Last week Steve pointed out that a lens-based optical system like a camera or the eye can only focus at one field of depth at a time — meaning, as some sample images he presented illustrated, that other parts of the scene will be out of focus. The images Steve presented were of a landscape. In one image, some plants in the foreground were in focus, whereas a distant mountain was a blur. In the second image, the mountain was in focus, the foreground a blur. [Here I have combined them in one image.] Why, Steve asked for the original images, did painters — at least before the invention of photography — not paint the blur like this? More generally, why don’t artists paint what they see?

The images Steve presented were provocative, but they were not entirely consistent with the way the eye filters and processes information. The major difference is that the photosensors in the digital camera form a regular matrix of identical elements, which is why a photograph of a uniform texture would appear as a uniform texture. In the eye, the photosensors are not uniformly distributed — they are much more densely packed in the center of vision, the fovea. Furthermore, they are not of the same nature across the retina. The sensors near the fovea are sensitive to color, whereas those in the periphery are less so.


Thus, if painters were to paint what they actually saw in one glimpse of a scene, the image would not only be blurred for depth, as Steve’s photos, but it would be also blurred radially around the center of fixation. Furthermore, the image would be colored only towards the center; around the edges it would be monochrome, although not necessarily black and white. [Here I show a quickly-made demonstration based on one of Steve’s images. For the size of the image on the monitor, the peripheral blur is exaggerated.] In short, a “physiological photo,” if such a thing could be made, would be even less like an artist’s painting than Steve’s sample photos. Which of course only emphasizes the question, why don’t painters paint what they see?

The photograph captures a moment in time, but the “physiological photo” would not do so. The reason is that the information that the eye records fades quite rapidly — if that input does not change. In the laboratory it is possible to present an image to a person in such a way that the light falling on the retina remains constant. The startling result is that after some seconds, the image will fade, the person will see nothing. In normal life we never experience this fading simply because it is impossible to stabilize an image in this way — the eye moves, the body moves, or the scene itself moves.

Thus, if an artist were interested in painting from the “physiological photo”, that is, what the eye really sees in a moment analogous to the moment it takes for a camera to record an image, he or she would need to paint very rapidly, or he or she would have nothing more to present than a blank canvas.

In normal life we are partially aware of the natural blur in peripheral vision, we don’t perceive our lack of color vision in the periphery unless we make an effort to do so, and the fast-fading of vision doesn’t come into play because what we look at changes so quickly. What we see is much more than the information the eye captures in a moment — by analogy, much more than the digital camera can capture in a moment — because our vision integrates over time, over space, over wavelength, across movements of the eye. Not to mention that we have two eyes, each receiving a somewhat different image, the two of which are seamless integrated in our vision. All of this integration — particularly that over time — means that comparing normal vision and the type of image recorded by a photograph is comparing things of quite different natures.

Why don’t artists paint blur? We could as easily ask, why doesn’t the digital camera focus consecutively at different depths and digitally process the information to present a uniformly focused image? Artists don’t paint the blur because their visual systems are so adept at integrating information that the blur (and all sorts of other distortions) are not noticeable — in short, because the visual system is miraculously advanced technology. The digital camera doesn’t produce a multi-focused image (despite the “auto-focus” feature) because, despite fine optics and a huge memory, it is a relatively crude device. This, of course, is why photography is an art form — in the end it is still the photographer doing the real work.

In answer to the question, why don’t artists paint what they see?, the answer I think is that they do paint what they see, or try to, but what they see is an integration over a much longer “expose” than a photographer is used to working with.