Quote Originally Posted by ntenny View Post
..surely everyone realizes (if they think about it) that analog processes are also full of stages in which information is lost or distorted, and that the feeling that a photo is somehow an accurate representation of "what was really there" or "what you would have seen" is an illusion that skips over a whole lot of mental modeling that we do unconsciously.
I guess I see a photograph as a naturally occurring phenomenon devoid, at the precise moment of rendering, of the Hand of Man.

At its most fundamental level it is a process that, although heavily refined by the chemical engineers to work well on a film substrate, and heavily altered by the optical engineers who designed the lenses to correct the things we don't like about the pinhole images that result from that hole created by the aperture blades, and conveniently encased into a useable form by the mechanical engineers who created the user interface we call a camera, essentially initiates, progresses, and concludes solely according to the laws of physics, chemistry, and thermodynamics.

The human photographer can chose the subject, camera, film, lens, shutter duration, aperture setting (pinhole geometry), along with an almost endless series of other variables. But all of these are chosen pre-rendering of the actual photographic image taking place.

At the moment the photographer releases the shutter, thus allowing light to strike the film medium, his contribution, and the contributions of all those generations of scientists and engineers, ceases to continue. At that moment the Hand of Man stops, and the Hand of Nature takes over.

Nature herself will spontaneously imprint an image onto the medium of whatever image was projected by the aperture pinhole. Nature does not want, and does not need, any input from Man for this process to proceed to completion. Once the shutter closes, the deed is done.

At that point Man steps back into the chain of events. He must tease that preserved image out and make it permanent. And he may later reproduce it many times over, into many different forms. But he cannot non-destructively alter that originally imprinted and preserved image. Nor can he ever identically recreate an exact duplicate, because the arrow of time has moved on. There are implications from the Second Law of Thermodynamics here, but we don't want to cause any more eyes to bleed than have already started to by this point.

The main difference between this process and the other technology is the presence of the Hand of Man at the actual point of rendition. Light still strikes a medium. But that medum no longer preserves the image. Instead, the Hand of Man, in the proxy form of computer software, endeavors to simulate the Hand of Nature by simulating the creation of a real three-dimensional negative.

It does this by reading voltages generated by (but not preserved by) the substitute medium and then logically abstracting them into the form of a zero-dimensional pattern. This pattern, consisting solely of abstract numbers that substitute for real-world negative densities, is itself a pure abstraction and is intended to be a substitute for a real negative.

The crucial difference in all of this is that numbers, being pure abstractions, can be altered non-destructively at any point in (or after) the original simulated rendering of the image. And the practical difference this implies is that at no point in the creation or subsequent viewing of a digital image can the viewer ever be 100% certain that none of the numbers have been altered from their original values.

If you suspect that the photograph in your hands was altered, you can immediately ask to see the original negative. There can be only one. And any after-the-fact alterations will be evident as destructive modifications. This chain of detectable events confers provenance.

If you suspect that the print from a digital image in your hands was altered, you cannot ask to see the original RAW file and be 100% certain that the file you are given is an accurate copy of the original values that existed at the original point of rendition. Any after-the-fact alterations cannot be detected because they were non-destructive, and so left no chain of detectable events. Thus, there is no provenance.

Changing a physical thing cannot happen non-destructively. Changing an abstract thing cannot happen destructively.

(Umm... anyone need eye bandages yet?)