Does a 50mm Equivalent Lens Really See the Same as a Human?
Everyone agrees that a wide angle lens sees a lot and that a telephoto sees a narrower field of view. But what makes a normal lens "normal?"
There are a number of ways to approach this question. The simplest one of course is that this is what most every beginning photography book teaches. On a "full frame" 35mm camera, the 50mm lens is considered normal by definition. That translates into roughly 46 degrees angle of view. I guess in a sense there has to be something in the middle of the scale between wide angle and telephoto and 50mm is a nice round number, but on balance isn't a very satisfactory answer for me at least. For one thing, in a world where the standardization on the 35mm film format (or equivalently sized sensors of course) just isn't what it used to, the magical number 50 isn't really 50 anymore. This same angle of view translates to around 35mm (technically 33 1/3) on Nikon DX and similarly sized APS-C sized sensors. Basic math gives us a variety of numbers on other formats.
Another, more technical answer is that a lens with a focal length equal to the diagonal size of the digital sensor (or film) is by definition "normal." With film measuring 24mm by 36mm, the diagonal works out to be about 43mm, a value sometimes quoted by people claiming to be in the know.
But regardless of the final number, these arguments completely ignore the mechanics and optics of human vision, an omission that seems relevant to the discussion at hand.
Some of the same beginning photography books reference earlier do use the argument that a normal lens is "normal" because is sees roughly the same angle of view as the human eye does. Scientists tell us that the human eye actually has a focal length of around 22mm with an angle of view of at least 120 degrees. This seems so far off the expected range that there's clearly more going on than just basic optics.
The most notable difference no doubt is that while your camera sees with just one lens, a full complement of human eyes generally totals two. That should tend to widen the collective angle of view even further though. But the field of view of those two eyes overlap so we can sense distance based on parallax. We really only see clearly in the central portion of that vast field of view. When people talk about seeing things out of the corner of their eye they are alluding to the fact that our visual acuity falls off towards the periphery of what we "see." If one pays attention only to the central portion, the angle of view we see clearly does fall somewhere in the 40 to 60 degree range.
There are other factors that complicate this whole thing though. For one, the back of our eyes are round. The flat surface of digital sensors and film allows the projected image to widen further than how we would see the same thing with our eyes.
Then there's the whole thing about our brains. So much of what we think we see is in fact a reconstruction of what our eyes actually see together with what they saw quite recently. We operate on a model of the real world more than we do on the real world itself. To illustrate, imagine that you're driving down the freeway. You're intently looking at the road in front of you as you should be. But for just a moment you glance down at the speedometer, your cellphone, or what have you. In your brain, the freeway continues to whiz past you uninterrupted even when you briefly look elsewhere. If you are so unfortunate as to have an accident when not paying attention, it would no doubt come as a surprise. In your mental model of the world around you, that accident wasn't supposed to happen. You thought the traffic was still behaving as before, but it wasn't. Surprise!
To complicate things even further, in your brain, you don't really see any specific angle of view. What you think you see is comprised of what is of interest. You pretty much disregard what isn't. You might be aware of some things far from your center of vision while being pretty much unaware of other objects much closer. After factoring in the processing performed by your brain, your actual field of view is irregularly shaped, not round as your lens projects or rectangular as your camera sensor captures.
Portrait photographers will often recommend lenses as long as 85mm equivalent since they result in happier customers than when they shoot closer to the literally correct 50mm range. I remember once shooting a picture of my parents and me using a camera with a self-timer. We were all sitting on my couch with the camera on the other side of the small living room in my apartment at the time. To get us all to fit in the picture, I zoomed the lens out to around 35mm. Big mistake. If you've ever done likewise, you will have noticed that the close camera position accentuates perspective in a less than flattering way. Everyone's knees being the closest thing to the lens appeared huge in the image. But aging bellies that protruded only nominally closer to the lens than the rest of the bodies they were attached to also looked bigger than expected. Needless to say, I wasn't too proud of that image. I goofed by using a wider lens than the more "normal" 50 to 55mm focal length. People who want more flattering portraits do the opposite by using lenses longer than "normal." The concept of "normal" is all subjective I guess.
Trying to compare camera vision to human vision can never be an exact side by side contest. Trying to pick the exact focal length that is equivalent to human vision is like trying to determine which variety of apple is most like a banana. Both are fruit, but they're not really the same thing. The correct answer falls somewhere around 50mm, but there is no exact answer.