Depth of Field is a Fallacy
Ahh, the good ole days of film. Life was so easy back then. Photographic principles remained relatively consistent for more than a hundred years. Emulsions, resolving power, and the technology of our film had normal gradual advances. These small changes had very little effect on how we viewed and calculated standard rules of photography. And then the digital sensor asteroid came crashing into our photographic planet and killed the film dinosaur. It has been the largest disruptive technology that our photographic medium has ever seen and has forced us to rethink almost every single aspect of our process. To cover the full disruption, it would need a Tolkien-esque multi volume novel. I am going to discuss just one topic today, Depth of Field.
My career in photography began as an employee of a view camera manufacturer. Thus subjects like DoF, plane of focus (PoF), and Scheimpflug’s principal have been near and dear to this nerd’s heart. We can get completely technical in discussing this topic but I don’t think that it is necessary.
DoF = The total distance in a photograph that is acceptably sharp starting in front of the plane of focus and ending behind the plane of focus.
Seems easy enough, but let’s look at it deeper. What is the plane of focus (PoF)?
PoF = A 2 dimensional plane in front of our optical system where focus has been achieved
Lets call this the “subject plane.” This is the same for any optical system, not just a photographic system. Our eyes work by shaping their crystalline lens to focus light from the subject to our retina. Our digital camera works by physically moving an optic to a point where the subject is focused on the sensor plane. Pretty simple right? This is such a basic principle that we truly never really have to think about it.
Here is where I want you to think a little deeper. There is just a single subject plane in an optical system. That means there is only one point in our photographic system that is truly sharp. Remember that the PoF is 2 dimensional. It consists of length and width, but does not have depth. In a focused image, any point on that 2 dimensional plane is sharp, but technically nothing in front nor behind it is. The concept of DoF only works because we “accept” unsharp subject matter as “sharp enough.”
So what is “acceptable sharpness?” With film it was easy. We viewed images by looking at prints (remember those). We had a closed loop system. We could only view them in print form as there was no other choice. DoF calculations used the factors of print size, print viewing distance, film capture size, magnification, and circles of confusion (CoC). We could use these formulas, albeit complicated, to consistently compare systems to each other. We accepted that these formulas and DOF were solid photographic principles that did not change and we built charts and scales to illustrate them. Now in the digital age, fortunately or unfortunately, those formulas and charts are all broken dinosaurs.
Why? My main problem is that today’s definition of DoF ignores resolution. And in my experience, DoF is dramatically reduced as you achieve more resolution in your system.
By definition, DoF is what is “acceptably” sharp in a scene. When resolutions increase in a system, subjects that were once “acceptably” sharp are now resolved clearly and become “unacceptable.” Your lower resolution systems did not have the resolving power to define its unsharp quality and thus we found it “good enough” and accepted it. With the higher resolution systems, the full area that we perceive as acceptably sharp in a scene naturally decreases and therefore DoF decreases. Its just logical.
Depth of field calculations of the past used film size as a parameter. This made sense since a larger area of film capture meant more resolving power and less magnification in the print. Online, you will find many calculators that use “sensor size” as the parameter to calculate DoF. And if all sensors were equal in pixel pitch, as film was equal in grain size, then this would work. But this is not the case. The same physical size sensor has multiple different resolutions today. Resolution is ignored and thus our calculations, scales, and charts are invalid.
Rochester, we have a problem
Kodak Ektachrome 100 film had a silver grain structure that averaged in size at about 11 microns. When you pack 100mp into a 33×44 sensor, your pixel wells end up at 3.7 microns. Quick and rounding math illustrates that our digital recording medium is today 9 times smaller than the film system we used just 15 years ago. Let me repeat, it’s a factor of 9! This means your full capture process needs to be 9 times more precise than the systems we used for film. And this is just one of the examples of the challenges we face.
I first noticed that we had a problem when our digital capture sensors achieved the 40mp resolution capability. At that point our testing and some quality critical customers proved that we had begun to have a system failure. We instantly lost sharpness and DoF in certain back/body combinations. The same systems that were “sharp” at 22 and 30mp resolution became “unacceptable” by just upgrading to a 40mp sensor. Our systems that were based on the technical quality of film began to fail us at that resolution.
Secondly, The viewing process has changed. Instantly, I can look at an image captured and view it at 400% on my 4k monitor. Immediately I can compare sharpness in any system. Lenses designed for film will begin to fail at certain resolutions. Customers will upgrade and see problems in their process that they had never seen before like camera shake, diffraction, and mirror slap. Importantly, these problems were ALWAYS present. It is the higher capture resolution that will test every variable of quality throughout your process. And when it comes to DoF, we are no longer the “standard viewer” at a print size and distance. We are now in front of a monitor and have become the pixel peeper. Anyone who conducts a balanced test can see that DoF is getting smaller and smaller.
Lens Scales and Hyperfocal Distances
DOF scales on lenses didn’t need to change for years. Film never had the large technical advances where resolution doubled overnight. Just in the last year, we moved from 24 to 40mp on my Leica M, 50 to 100mp with Fujifilm GFX, and 100 to 150mp with Phase One IQ’s. The fact is that DOF scales on all those corresponding lenses were designed with a single Circle of Confusion for the corresponding medium at one point in time. With resolutions of sensors changing so quickly, lens manufacturers can’t keep up. And photographers keep high quality lenses through multiple sensor upgrades in their career. The formula or scale that works today will not work tomorrow, thus rendering DoF scales as only useful for a fraction of their usual life.
This also applies to Hyperfocal distance calibrations. There are a few very technical image scientists that state we shouldn’t even use hyperfocal distances anymore and I understand their argument. However, with modern cameras we can calibrate our full resolving system easily and quickly through live view or tethering to our monitor. So it is important to remember that every time you upgrade your sensor, you need to recalibrate your hyperfocal distance for that lens for the new resolution of your system. Hyperfocal distances in our scene are getting closer and closer to infinity focus.
Agree to Disagree
Today there are articles, videos, and blogs that state medium format DoF is the same as DSLR. They state that the size or resolution of the sensor does not affect DoF. Let me unequivocally state that these arguments are completely flawed and wrong. They are online clickbait to enhance ad revenue by counting “likes.” They are for individuals who want to justify their choice in a lower quality system by saying “my camera is just as good.” I am sorry but I have to call these guys out. Its utter nonsense and not logical.
“If it’s on the web, it must be true.” “But these guys have such a polished website.” Yes, very few online posts or calculators will address my argument, but if you dig enough you will find the few that actually have the experience or have the technical background to admit there is a flaw in our system today. One of my favorite quotes I have found is: “Those who use depth of field scales, tables, and formulas (e. g. for hyperfocal settings), restrict themselves – most probably without knowing why – to the image quality potential of an average pre-World-War-II emulsion.” This completely backs up my experience and my findings on this subject matter.
I have been told that I have a black and white personality and it’s not because I shoot an Achromatic or Monochrom camera. A subject is either sharp or its not. You either shoot the highest quality system or you don’t. Is DoF a fallacy? I firmly state that we will continue to lose DoF as we continue to achieve higher capture resolutions. I would love to continue that debate over a bourbon someday. First one is on me. What is certain is that we should be knowledgeable in how we can gain or maintain DoF as resolution increases. There are many factors to consider as how to achieve the highest resolution and sharpness in our photographic process but those posts are for another day.