Does QM itself prevent removal of noise without comparison

[San K]

Is there some fundamental fact of the universe (Quantum Mechanics) that prevents us from removing noise (without having to compare the entangled particles)?

In the quantum eraser experiments:

To get ("un-embed") the interference pattern i.e. to separate the entangled particles from the noise both the entangled particles have to be compared via a coincidence counter.

Is there a way to eliminate noise in advance? i.e. without having to compare the entangled particles

 

[mfb]

Noise can appear from random photons flying through your setup, or other background sources. You can reduce the number of background photons, but no experiment is perfect.
Do you have some specific requirement for low noise in mind?

 

[San K]

Noise can appear from random photons flying through your setup, or other background sources. You can reduce the number of background photons, but no experiment is perfect.
Do you have some specific requirement for low noise in mind?

In a quantum eraser experiment (for example) involving entangled photons (signal and idler)

If we were able to get (close to) zero noise then what stops us from sending information by performing which-way (or no-which-way) on one of the photons and observing it on its distance twin?

*without having to compare the photons, via co-incidence counter *

let me know if you want me to explain the experiment in detail

 

[mfb]

Wait - what do you consider as noise?
The analysis of quantum eraser experiments can be done without considering any background.

 

[San K]

Wait - what do you consider as noise? The analysis of quantum eraser experiments can be done without considering any background.

In my definition of noise:

It's all those photons other than the entangled ones.

in other words

It is all those photons that are not entangled...



We want such photons somehow removed/filtered ...*without the use of a co-incidence counter or similar apparatus*

the below is not from literature but a construct of my imagination

Noise type 1 -- stray photons
Noise type 2 -- photons that get un-entangled prior to detection
Noise type 3 -- un-entangled photonsAlternatively

In the quantum eraser experiments that photons need to be compared via co-incidence counter in order to get the interference pattern.

Now, if there were no noise then :

could we get an interference pattern (from the idler photon) without having to compare?

 

[mfb]

You should specify which experiment exactly you mean. Simple quantum eraser experiments themself do not need any coincidence counter.
If (and only if) you use entangled photons, you always get photons which are not entangled as well - this is not related to the experiment itself, it is just a problem of the production of those photons. If you do not remove those from the experiment, you get a very small interference pattern on top of something else (from single photons without partner).

 

[Cthugha]

In the quantum eraser experiments that photons need to be compared via co-incidence counter in order to get the interference pattern.

Now, if there were no noise then :

could we get an interference pattern (from the idler photon) without having to compare?

You are discussing DCQE here. No, you could not get interference patterns looking at the idler side alone because the first-order coherence properties of the light field in question are not sufficient to create interference patterns under the experimental conditions needed for such experiments. This is not a question of noise. Or Simply put: the light is just too incoherent.

 

[San K]

You should specify which experiment exactly you mean. Simple quantum eraser experiments themself do not need any coincidence counter.
If (and only if) you use entangled photons, you always get photons which are not entangled as well - this is not related to the experiment itself, it is just a problem of the production of those photons. If you do not remove those from the experiment, you get a very small interference pattern on top of something else (from single photons without partner).

MFB, As I open physicsforums to answer to your post above, I see that Cthuga has answered my question. though I have to understand/learn what first order coherence means.

thanks Cthugha and Mfb

nevertheless I think I should still respond to your/mfb post above:

an example of an experiment could be:

http://en.wikipedia.org/wiki/Delayed_choice_quantum_eraser

So my question is if we assume in the future we have devices a way to have only entangled photons emanate from the crystal via SPDC, could we then have an interference pattern without having to compare the entangled/twin photons?

because if that was possible then information could be sent ftl (i.e. Alice could signal to Bob, by doing which-way or no-which-way on her photon(s) and Bob would see the effect instantaneously on his other half of the entangled photons in the form of interference pattern or no-interference pattern) <--- is the logic correct (even though the assumption might be wrong)?

