Thom Hogan

As fast (and slow) lenses come out with aperture values that weren't common in the DSLR era, I'm finding a lot of misunderstanding about the aperture progression. 

Oh, the primary apertures aren't the problem: f/1, f/1.4, f/2, f/2.8, etc. These all are easy enough to calculate (multiply by 1.4 to get the next aperture in the sequence). Well, okay, that's not exactly true, as rounding starts to come into play (and the 1.4 value itself is a rounded number); still it's close enough for quick understanding of the one-stop differences. 

The biggest problem I'm seeing with photographers comes up in comparing f/1.2 versus f/1.4, or f/1.4 versus f/1.8. These don't at first seem to be big differences, but each step is a half or two-thirds of a stop, which can make a significant difference in exposure in lower light conditions and has a fairly clear depth of field difference at the 35mm+ focal lengths. 

When you look at Nikon's choices for Z-mount primes so far, you'll see that they've almost always chosen wide, clear differences: f/1.2, f/1.8, and f/2.8 (the exception is the 40mm f/2). That's an exposure change of ~1.2 stops from f/1.2 to f/1.8, and 1.3 stops from f/1.8 to f/2.8. Those detail-oriented Nikon engineers are nothing if not (mostly) careful and consistent in their design tactics. 

Sigma now has 35mm and 50mm lenses at f/1.2, f/1.4 (half stop change), and f/2 (full stop change). Sony is a bit less consistent with 50mm lenses at f/1.2, f/1.4 (half stop change), f/1.8 (two-third stop change), f/2.5 (full stop change), and f/2.8 (third stop change). 

So what is the full aperture progression? Here you go:

Technically, before the CIPA-agreed rounding, the full aperture sequence goes 1, 1.41, 1.99, 2.78, 3.92, 5.53, 7.80, 11, 15.51, 21.87. The above charts use the agreed-upon rounding numbers. The marks on a Japanese-produced lens will conform to the CIPA rounding with the lens focused at infinity. In other words, there's a lot of wiggle room in the actual numbers, which is partly the reason why trying to stick more lenses into the equation at minimal aperture differential is a fool's errand.

At the slow end of the aperture range (e.g. with telephoto focal lengths), we're now seeing lenses that weren't really possible (with autofocus) in the DSLR era, where there was a fairly strong cut-out of phase detect focus performance beyond f/5.6 due to the geometries involved. It's now more common to see f/6.3, f/6.7, f/7, and f/8 as maximum apertures in mirrorless at the telephoto focal lengths. Some people panic over f/6.3 versus f/5.6, but that's only a third of a stop, not something dramatic. Even f/8 is only a stop slower than we used to see in the DSLRs. 

Of course light is light, and light is a key ingredient in exposure. 

Remember, the equation for exposure is really: 

     EXPOSURE = LIGHT filtered by APERTURE filtered by SHUTTER SPEED

Thus, the implication of a faster or slower aperture in the same light—less or more filtering—means that your shutter speed changes to create the same exposure at the image sensor.

At the fast end of the maximum aperture spectrum (f/1 to f/2.8), the implication is that you can hold a 1/60 second shutter speed into dimmer and dimmer light with faster apertures. 1/60 is a good marker because slower than that is difficult to handhold well and anything slower is absolutely subject to subject motion. 

At the slow end of the maximum aperture spectrum (f/5.6 to f/11), the implication is that your shutter speed is going to get clipped and/or your ISO boosted, potentially even in decent light. The Sunny 16 formula is 1/ISO shutter speed at f/16. So to hold a 1/1000 shutter speed in sunlight on a base ISO 64 camera you need f/4 or faster as the aperture, otherwise you need to start bumping your ISO value. A rule of thumb is that you'll lose about a stop of dynamic range with each stop of ISO bump, though, so you want to avoid that, if possible. 

So at both ends of the aperture range there are intersecting issues you need to consider. The real question is how often do you encounter those, and what would you do about them if you did? Generally speaking, the thing that most people end up doing is bumping up their ISO when they don't have a "fast enough" lens. So the real question is this: how often are you bumping ISO up because you don't have a fast enough lens? And are you sure that you can't do the other obvious thing, and add light? 

That last bit is what we had to do in the film world, as our ISO constraints were much more rigid and filled with peril. But given the state of noise reduction software these days, I feel ISO bump is a reasonable choice now, particularly since a faster lens probably is only going to give you a stop or so advantage from what you'd normally use. 

But getting back to the misunderstandings that prompted this article, I see, for instance, quite a few people saying Nikon should create a line of f/1.4 primes in addition to their f/1.2 and f/1.8 ones. Personally, I don't see how chopping the choices so finely—f/1.2, f/1.4, f/1.8—does anything other than increase the cost of your gear closet. The half to two-thirds stop difference that f/1.4 option would make isn't enough for me to try to add another set of lenses: I can deal with a half to two-thirds stop ISO change.