timeUnit said:
(I'm not a chemist/physicist, don't shoot me)
Isn't it so that on a molecular level, each silver halide molecule has either reacted to the light or it hasn't, it's either "black" or "white". When I look at the grain in my grain focuser I don't see "grey" grains, but grains and the space between. So in a sense, film is discrete, each grain is "on" or "off".
The answer is: who knows? As far as I understand (there may have been many recent findings of which I'm not aware, of course) the theories regarding physical/chemical processes which occur in sensitive materials are mostly based on speculation and are constant matter of debate since 150 years, far from being clearly established.
Leaving silver halide theories aside, however, the more you get in smaller dimensions, the more you loose contact with reality as we perceive it. Your sentence should rather read "there is some probability that this unknown energy level may have been altered by one of various photons". The fact itself that the photon is an entity that travels at the speed of light but has NO MASS and may be (statistically speaking) ubiquitous as God should warn you that it's not as easy as figuring pool balls and planet's orbits. That, of course, if you blindly accept the so-called "Standard Theory" of the Copenhagen's School, which interpretation is in turn matter of debate since 100 years.
The more you get at atomic level, the more you get into indetermination. Quantic state reading - by the way - doesn't exist, or better may be considered a memory-destructive operation. When you check for quantic states, you instead IMPOSE a quantic state over a particle which had previously an undetermined quanitc state which you will never know. At a maximum, you can perceive that you altered its quantic state with reading. So you understand that you can't treat this matter as if saying "well I have a silver halide molecule at such and such energy level, here comes ONE photon of such and such frequency and one black piece of the grain comes out".
If it was so easy, engineers would have covered the sensors with a layer of silver halide, and then some magic machine would read all quantic states of every single particle in the layer, then storing the results in some imaginary medium. Well, THAT would mean beat resolution and dynamic of film photography! Too bad that's just dreaming, for now and most probably for ever...
Of course you understand that what you see as "grain" are not molecules, but enormous aggregates of an incredible number of molecules, which in turn tend to aggregate in bigger bunches of all shapes and sizes. You can easily see how grain has a total different shape and size in areas of uniform tone compared to more dynamic areas. If you take a uniform grey sky, you'd see grains as big as peas, while in areas with a lot of details and varying tonal values grain is virtually invisible, even in areas which have the same tonal value of the before mentioned uniform grey sky. So, if you ask me, it's not again simply a matter of "molecule on" or "molecule off", and not even "grain on" or "grain off", but perhaps how "structured" the compounds forming the developed emulsion are. It recalls much more a fractal structure than a lattice or a CCD sensor.
And in the end, the basis of all differential calculus is that an infinite amount of infinitesimal units makes a finite quantity. This is a fundamental concept of all sciences, and was already fully comprehended by Galileo Galilei. When you treat the number of atoms or molecules in a sensitive layer, that number is not mathematically and physically infinite, but is so big that can be treated as such for every practical purpose. I, personally, have no problem to think that traditional photography happens in a continuum realm. And I defy every megapixel maniac to treat traditional photography atom-by-atom, as if it was a bit-to-bit operation