When I started getting into photography, I was amazed at how my eyes could see just fine, but when I tried to take a picture, I’d discover that the camera couldn’t. Even when cranking up the ISO and opening the aperture, the shutter speeds weren’t hand-holdable. Just how light-sensitive is the eye, anyway?
Of course, the human eye is not a camera, but it sure has a lot of commonalities. There is a lens, and there’s an aperture through which light falls. There’s a sensor in the form of a bunch of rod and cone cells. The f-number of a camera lens can be calculated by dividing focal length by the diameter of the aperture, i.e., the hole through which light falls; the same can be done for the eye. Various sources estimate that at its widest, the eye is hovering slightly above f/2. A photographic exposure is determined by the aperture, the exposure time (usually measured in seconds) and the recording medium’s light sensitivity (usually measured on the ISO scale). So how does one estimate the eye’s “shutter speed” and “ISO value”?
The eye doesn’t have a shutter, of course. But there must be some interval of time from which the brain integrates visual data to form an image. If we saw every photon that hit the eye individually, our vision would be a chaotic pattern of tiny, not very bright dots that jump around our visual field, and we wouldn’t be able to distinguish between different luminosities. On the other hand, if the brain kept adding up every photon that hit every rod or cone forever, our entire visual field would start as pure black when we’re born, and quickly turn pure white. If you take a picture of a night scene with an exposure time of minutes or hours, it looks like day. But if you stay in a dim room for minutes or hours, the room doesn’t get gradually brighter until it’s bright as day. We do see better, but that’s not because our eyes have had more time to gather light; instead, it’s because the eye adapts to the darkness by upping the sensitivity of the rods and cones (equivalent to upping the ISO on a camera).
The best approximation I could find was this, a big experiment on contrast thresholds in the human eye done during WWII and published in 1946. The procedure called for people to observe a white dot of various luminosities from some distance. The study notes that “in general, 15 seconds of exposure time were adequate for minimum thresholds.” Extrapolating slightly from this, we can say, then, that in near-total darkness, the human eye’s equivalent to “shutter speed” is up to 15 seconds. Of course, there must be some marvelous signal processing going on to ensure that nothing looks sluggish. Probably recent photons are given priority; after all, we don’t observe things fifteen seconds after the fact even when it’s dark. So if we want to calculate the brain’s “exposure” to compare it to a camera, 15 whole seconds is probably not accurate.
Even if we fix the human visual system at f/2.0 and 15 seconds exposure, which is already too approximate to be of much interest, that still doesn’t tell us the eye’s “ISO”. After all, a camera exposes to average the picture out at 18 percent gray. The eye, however, probably doesn’t; when you look at a well-exposed photograph of a dark night scene, it looks lighter than it does to the naked eye. The brain, however, perceives lots of detail even if it’s mostly in the dark range, whereas photographs become fuzzy and difficult to make sense of if all the info is within a stop or two of absolute darkness. In addition, the eye doesn’t stay fixed during the entire “exposure”: it moves around all the time, whether we want it to or not.
All this speculation, and we aren’t much closer to finding out how the eye’s sensitivity to light compares to a modern camera’s. What should be clear, though, is that the human visual system is a wonder of nature.
dailymeh posted this on April 6, 2011
