Someone asked this question on Ask Metafilter: why isn’t there a camera that takes a picture exactly as my eye sees it? I’ve actually wondered about this before, and was never quite satisfied with the answers that I got, so I was really happy to see this question.
Essentially, the answer is:
- The camera only has one lens, but you have two eyes, and your eyes have specs that will beat any camera on the market by a wide margin.
- Your brain does a lot of interpolation and post-processing.
- You can focus much faster than a camera can.
- Different parts of your eye can have different levels of sensitivity to light, but a camera can only have one level for the entire sensor. (That’s why you can look at a scene with both light and dark areas and still distinguish details in each area, but a camera will produce pictures with blown out areas or completely black areas.)
In one answer:
In other words, your vision can’t be replicated by a single camera because it’s essentially the result of two cameras (that are better than even the best cameras available now) taped next to each other and hooked up to a supercomputer doing real-time image interpolation and processing.
I wonder if we have the technology now (or will soon) to do that. I don’t think the technology is up to snuff on the speed side yet, but couldn’t you theoretically hook up two cameras to a computer, and do the image processing? Two cameras would give you a wider dynamic range and two focal points. I imagine the image processing part would be the hardest though, particularly combining the two images properly. I remember reading some papers where people used two side-by-side cameras to build depth maps and other interesting features, so maybe it’s in the works.
Still, even if all that was possible, the brain probably does too much stuff that a computer can’t replicate (at least not currently):
Your eyes and brain have evolved over the years to make high probability assumptions. Interestingly, when presented with a field of Gaussian noise — where every pixel was a randomly selected shade of gray — the eyes and brain do no better than a camera. Not so for any natural scene — there your highly evolved and specialized brains do far better than any camera so far produced.
I couldn’t find the study mentioned after a quick search; it would have been interesting to see how performance was measured. I can see how it could be true though. The brain does fill in a good deal of the scenery, especially in a familiar environment.
Anyway, it’s an intriguing question. I’ll have to take the Computational Photography class when it’s offered again.
Currently listening to: Why by Shattered Atom
In response to 3: so just how much faster is the eye compared to the fastest consumer focusing mechanism out there?
In response to 4: as technology progresses, we can start interspersing sensors of various sensitivities in there to deal with the dynamic range, can we not? That would be cool.
What the hey? You changed the theme again and I was confused. I like this picture too
How much faster: I have no idea.
Interspersing: Yes, the thread that I linked mentions that already.
Yup, I changed the theme because the other one didn’t have any delineation between the white area with content and the background. I just picked another theme that lets me set the picture, and this picture seemed to go well with the background color, in terms of color scheme.
I totally agree with the theme change. I wanted to tell you that there didn’t seem to be enough vertical lines in the other theme, and I also wanted to say that the overall color of the picture you chose for this theme goes with the color scheme very well. Good job!
CS448 did something like this for a few years:
http://graphics.stanford.edu/projects/array/
Whoa, that’s kinda cool. I wonder how they compose all the results into one image, if they do that at all…