In 1805 or thereabouts, Thomas Yound performed the famous double slit experiment:

The purpose of the first slit is to make the light "spatially coherent," meaning it comes from a point-source, of line-source in this case. At the slit, the light diffracts so that is spreads out and covers both of the next two slits. The waves from each of the two slits on the rightr interfere with each other, creating the light and dark bands seen on a screen to the far right.

The waves, in this case, are not electromagnetic waves but rather are quantum waves of the square-root of the probability of finding a photon anywhere on the wave. The double-slit experiment can be done with electrons, protons, atoms, or buckyballs (which have a very small wavelength).

What this all means is that a photon (or electron, proton, buckyball) is initially in a state of superposition with respect to its position. when "measured," the photon stops being in many places and instead is in one place where it gets measured.,

Where is gets interesting is when two photons (or other quanta) are created in such a way that there is a relationship between them. When a violet photon is absorbed by a beta-barium-borate (BBO) crystal and two infrared photons are emitted, their momentum must add up to the momentum of the original violet photon. Also, their polarizations can be related - either the same or orthogonal. These properties are in a state of superposition until they interact with stuff, whereby one photon takes on a value and the other photon takes on the corrosponding value.

If the two photons didn't take on correlated values, it would be possible to violate conservation of momentum. When measured, each photon takes on a value instantaneously regadless of separation distance and light speed, and since the universe is relativistic, the idea of "instantaneous" can get very interesting.

The obviousl question to ask is whether the photons really do have values for their properties and that we just can't predict them. The answer to that question is definitely "NO!" John Bell came up with a theorem that showed the correlation rates could only reach a certain percentage if they did actually have values, and quantum mechanics predicted a higher rate. See Bell's Inequality. Many experiments show that the rate of correlation is what QM predicts, ruling out these "hidden variables" theories.