Is a polarizer a filter or a converter?

In summary: It seems you could avoid even having to ask your question if you perhaps decided to abandon classical thinking right off the bat, i.e. what does it mean for an individual photon to be polarized at 45 degrees? You could take the math for what it's worth in experiment (a lot) and say that polarization of individual photons is defined as|\psi\rangle=\sin\alpha|\psi\rangle_{||} + \cos\alpha|\psi\rangle_{\perp}
  • #36
mr. vodka said:
Okay as I seem to be ignored trying to explain it nicely, I'll be blunt: you're wrong. Say system S is a construction of two polarizers at different angle, then S is a filter because whatever comes out was already part of the original wave, there's just something left out.
No, it wasn't. This can be seen easily if you use a third polarizer at 90° to the first to analyze your wave. The first polarizer filters out all parts of the wave which could pass the third. So if the second would just filter out more parts of the wave, there would still be no transmission through the third polarizer. Actually, there is a transmission, so the second polarizer has created new parts of the wave.

mr. vodka said:
How does a polarizer actually work on a QM-level? I can't figure out what it does to one photon. What is the measuring aparatus?
I'm not sure if this is a satisfying answer to your question, but like every measurement device, its a big system which can be treated classically. Similar to a Stern-Gerlach-Apparatu for measurements of the spin of the electron.
 
Physics news on Phys.org
  • #37
jfy4 said:
But consider it as a "converter"; what is being converted!?
In the CI, the answer is "the state of the photon". It enters with initial state |i> and exits with an eigenstate of the measuring device |f> (as DrChinese has pointed out).

However, I just realized that the answer is interpretation-dependent. In the statistical interpretation, I think the situation is analogous to the classical case: the first polarisator is a simple filter, the second a converter (with limited efficiency).
 
  • #38
kith said:
No, it wasn't. This can be seen easily if you use a third polarizer at 90° to the first to analyze your wave. The first polarizer filters out all parts of the wave which could pass the third. So if the second would just filter out more parts of the wave, there would still be no transmission through the third polarizer. Actually, there is a transmission, so the second polarizer has created new parts of the wave.

Well, I still stand by my case that it's a filter, but I also understand your argument; I think are points of view aren't mutually exclusive, and this because the words "being a part of the original wave" actually seem to be pretty stupid:

Take for example a sine wave (name: X) at 45° degrees, that you can view as a sum of a sine at 0° (name: A) and at 90° (name: B). Now a polarizer at -45° won't let anything through (making a drawing helps). But if you first put a polarizer at 0° (only keeping A), and then the one at -45°, you will (not surprisingly) get something. The wave you get at the end is actually a component of A, which on its turn was a part of the original wave. But of course you might sensefully argue that that component is not a true component, because it is canceled out by a certain component of B. So in that sense, the wave you get in the end is not a part of the original wave. But as I see it, each polarizer takes a component of the wave, and to me a component of a component can still be regarded as a part of the original wave. My main point is this: it seems to be a matter of taste, wouldn't you agree?
 
  • #39
mr. vodka said:
My main point is this: it seems to be a matter of taste, wouldn't you agree?
I agree that it is mainly a semantic question. I also understand your argument, which seems similar to the last paragraph of jfy4's last post: with every polarizer, you "filter out" some intensity.

Still, I don't think that calling them filters is appropriate. If I have a quantity of something and I filter out all parts with certain properties, I can't get parts with the same properties back by filtering out all parts with certain other properties.

Maybe we should just agree to disagree. As far as I see it, everything relevant has been said. ;-)
 
  • #40
If I have a quantity of something and I filter out all parts with certain properties, I can't get parts with the same properties back by filtering out part with other properties.
Can you maybe rephrase, I don't understand the construction of your sentence (English is not my mother tongue)
 
  • #41
mr. vodka said:
Can you maybe rephrase, I don't understand the construction of your sentence (English is not my mother tongue)
It isn't mine either, so maybe my sentence is more complicated than needed. ;-) I try it again with a concrete example.

We have a bowl filled with objects in different colors and different shapes. If we filter out all red objects, there is no way to get any red objects back by filtering out other objects (cubes for example). This is the behaviour I expect from filters, but it's not the way polarizers behave. If I filter out one part of the whole wave (for example the 0°-polarized), I can get it (partially) back by filtering out another part (the 45°-polarized).
 
  • #42
Well, as you said: "partially" :p that's the key word for me there

Okay I agree to disagree ;)
 
  • #43
kith said:
It isn't mine either, so maybe my sentence is more complicated than needed. ;-) I try it again with a concrete example.

