Thoughts on measuring the ISO of the human eye
Just read a little article on what the ISO of the human eye is: http://www.pixiq.com/article/eyes-vs-cameras (short answer: can't really say) and started toying with the idea of making a controlled experiment to determine the speed and ISO of the eye -- the aperture is said to be between f/2 (maybe f/3) and f/8.
Tests have shown that images shown for as little as 1/200th of a second can be seen and recognized (tested on fighter pilots). The problem is that to be able to compare what a camera captures with what the eye captures, we'd need some way to quantify how much detail we can actually see. The longer we look at something, the more detail we get out of it, much like frame stacking can be used to reduce noise. But we can't just show a picture to a person for a fraction of a second and ask how much detail there was. One approach would be to show a text and ask the viewer to read it, but that would include the time it takes to actually read the text. We'd need something where fine detail makes the difference between seeing different things. If we could find that, we could see what amount of noise renders the same detail unrecognizable.
I also hold that I've seen the spoke-wheel effect in daylight, but I will have to keep an eye out for seeing it again. It could be interesting to take, say, a bike wheel and rotate it at a certain rate. By covering it with black and white striped cardboard we can test if and at what speed there's a spoke-wheel effect. But I'll have to see it in real life first before I'll go to that bother.
But people with more time on their hands than me have looked into these matters before. I just like to think about them.
Tests have shown that images shown for as little as 1/200th of a second can be seen and recognized (tested on fighter pilots). The problem is that to be able to compare what a camera captures with what the eye captures, we'd need some way to quantify how much detail we can actually see. The longer we look at something, the more detail we get out of it, much like frame stacking can be used to reduce noise. But we can't just show a picture to a person for a fraction of a second and ask how much detail there was. One approach would be to show a text and ask the viewer to read it, but that would include the time it takes to actually read the text. We'd need something where fine detail makes the difference between seeing different things. If we could find that, we could see what amount of noise renders the same detail unrecognizable.
I also hold that I've seen the spoke-wheel effect in daylight, but I will have to keep an eye out for seeing it again. It could be interesting to take, say, a bike wheel and rotate it at a certain rate. By covering it with black and white striped cardboard we can test if and at what speed there's a spoke-wheel effect. But I'll have to see it in real life first before I'll go to that bother.
But people with more time on their hands than me have looked into these matters before. I just like to think about them.