I understand the info re: 'correlation', but my understanding is that when one photon was stimulated, something occurred with the other photon, even though many miles apart.
If, as you say, the angle of measurement is changed for one photon, does that mean different angles of correlation will be present for the other photon?
Let me go into this a bit more for understanding.
First, photons can be polarized. We can detect this using polarized film (like in glasses). Photons polarized 'in line' with the film get through and those polarized perpendicular to the line of the film are blocked. After going through the film, ALL photons are polarized in line with the film.
Photons polarized at a different angle will have a *probability* of getting through that depends on the angle between the polarization of the photon and the line of the film.
Now, we can construct devices the produce pairs of entangled photons and direct them in opposite directions. Suppose we put polarized films far away from each others, but such that the photons go through them. The characteristic of entangled photons is that if you measure their polarizations in the same line, you will always get the same result. Either both go through or neither does.
Now, suppose we use this device to create many pairs of photons. At first, we set things up so the films at either end are in line with each other. What do we see?
Well, *both* sides see a random set of photons getting through.There is no way to predict which photons will and which will not get through and the randomness passes all standards of randomness we want to check.
But, when we bring those measurements together, we find that from each pair, either both went through or both were blocked. There is a perfect correlation here.
Next, we set the polarized films so that they are perpendicular to each other. What happens then?
Well, again both sides see a random set of photons going through. There is no way to predict which will go through and which will not. The randomness at both ends passes all standards of randomness we care to check.
But, when we bring the measurements together this time, we find that whenever a photon went through one side, the partner was blocked on the other. We have a perfect anti-correlation.
So, now we try to send a signal by adjusting our polarized film to send a sequence of 0's and 1's depending on our message using, say a film in line to represent 1 and perpendicular to be 0.
What is seen now? Well once again, *both* sides see a perfect set of random photons getting through. All standards of randomness are passed on both sides.
But when we bring the measurements together this time, we find perfect correlation when the films were in line and perfect anti-correlation when perpendicular. In that way, we can recover the message we sent (1 for same, 0 for different).
So, is there information sent between the two ends? NO! Both ends are random. There is no way to tell the difference *by looking only at one end* between aligned films and perpendicular films. No faster-than-light information is transfered because both ends are completely random. It is only when the measurements from both ends are brought together that we can get the message, but it is a slower than light thing to bring them together.
What Aspect's experiment did was essentially this, only also checked intermediate angles for the polarizations of the films. Again, at both ends, all measurements were random. All tests of randomness were passed. I tis only after the measurements were brought together that the *correlations* were seen.
I hope this helps some.