We have a bowl filled with objects in different colors and different shapes. If we filter out all red objects, there is no way to get any red objects back by filtering out other objects (cubes for example). This is the behaviour I expect from filters, but it's not the way polarizers behave. If I filter out one part of the whole wave (for example the 0°-polarized), I can get it (partially) back by filtering out another part (the 45°-polarized).

Some polarizers behave like waveguides as well depending on the polarization i/o relationships. The one posted on the video is an example of this. Internal structure of these polarizers is multiple layers of waveguides (at micro or nano level - depending on the frequency) rotating respectively from original polarization state to aimed (output) polarization state made from very low refractive index material. Linear to circular polarizer is an example of this (waveguide). Shifting from circular polarization to linear can be done by filtering and shifting from linear to circular polarization by waveguiding. There are also various polarizer types like IR polarizers which operates independent of the polarization state. 'Polarizer' is not enough to decide whether it's a waveguide or a filter, it depends on what kind of polarizer it is.
 
  • #44
kith said:
I agree that it is mainly a semantic question. I also understand your argument, which seems similar to the last paragraph of jfy4's last post: with every polarizer, you "filter out" some intensity.

Still, I don't think that calling them filters is appropriate. If I have a quantity of something and I filter out all parts with certain properties, I can't get parts with the same properties back by filtering out all parts with certain other properties.

Maybe we should just agree to disagree. As far as I see it, everything relevant has been said. ;-)

Why does that not make it a filter?
 
  • #45
Drakkith said:
Why does that not make it a filter?

Again: a pure filter only removes, it doesn't modify.
Perhaps a counter example is useful: An attenuator is a pure filter, as it doesn't modify the polarization.
 
  • #46
harrylin said:
Again: a pure filter only removes, it doesn't modify.
Perhaps a counter example is useful: An attenuator is a pure filter, as it doesn't modify the polarization.

I have yet to see where a polarizer modifies the polarization of individual photons.
 
  • #47
Drakkith said:
I have yet to see where a polarizer modifies the polarization of individual photons.

So if a photon polarized at 45° to the vertical passes through a polarizer set at 0° to the vertical, would you say that photon is still polarized at 45°?
 
  • #48
Joncon said:
So if a photon polarized at 45° to the vertical passes through a polarizer set at 0° to the vertical, would you say that photon is still polarized at 45°?

Yes. To my knowledge, at that polarization the photon still has a chance of passing through the polarizer. Another photon polarized also at 45 degrees might not make it through.
 
  • #49
Drakkith said:
Yes. To my knowledge, at that polarization the photon still has a chance of passing through the polarizer. Another photon polarized also at 45 degrees might not make it through.

OK, so if I hold two identical polarizers together, at the same angle, are you saying the light passing through is dimmer than if I only used one polarizer? i.e. a photon at 45° has a 50/50 chance of passing the first and then a 50/50 chance of passing the second.
 
  • #50
Drakkith said:
Joncon: So if a photon polarized at 45° to the vertical passes through a polarizer set at 0° to the vertical, would you say that photon is still polarized at 45°?

Yes. To my knowledge, at that polarization the photon still has a chance of passing through the polarizer. Another photon polarized also at 45 degrees might not make it through.

There is absolutely no evidence this is a true assessment. After any photon passes through a 0 degree polarizer, it will continue to pass similarly oriented polarizers - no problem. However, its orientation at 45 degrees is completely indeterminate. So I would NOT say it is oriented at 45 degrees. For if it were, it would pass 45 degree polarizers.
 
  • #51
DrChinese said:
There is absolutely no evidence this is a true assessment. After any photon passes through a 0 degree polarizer, it will continue to pass similarly oriented polarizers - no problem. However, its orientation at 45 degrees is completely indeterminate. So I would NOT say it is oriented at 45 degrees. For if it were, it would pass 45 degree polarizers.

Yes, that's what I was thinking.

Is it correct then to say that, while a polarizer acts like a filter (either letting photons pass or not), it also "rotates" the polarization of photons which pass successfully?
 
  • #52
Joncon said:
Is it correct then to say that, while a polarizer acts like a filter (either letting photons pass or not), it also "rotates" the polarization of photons which pass successfully?

It does appear to have that effect. Any quantum observation has this appearance (following the HUP). It is a bit difficult to know exactly what is happening during the process of collapse, and there are varying interpretations.
 
  • #53
DrChinese said:
It does appear to have that effect. Any quantum observation has this appearance (following the HUP). It is a bit difficult to know exactly what is happening during the process of collapse, and there are varying interpretations.

Are there any links/articles which go a little in depth about what happens during the collapse of the wavefunction ? I will be grateful if you know of any.

