Do photons propagate in a single direction? I’ve been researching this and can’t find a straight answer.
I’ve heard that the direction of a photon, when it’s emitted, is random. But I’ve also heard that if it’s emitted from point a to b, it takes all paths simultaneously. Other people say that it doesn’t take all paths, but that it’s only a probability. But with a polarizer, doesn’t a photon shift in order to pass(or not shift to not pass)?
I guess to clarify: if a photon is emitted from a light bulb and reaches my eye, will it go straight, or could it kind of swerve(or does it take all possible paths at once)? I keep getting conflicting answers.
Please elaborate on this for me. I’m a noob to this stuff. Plain English would be appreciated! Thanks!
 A: Photons, to the extent that it makes sense to model them as particles, follow the trajectories predicted by classical ray optics. On the other hand, for phenomena like diffraction, you must model electromagnetic radiation as waves regardless of whether your model is classical or quantum.
The classical version of the "all paths" idea is the Huygens–Fresnel principle. Quantum electrodynamics elaborates on this when interactions with other quantum fields are involved.
The difference between the classical and quantum views is that the intensity at the detector in the classical model is the energy that arrives at the detector, while in a quantum model it is the probability that a photon arrives.
A: The photon takes all possible paths from the light source to your eye. Coherently. That last little "coherently" part is a big deal: it means the photon goes in a straight line from the light source to your eye because the straight line is a stationary point in the phase from start to finish. Now if there are mirrors (or any surface, really) or refractive media in between, the photon can bounce or bend, but it is taking the path of least time (Fermat's Principle).
The question is: how straight is straight? Really straight. You can borrow some analysis from the communication industry, the Fresnel Zone (https://en.wikipedia.org/wiki/Fresnel_zone), in which a radio wave propagates from Tx to Rx in a straight line, but with some lateral influence being relevant. Thought radio is classical, the phase and propagation considerations don't differ from quantum mechanics. (If you want to know how a classically coherent wave is made from photons, see "The Glauber State").
A: A light source can emit single photons or radiate many photons, a light source can emit in a confined direction (laser, flashlight) or many directions (bulb).  Photons are best considered as being waves in the EM field and the EM field can handle many many photons due to superposition of waves .... but energy is never lost in the medium (EM field).
Only electron activity puts waves (photons) in the EM field as quanta (one at a time), only electrons (in atoms) eventually absorb photons (as quanta) .... there are likely many photons in space that are travelling energy in the EM field that will never get absorbed.
Nobody knows what a photon actually does in space, we can only absorb them to observe. We can also generate, for example we can bounce a laser off the moon.  Based on c and the time taken we determine they travel in a straight line.
An excited electron in an atom (even before real photon emission) is already interacting with the EM field (virtually, forces only) ... thus we can say it is considering all directions .... but it does not mean the real the photon is traveling all paths .... the EM field with the electron considers all paths. The real photon direction likely results from the EM field/electron interaction. In a laser cavity we heavily confine the EM field ... which gives the eventual photons a confined direction.
Maxwell proved that the E field and the M field were tied together and proposed that light was a unified concept of E and M, he also derived a propagation equation for light which would support straight line travel of a confined wave packet .... this accomplishment is as astounding as Einstein's E=mc^2.
So swerving is not likely ..... but since we can't observe the EM field directly we will never know .... we can only observe energy coming out of the EM field as quanta/photons.
