Couldn’t this mean that inserting a middle filter at 45 degrees means you are re-orienting the photons closer to the orientation of the filtering mechanism of B, hence now some of them are able to vary from their vector in such a way as to allow some of them to now pass through filter C whereas before filter B was inserted this would have been impossible, since the variance in the vector could only approach 90 degrees but never actually hit it?
-------> (A) --------> b[/b] --------> (B)
For example, the photon is a “wave-like particle”. It may be particle-like but it represents a finite quantity of energy bounded up in a wave traveling as a vector; this wave is oscillating in two directions or along two dimensions, an x and y axis. Both x and y are perpendicular to one another in orientation, as how waves naturally propagate. Now we also know that the energy inside the wave is not localized or “collapsed” into a point-like particle unless it is forced to do that by some “observation” taken of it, namely by the wave impacting something else that forces the energy to collapse from its wave-form into a more particle-like (non-waving) form.
Filter A is allowing only photons to pass through which happen to be oriented in the same direction as the filtering mechanism. However, if there is also a probability that for a photon which isn’t exactly oriented this way but it will still pass through the filter anyway, because the manner of its wave-function collapsing actually ends up re-orienting the x-y axes in such a way as now lines up with the filtering mechanism, then we can see that some light can get through the filter even though it wasn’t exactly oriented in line with A to begin with, but was “close enough” given the probabilities involved having to do with how its wave was collapsing. The probability of this happening would be greater the closer the orientation of the wave-function is to the orientation of the filtering mechanism at the time of impact. So, some photons are getting through even though they were not originally oriented exactly aligned with the filtering mechanism.
This does mean that the light having passed through the filter is now homogenized in such a way that every photon that has passed through is now aligned in the same way as all other photons that passed through. Being now “unobserved” they become more wave-like or remain wave-like. This also assumes the photons keep their orientation between filter A and B (they don’t spin around to find new orientations, which would defeat the function of the filter entirely hence doesn’t seem to be the case (but see * below)).
The same situation occurs with filter B as it did with filter A, only this time all the photons going to filter B are pre-aligned, which wasn’t the case with filter A. Now filter B is a perfect expression of the probability of a photon not perfectly aligned with the filtering mechanism in B having its wave-function collapse in such a way as to happen to reorient it and allow it to pass through B anyway. Basically, if its orientation is already very close to that of the filtering mechanism, so tiny perturbations in precisely how that wave is changing/collapses allows it to assume the orientation needed to pass through the filter anyway. This becomes less probable the greater the difference in orientation is between a given photon arriving at B and the filtering mechanism of B, culminating in an impossibility of being able to pass through B if B is exactly perpendicular to the orientation of the incoming photon (because the wave cannot probabilistically collapse in such a way as to shift its orientation completely from x onto y, because x and y represent polarities or exact opposite dimensions within the structure of the wave). And it becomes less likely that x will shift toward y the closer this would end up coming to y.
This explains why adding filter C in the middle would allow light to move through B whereas without C being present then no light would move through C. What C does is reorient the orientation of the light closer to the filtering mechanism of B, namely to within 45 degrees instead of within 90 degrees. And this isn’t a perfect 75% increase because of the curvature of the probability index being circular in nature (it has to do with how the photon’s wave orientation is rotating: instead of a point going down in both directions like a triangle, you have a point going down in both directions in a gradual curve that becomes steeper the further out it goes, like a bell curve). This explains why there is more area under the curve as the orientation of the photon is rotating compared to the area under the curve might be if we were simply assumed a numerical mid-point between 0 degrees and 45 degrees.
*Note that an alternate explanation exists for if we don’t want to imagine the process of encounter or wave-collapse as being something subject to a possibility of that encounter/collapse happening to change the orientation of the wave in such a way as happens to allow it to pass through the filter anyway. You can instead think about the wave as already having a natural variation/change in its orientation as it passes through space on its own (before it arrives at the next filter). This change in its orientation could be probabilistic in the sense of being essentially chaotic or idiosyncratic and unpredictable, subject to tiny influences impossible for us to map and know but which can cause the wave to change slightly its orientation as it moves through space on its own. This change would, however, theoretically be unable to cause such a significant change in orientation as to allow x to completely map onto where y is, so the probability that the orientation will change a lot is less than that the orientation will change a little. Or, the variation or change in the orientation of the wave as it is traveling between filters is more like a probabilistic effect that only materializes once the wave is forced to collapse and become more particle-like.
One thing to note, if the filter C is instead added before A or after B rather than inserted in the middle between A and B, and the same effect is observed as when C is put between A and B, then I would need to re-think all of this. But the video made it seem like the 45 degree tilted filter C needed to be added between A and B to produce this effect.