P.S: Sorry for having hijacked this thread, slightly.


-ibysaiyan
 
  • #54
Drakkith said:
I have yet to see where a polarizer modifies the polarization of individual photons.

The point of the demonstration with three polarizers - as also Joncon and others tried to explain - is that evidently photons that come out of the middle polarizer have a different polarization than the photons that enter it. If the photons coming out had the same polarization as the ones that enter it, then no light would come out. It's a basic characteristic of polarizers that they attenuate by changing the polarization, in contrast with pure attenuators that do not change the polarization.
 
  • #55
DrChinese said:
There is absolutely no evidence this is a true assessment. After any photon passes through a 0 degree polarizer, it will continue to pass similarly oriented polarizers - no problem. However, its orientation at 45 degrees is completely indeterminate. So I would NOT say it is oriented at 45 degrees. For if it were, it would pass 45 degree polarizers.

Ah, ok. I had not seen anything like that before.

I found this page that explains the effect: http://demonstrations.wolfram.com/LightBeamsThroughMultiplePolarizers/

harrylin said:
The point of the demonstration with three polarizers - as also Joncon and others tried to explain - is that evidently photons that come out of the middle polarizer have a different polarization than the photons that enter it. If the photons coming out had the same polarization as the ones that enter it, then no light would come out. It's a basic characteristic of polarizers that they attenuate by changing the polarization, in contrast with pure attenuators that do not change the polarization.

Yes, I see now. It would appear that a polarizer is both a filter and a converter then.
 
  • #56
kith said:
We have a bowl filled with objects in different colors and different shapes. If we filter out all red objects, there is no way to get any red objects back by filtering out other objects (cubes for example). This is the behaviour I expect from filters, but it's not the way polarizers behave. If I filter out one part of the whole wave (for example the 0°-polarized), I can get it (partially) back by filtering out another part (the 45°-polarized).

Consider this. The 1st polariser filters out red balls, the 2nd yellow balls, and the 3rd green balls. Now consider that red balls are equal to (yellow-green) balls, and green balls are equal to (yellow-red) balls. You'll see that the 2nd filter will "turn" green balls into (negative) red balls, but it's not adding anything, only subtracting the yellow component.

Maybe the bit that is confusing is that a light wave polarised in the negative "red" direction is just as visible as one polarised in the positive "red" direction. With real objects you can't see a "negative red" ball.
 
  • #57
sgb27 said:
Consider this. The 1st polariser filters out red balls, the 2nd yellow balls, and the 3rd green balls. Now consider that red balls are equal to (yellow-green) balls, and green balls are equal to (yellow-red) balls. You'll see that the 2nd filter will "turn" green balls into (negative) red balls, but it's not adding anything, only subtracting the yellow component. [..]

Your illustration doesn't work. For it to work, no ball at all must pass with two filters, and some balls must pass with an additional filter added between them. That won't be easy. :wink:

Harald
 
  • #58
harrylin said:
Your illustration doesn't work. For it to work, no ball at all must pass with two filters, and some balls must pass with an additional filter added between them. That won't be easy. :wink:

Harald

Assuming you just start with red and green balls, the red will be filtered out by the 1st polariser and leave the green unchanged, and the 3rd polariser (at 90 degrees) will filter out the green leaving no balls passing. The key point is that each polariser considers the "colours" relative to it's optical axis, the 1st and 3rd use red/green as their axes, and the 2nd uses another set of colours. I think that's about as good as you can get with this analogy.
 
  • #59
sgb27 said:
Assuming you just start with red and green balls, the red will be filtered out by the 1st polariser and leave the green unchanged, and the 3rd polariser (at 90 degrees) will filter out the green leaving no balls passing. The key point is that each polariser considers the "colours" relative to it's optical axis, the 1st and 3rd use red/green as their axes, and the 2nd uses another set of colours. I think that's about as good as you can get with this analogy.

Indeed the analogy only serves to show that a mere filtering doesn't work: either a filter filters out a full colour (which isn't the case in your example, the second filter distinguishes and filters a component), or it only filters out a component (which isn't the case in your example, the third filter doesn't let the yellow component pass).
 
  • #60
In terms of polarisers there is no such thing as a "full colour", ie an absolute or reference axis for polarisation. You must consider every case with reference to the optical axis of the material being considered.

Even in the analogy, who said that red was a "full colour" and yellow wasn't? Red can be made up of yellow and negative green, and also yellow can be made up of red and positive green. It all depends on the point of view of the specific filter/polariser as to what is the "full" colour. In terms of the 2nd polariser, yellow is the full colour, it no more or less full than the red the first polariser sees.
 
Back
Top