Note: I have asked this question before however the answer I got was it was not possible.
Ctugha has given a clear answer -- it has something to do with first order coherence, which I have to read up on.

 

[San K]

This is not a question of noise. Or Simply put: the light is just too incoherent.

Sorry I don't get this. why does the light get too incoherent?

also we are talking about single photons, with whom do single photons need to be coherent with?

 

[Cthugha]

it has something to do with first order coherence, which I have to read up on.
[...]
Sorry I don't get this. why does the light get too incoherent?

also we are talking about single photons, with whom do single photons need to be coherent with?

Ok, first-order coherence is spatial or temporal coherence. This is what you test in a double slit experiment (spatial coherence) or a Mach-Zehnder-interferometer (temporal coherence). It is a measure of the time scale or spatial scale over which the phase of a light field randomizes. The first-order coherence function in time is basically the Fourier transform of the spectral width of your light field. Therefore, if the emission is very narrow in energy, it will have a long coherence time and if it is very broad, the coherence time will be short.

For spatial coherence, you get a similar relation with the angular width of the source as seen by the double slit. The broader the range of angles is, from which the light field may reach the double slit from the light source, the smaller the spatial coherence will be. You can try that yourself. build a simple double slit experiment and vary the distance between the light source and the double slit. If the distance becomes too small, the interference pattern will vanish. This is also why a pinhole was used prior to the double slit in the original Young double slit experiment. Both temporal and spatial coherence apply to ensembles of single photons as well.

Now the problem is that you want a large spread of emission angles for (momentum)-entangled light because this is the entangled quantity. Filtering a small range of emission angles therefore reduces the entanglement. However, the entanglement gives you second-order coherence of the total light field (which is what creates the interference pattern in coincidence counting experiments). It turns out that one cannot have both - entanglement and first-order coherence - at the same kind in these setups.

A detailed description is given in: "Demonstration of the complementarity of one- and two-photon interference", Phys. Rev. A 63, 063803 (2001), also available on the ArXiv for free: http://arxiv.org/abs/quant-ph/0112065.

 

[San K]

Ok, first-order coherence is spatial or temporal coherence. This is what you test in a double slit experiment (spatial coherence) or a Mach-Zehnder-interferometer (temporal coherence). It is a measure of the time scale or spatial scale over which the phase of a light field randomizes. The first-order coherence function in time is basically the Fourier transform of the spectral width of your light field. Therefore, if the emission is very narrow in energy, it will have a long coherence time and if it is very broad, the coherence time will be short.

For spatial coherence, you get a similar relation with the angular width of the source as seen by the double slit. The broader the range of angles is, from which the light field may reach the double slit from the light source, the smaller the spatial coherence will be. You can try that yourself. build a simple double slit experiment and vary the distance between the light source and the double slit. If the distance becomes too small, the interference pattern will vanish. This is also why a pinhole was used prior to the double slit in the original Young double slit experiment. Both temporal and spatial coherence apply to ensembles of single photons as well.

Now the problem is that you want a large spread of emission angles for (momentum)-entangled light because this is the entangled quantity. Filtering a small range of emission angles therefore reduces the entanglement. However, the entanglement gives you second-order coherence of the total light field (which is what creates the interference pattern in coincidence counting experiments). It turns out that one cannot have both - entanglement and first-order coherence - at the same kind in these setups.

A detailed description is given in: "Demonstration of the complementarity of one- and two-photon interference", Phys. Rev. A 63, 063803 (2001), also available on the ArXiv for free: http://arxiv.org/abs/quant-ph/0112065.

Thanks for the great post and reference paper, Cthuga. I am going over it and it will take me some time.

I wonder if entanglement and (first-order) coherence have anything to do with complementarity.

or alternatively

if we could figure out a way to keep the coherence (at required levels) would that mean we can send information without having to compare (via co-incidence counter)

would that, in turn, mean

information could be sent FTL?