/2.8 versions differ quite a bit: The Zeiss bokeh is softer than the rest, followed by the Nikkor 50/1.8G, and then the others.
But: The 1.8G is NOT the second “creamiest” in the setup – it’s just at f/2.8. Sounds complicated? No it’s actually not, it’s just that every lens has a different “curve” of its bokeh smoothness progression. So if you’re not shooting at f/1.4, the right aperture for a desired background smoothing will be different from lens to lens. E.g., for the desired effect in this motive as displayed above, I had to use f/2.8 with the Zeiss, but f/2.2 with the Sigma and the Nikkor 58G and 50/1.4G, and f/2.0 with the Nikkor 50/1.8G.
As esoteric as this sounds, it’s quite simple, as you’ll see this in the next motive:
In order to see if these results are more than just an impression, I wanted to look at a more standardized motive that also provides a more “challenging” background.
Step 2: Standardized Motive (Sorry, Contains Leaves…)
As much as I hate having green on this website, leaves are just a very good way to compare the bokeh, because of the high frequency of highlights in the pattern. The following motive does a good job at illustrating what I said above:
When you look at the following comparisons now, in addition to the leaves, take a look at the renderings of the windows (black panels with white frames) and at the pattern on the wall going into the far on the left side of the picture. The object that served as the focus point was at a distance of 1 to 1.2 m (approx. 3 feet), depending on the focal length, the trees were approx. 4 to 4.2 m (9 feet and 2 imperial hedgehogs) away.
All images are not edited except for slight cropping for direct comparability. The images show each lens at f/1.4, f/1.8, f/2, f/2.8 and f/4 and f/5.6, and since there are 29 images to compare, the gallery only shows what I want to point out.
Instead of giving you thousands of comparisons, again you can download the complete package and compare whatever you want. For traffic reasons, they’re sized down and compressed, but it’s enough to compare (they’re the same files that are used on this website).
Before you start seeing any “magic” in an image, bear in mind that sometimes the light changed a bit, that’s why the tree is sometimes brighter. It’s not divine light or anything. Also, don’t get fooled by vignetting, either! I tested this with a couple of people and everyone always loves the images with vignetting “I don’t know why, but…”. Concentrate on the bokeh, not on the image.
- They all have beautiful bokeh that isn’t harsh at all.
- At f/1.4, the lenses are all very similar, one could say pretty much the same. And, though not in the comparisons (but in the download), I can tell you that they “meet again” at f/5.6, where they’re also almost indistinguishable. The difference is what happens in between, with especially two lenses standing out:
- The character at f/1.4 and f/1.8 is not very different for all lenses except the Nikkor 50/1.8G, which is “clearer”, but not harsh.
- From f/1.8 on, the Zeiss Otus is always a bit more unsharp, or “creamier” than the rest. This is especially visible comparing it at f/4 with the Sigma only at f/2.8, as you can find in the gallery. They look the same although there’s one stop difference! As said, all this difference stuff is over at f/5.6, where all lenses look pretty much the same again.
- When you compare in a standardized and adjusted setting, as in: compensating for the focal length, the NIkkor 58/1.4G is not overall “creamier” than the rest. It is at f/1.4, but only slightestly. However, from f/2.8 on, it looks just like the other Nikkors.
- Wide open at f/1.8, the Nikkor 50/1.8G is slightly “clearer” than the rest. But: This is reversed by f/2.8, where it is now creamier than the others, approaching the Zeiss character.
One thing that should be very apparent from all the images is that all lenses have a very pleasant bokeh quality. Seriously, as opposed to sharpness, which is easily identified, you will have a hard time telling the lenses apart by their bokeh only if you don’t have any other information about the picture.
Ok, last bokeh chapter, here’s a shot of bokeh balls (bright points of light that are out of focus) wide open for all lenses. Just so you see the pattern and the size relations, here’s a chart of all lenses:
As you can see, Sigma has the biggest balls. Yes, the sizes are a little different, but don’t make too much fuss about it, this is due to the different focal lengths and close focusing distances, because all these pictures were shot at the same distance, wide open, and at closest focus, which is different per lens, hence the differences. The same is valid for the shape of the bokeh balls: It’s normal that they are shaped more cat-eye style towards the outer edges of the frame. This is the result of vignetting by the lenses hardware, pretty much like if you get a vignetting effect from the frame of a filter you put on the lens. Again, the Sigma looks like it’s rounder and not prone to cat-eye shaping, but again this is the result of the Sigma being the most out-of-focus. (We’ve already tested the bokeh style above, so this one was only about the details of the light bokeh balls.)
You will want to look at those balls at a 100%, right? Here you go, this is from the middle of the frame, so you can see different light intensities at once (high-dpi 100% crops):
As you can see, the Zeiss and the Nikkkor 58/1.4G have onion ring issues, the Nikkor more pronounced than the Zeiss. Onion rings are a result of aspherical lens elements (they are basically the polishing marks of those elements), so it’s logical that the Nikkor 50/1.4G is not prone to them, as there’s no aspherical element in its lens design. All the other four, though, have aspherical elements, so you can conclude that the Nikkor 50/1.G and Sigma Art do a good job supressing them.
Can you see this in an image if not at 100%? Yes, you can, but of course it depends on the size of the light balls in relation to the rest of the image. And just in case: Onion rings are usually only present in relatively lower-light-intensity spots. Just like in the above examples, where you can’t see any rings in the brighter bokeh balls.
Ok, basta with the bokeh stuff. Next!
Go to next page 4: Sensitivity to bright light inside/outside the frame