How does ZEISS define Bokeh – An Interview with Dr. Stefan Ballmann
Are you confused about what bokeh really means? In this interview, Dr. Stefan Ballmann talks about the different characteristics of the out-of-focus areas in an image and how and why lenses differ in the quality of this bokeh or blur.
Bokeh is a term that has prompted plenty of discussion and causes a lot of confusion. Learn more about how ZEISS defines bokeh in this interview with ZEISS research assistant, Dr Stefan Ballmann.
Bokeh is a term that prompts plenty of discussion in online forums. How do you define bokeh at ZEISS?
The term bokeh certainly causes a lot of confusion, and funnily enough it actually means “mental haze” or “blur” in the original Japanese! In photography, the word stands for the “confusion” of rays of light, in other words the way in which the light beams emitted by an object outside the focal plane do not meet in an orderly fashion at a single point on the camera sensor when they pass through a lens. That’s why those parts of an image appear blurred. Bokeh can therefore be used as a general term for all the properties of blur. For example, is the transition from the in-focus to the out-of-focus parts of an image hard or soft? How high is the contrast between the sharp and out-of-focus areas? How does the background blur differ from the foreground blur? And what kinds of chromatic aberration occur?
When people post on forums, they often use the word bokeh to refer to the shape and structure of out-of-focus highlight areas, sometimes referred to as aperture iris images. The appearance of these iris images is certainly important to the overall composition of the image, but it is only one of many facets of bokeh. We often come across sentences like “The rounded iris blades produce images with soft bokeh.” But scientifically speaking that doesn’t make sense, so it actually causes even more confusion.
What influence does design have on the bokeh of a new lens?
Influencing bokeh during the design phase of a lens is extremely challenging. Obviously factors such as the number and shape of the iris blades and the choice of lenses affect how the highlight areas look in parts of the image that are significantly out of focus. That is a key issue for us when we’re designing a lens, but it doesn’t tell us much about the actual bokeh. Without going overboard on the technical details, it is generally the case that lenses with spherical overcorrection tend to exhibit a harsher transition to out-of-focus areas. Lenses that are undercorrected for spherical aberration generally have a much larger depth of field, which creates a softer transition. Distracting double images – in other words sharp edges within out-of-focus areas – are regarded as “bad” bokeh. They are essentially undesirable side-effects that need to be corrected.
Does it make sense to quantify the properties of blurriness? Is that even possible?
Plotting values in defocus curves can give us useful information on out-of-focus areas, both during the design phase and using measurements taken from real lenses. But ultimately blurriness is a somewhat hazy term that primarily comes down to a subjective judgement. That’s why no set of scientific measurements can ever be a proper substitute for the visual impression of a photograph taken under realistic conditions. We appreciate that fact – and that is why we equip our photographers and brand ambassadors with lenses right from the start of the prototype phase so that we can gather as many images and appraisals as possible.
Why do people focus so heavily on the qualities of blur with 85mm lenses? And what kind of bokeh are you looking for in an 85mm lens?
An 85mm lens is a classic focal length for portrait photography, so bokeh is obviously very important. That is especially true of fast portrait lenses that exhibit a shallow depth of field at wide apertures. The idea is to make the subject stand out as sharply as possible against the blurred background. That gives an impression of spatial depth and focuses the viewer’s attention on the key part of the image. As a photographer, I strive to ensure that the contrast of the sharp part of the scene such as the face or eyes is nicely in harmony with the blur gradient and background blur.
How does that differ from a wide-angle fixed focal length lens, for example?
Due to their shorter focal length, wide-angle lenses exhibit a larger depth of field compared to normal and telephoto lenses (in situations where you have the same f-stop and distance to the object). That generally produces a softer blur gradient than you would get with a telephoto lens of a comparable speed.
What’s the difference between zoom lenses and fixed focal lengths in this context?
People often assume that zoom lenses will produce a less harmonious bokeh than fixed focal length lenses. But that’s not normally true, at least as far as modern zoom lenses are concerned. That assumption basically stems from the fact that the out-of-focus highlight areas of certain zooms are not considered to be as “aesthetically pleasing” as those of a fixed focal length lens. But it is true that prime lenses usually have a higher max. aperture than zoom lenses. Therefore, prime lenses show a shallower depth-of-field and thus demonstrate a faster transition from focus to out-of-focus areas.
What would you say are the key factors that influence the bokeh of a lens?
Let me use as an example, the new Batis tele portrait lens which was introduced recently. With its 135 mm it has the longest focal length in the Batis family. As a result, it has a shallower depth of field at the opening aperture of f/2.8 than the Batis 1.8/85 at full aperture. It is therefore ideal for separating the main subject from the background. In common with the other members of the Batis lens family the Apo Sonnar 2.8/135 has a harmonious transition between the focus and out-of-focus areas. The apochromatic correction also ensures a colour-neutral reproduction of contrast-rich transitions and conveys a particularly pleasant bokeh, free from colour fringes, as the pictures below show.
Depth of field of the ZEISS Batis 2.8/135 (Sony A7 with 24MPix):
Object Distance 1m | Depth-of-field 0.2 cm
Object Distance 2m | Depth-of-field 1.3 cm
Object Distance 5m | Depth-of-field 10.0 cm
Object Distance 10m | Depth-of-field 41.0 cm
How can a photographer know whether a lens has the potential to produce harmonious, soft bokeh?
As I mentioned before, lenses with spherical overcorrection tend to exhibit a faster transition from sharpness to blur, while lenses with spherical undercorrection make a softer transition. You can see that clearly in the MTF charts for each lens. You can also look at how the contrast changes between very bright and very dark elements in the out-of-focus areas of the image, for example by including a branch set against the bright sky in the background of a portrait photo. Some lenses produce hard edges in that situation, which are often regarded as rather artificial and aesthetically unappealing. The shape and structure of the aperture iris is also important. Currently, rounded iris images are considered to be attractive. The regular pentagons or hexagons that used to be so common are nowadays frequently felt to be “too technical”. With a sufficiently large number of rounded iris blades – for example eight or nine – it is possible to come close to the ideal of a circular aperture.
Do you have any recommendations on how photographers can track down lenses with very special types of bokeh?
We are often approached by photographers who have adapted their analogue lenses to a digital camera body and are over the moon with the bokeh of these sometimes very simple lenses. They may not have been designed for optimum image quality, but these old lenses often produce soft and attractive bokeh. Ultimately, though, it is all very much a matter of taste: bokeh that appears harmonious and aesthetically pleasing to some people may seem too soft and uneven to others. My advice is to simply try out a couple of lenses under real-life conditions, because you can normally judge what you prefer quite quickly. For example, ZEISS runs a website called lens4rent.de where you can hire out a large selection of ZEISS lenses.
About Dr. Stefan Ballmann
Dr. Stefan Ballmann has worked as a research assistant in the ZEISS research and development department since 2013. He studied physics at the University of Erlangen-Nuremberg as part of the Elite Network of Bavaria graduate program, completing his doctorate in 2012. In his free time, he is a keen portrait and nature photographer.
He prefers working with fast telephoto lenses because of their ability to isolate the subject from the background – and of course because of bokeh. For a more in-depth look at this topic, Dr. Stefan Ballmann recommends reading the article “Depth of Field and Bokeh” by Hubert Nasse.
This article was first published in German photo magazine c’t Digitale Fotografie, issue 1/2017.