I’ve been doing a lot of opining on my blogs of late, and much less science-ing that I would like. So I thought I’d try bringing a little science to the photo-a-day project, by playing around with f-numbers.
I put the camera on the tripod, with my fastest lens (a 50mm f/1.8 prime) and set up an array of SteelyKid’s Lego minifigs to be targets. Then I shot pictures of the scene at different aperture settings spanning the full range I could select. The two extremes are shown here:
I had to put it on manual focus, otherwise it went nuts trying to decide what to autofocus on, and as you can see, I didn’t get it quite dialed in on the frontmost minifigs. I kind of like the fact that it’s got that one bit of dog hair on the wagon zeroed in perfectly, though. The lens is a bit over 20cm from the table, because the tape measure I have here doesn’t have SI units on it, so I put it about 8 inches away (about the closest it would focus on). Each Lego set-up is 8 inches behind the next.
Because I’m a great big nerd, I took photos all the way through this sequence, and here’s a representative set of f-numbers (I left the exposure time on auto, which means some of the smaller apertures had rather long exposures, and a few of those pictures got some motion blur from me pushing the button; I left those out):
f/22:
f/16:
f/11:
f/8:
f/5.6:
f/4.5:
f/3.5:
f/2.8:
f/1.8:
Kind of impressive how rapidly the depth of field shrinks as you get toward the bottom. The factor-of-2 difference between f/22 and f/11 doesn’t make much difference, but going from f/3.5 to f/1.8, wow. which, you know, is not that surprising on an intellectual numbers-on-paper sort of level– it’s optical physics, after all– but it’s something else to see the effect on the images.
Anyway, there’s your highly nerdy photo-of-the-day. Next time I feel the need to do this, we’ll play with ISO values in the dark basement…
Uh, reminds me of a series of photos we had to do in school to understand depth of field…
I think we pretty much have the same camera, though yours is a newer version. You can change the auto-zoom options to center-pixel only. I forgot how to do it, but the internet will undoubtedly tell you. You can then pretty precisely aim at whatever you want in focus.
I have a degree in physics, and “understood” how depth of focus worked, but it wasn’t until I was in my 50s and needing bright lights to read small print that it hit me how this applied.
In a bright light, your eyes undilate, effectively raising the f-stop. That gives you a greater depth of focus which means that my old eyes could still bring the print within that depth of focus and I could still read it.
Of course the converse was true. In dim light, my eyes were no longer able to bend the lens sufficiently for the limited depth of focus at that low f-stop.
I think it is a plus that your focus was on the wagon rather than the front-most objects. What you lost in additional blurring of the background, you gain by seeing just how thin the depth of field actually is.
And apropos the comment from Ahcuah @2, I always tell my students that this is why people squint when they have trouble getting something in focus. I never thought about the effect of lighting on the iris, but that could be another reason some people have a lot of trouble driving at night. (Cataracts are the other.)
The issue with night driving, at least for me, is more that oncoming bright headlights look that much brighter when you are driving with dark-adapted eyes. (Much of my driving is on rural single carriageway roads; this may be less of a problem in urban areas, and is definitely less of a problem on motorways.) Driving also tends to involve distant rather than up-close focusing. However, I had similar symptoms to what Ahcuah reported, and that was one of the signs that it was time for me to get reading glasses.
Chad, Thank you for this fascinating and well-illustrated post.
As a photographer, I’ve had a question that’s been puzzling me for decades, but I think only physicists can provide me with the correct answer. A photographic image (be it digital or a print) doesn’t actually [as in physical reality] have a quantifiable depth of field (DoF) per se, despite the fact that there are equations and online DoF calculators that provide a result in metres or feet.
The DoF is the range of object space distance that we perceive as being in sharp focus when we are viewing the captured image. E.g. if we view side-by-side a 6×4 inch and a 12×8 inch print from the same image, the larger print will have a smaller DoF. In fact, the only way to make the two prints appear to have exactly the same DOF is by capturing two images: if the 6×4 print is captured at, say, f/4 then the 12×8 print must be captured using half of the aperture opening diameter, f/8. For a more extreme example, compare the wide DoF of a thumbnail image to the much narrower DoF of its full image displayed full-screen.
So, my main question is, what is the correct technical name for an attribute of an object (such as DoF) that depends entirely upon the conditions under which it is viewed (such as the size of a print and the distance from which it is viewed)? I’m tempted to say that it is a virtual attribute or virtual parameter, but the word “virtual” has become so overloaded with various branch-specific scientific meanings, and misused by laypersons and journalists, that I try to avoid using it.
I think equally important is my supplementary question: is a DoF (even when expressed in metres/feet) a qualitative or a quantitate attribute/parameter?
NB: This isn’t about pedantry, it’s all about trying my best to be scientifically accurate and consistent in my usage of terminology.