New loophole free EPR test with photons by Wittmann et al

[harrylin]

New "loophole free" EPR test with photons by Wittmann et al

Recently posted under Recent Noteworthy Physics Papers,
https://www.physicsforums.com/showthread.php?t=127314&page=10:

Bernhard Wittmann, Sven Ramelow, Fabian Steinlechner, Nathan K Langford, Nicolas Brunner, Howard M Wiseman, Rupert Ursin and Anton Zeilinger, "Loophole-free Einstein–Podolsky–Rosen experiment via quantum steering", New J. Phys. 14, 053030 (2012).

http://iopscience.iop.org/1367-2630/14/5/053030

Did anyone study that paper?

I quickly looked at that paper, and while it is said to have a high overall efficiency I also noticed mention of coincidence detection... so I wonder, is it indeed "loophole free" as is suggested, or would for example De Raedt et al's simulation program give about the same results?

 

[Demystifier]

Interesting!

But the result seems to be too important (and authors too famous) to be published in New J. Phys. instead of Nature, Science, or at least Phys. Rev. Lett.

So maybe there is still something fishy about the result ...

 

[zonde]

There are three somewhat similar papers:
Conclusive quantum steering with superconducting transition edge sensors
Journal reference: Nature Communications 3, 625 (2012)
Loophole-free Einstein-Podolsky-Rosen experiment via quantum steering
Journal reference: New J. Phys. 14, 053030 (2012)
Arbitrarily loss-tolerant Einstein-Podolsky-Rosen steering allowing a demonstration over 1 km of optical fiber with no detection loophole

All three submitted to arxiv on 3 Nov 2011

I have looked into all three of them but the first one is more interesting (at least for me).
Anyway all of them are testing "quantum steering" (instead of Bell inequalities, btw setup is basically the same for both). Try to guess what is this "quantum steering" and then we could compare how do we understand it.

 

[Cthugha]

One could also add A. Cabello and F. Sciarrino, "Loophole-Free Bell Test Based on Local Precertification of Photon’s Presence", Phys. Rev. X 2, 021010 (2012)
http://prx.aps.org/abstract/PRX/v2/i2/e021010

But the result seems to be too important (and authors too famous) to be published in New J. Phys. instead of Nature, Science, or at least Phys. Rev. Lett.

So maybe there is still something fishy about the result ...

Looking at the list of recent papers on loophole-free Bell tests, it seems to me that the authors of all these papers "conspired" to publish around the same time and explicitly avoided Nature/Science/PRL to publish in open access journals like PRX, Nat. Comm and NJP instead.

That might just be an odd coincidence, but at first sight it looks like open access was some priority in choosing the journals.

 

[zonde]

One could also add A. Cabello and F. Sciarrino, "Loophole-Free Bell Test Based on Local Precertification of Photon’s Presence", Phys. Rev. X 2, 021010 (2012)
http://prx.aps.org/abstract/PRX/v2/i2/e021010

Looking at the list of recent papers on loophole-free Bell tests, ...

The one you have given is proposal for Bell test. But the others are actual experiments of "quantum steering".
Please notice that test of "quantum steering" is not a Bell test!

 

[Cthugha]

Yes, of course.
I was just trying to point out that these papers cite each other and that there seems to be a general tendency towards open-access publication in that field. I was not trying to imply something about the scientific content or the validity of these papers.

 

[harrylin]

[..] Please notice that test of "quantum steering" is not a Bell test!

They seem to be closely linked and with similar claimed implications. I haven't figured it out yet, although the first paper to which you linked clarifies it somewhat better - thanks!

 

[Demystifier]

Please notice that test of "quantum steering" is not a Bell test!

So, if I understood it correctly, a loophole-free test of steering is not a loophole-free test of nonlocality. Is that correct?

 

[harrylin]

So, if I understood it correctly, a loophole-free test of steering is not a loophole-free test of nonlocality. Is that correct?

Hmm I understood just the contrary: that a loophole-free test of steering is also supposed to be a loophole-free test of nonlocality. At least, that's how I understand the paper's summary:

"Tests of the predictions of quantum mechanics for entangled systems have provided increasing evidence against local realistic theories. However, there remains the crucial challenge of simultaneously closing all major loopholes [..]. An important sub-class of local realistic theories can be tested with the concept of 'steering' [which] would seem to allow an experimenter to remotely steer the state of a distant system [..]. Einstein called this 'spooky action at a distance'. [..] we exclude—for the first time loophole-free—an important class of local realistic theories considered by EPR.[..]"

 

[zonde]

So, if I understood it correctly, a loophole-free test of steering is not a loophole-free test of nonlocality. Is that correct?

Yes
Pay attention that in quantum steering experiment measurements by Alice and Bob are always made in the same basis. This should tell you enough.

 

[zonde]

Hmm I understood just the contrary: that a loophole-free test of steering is also supposed to be a loophole-free test of nonlocality. At least, that's how I understand the paper's summary:

"Tests of the predictions of quantum mechanics for entangled systems have provided increasing evidence against local realistic theories. However, there remains the crucial challenge of simultaneously closing all major loopholes [..]. An important sub-class of local realistic theories can be tested with the concept of 'steering' [which] would seem to allow an experimenter to remotely steer the state of a distant system [..]. Einstein called this 'spooky action at a distance'. [..] we exclude—for the first time loophole-free—an important class of local realistic theories considered by EPR.[..]"

As I understand this sub-class of LHV theories assume that you detect one sub-sample for one measurement base and another sub-sample for another measurement base. But that sub-class of LHV theories should be ruled out by the very first experiments that performed Bell tests and roughly tested cos^2(theta) relationship IMHO.
Besides in quantum steering experiment they assume that Bob's side works according to quantum laws - this is complete mystery for me. Either you assume LHV or you assume QM or you assume neither. But how can you assume two supposedly conflicting approaches in single experiment and claim that one of them is ruled out by that experiment. It just seems such a crap that I am really confused.
On the other hand that setup with high efficiency thermal sensors is first of this kind and first reporting such a high heralded coincidence rate. But they give such an limited data from that setup.

And then this:
"The dominant source of optical loss, which leads to these less-than-optimal figures, was a splice between the single mode 820 nm fibres connected to the source and the fibres connected on the TES, which were single mode at 1550 nm."
There are latest coolest very high efficiency detectors and then 25% loss is introduced by guess what ... splice. This is such a disappointment.

 

[Demystifier]

Yes
Pay attention that in quantum steering experiment measurements by Alice and Bob are always made in the same basis. This should tell you enough.

Thanks for the tip, which does tell me enough.

Or maybe not quite. If both measurements are in the same basis, then how steering differs from "Bertlmann socks" (if you know what I mean)? In other words, how steering differs from classical correlations caused by local predefined properties of objects?

 

[Demystifier]

Hmm I understood just the contrary: that a loophole-free test of steering is also supposed to be a loophole-free test of nonlocality. At least, that's how I understand the paper's summary:

"Tests of the predictions of quantum mechanics for entangled systems have provided increasing evidence against local realistic theories. However, there remains the crucial challenge of simultaneously closing all major loopholes [..]. An important sub-class of local realistic theories can be tested with the concept of 'steering' [which] would seem to allow an experimenter to remotely steer the state of a distant system [..]. Einstein called this 'spooky action at a distance'. [..] we exclude—for the first time loophole-free—an important class of local realistic theories considered by EPR.[..]"

Yes, that's how I also understood it at first. But zonde's remarks suggested the opposite, and if he is right, that would explain why the paper is not published in a more prestigious journal, such as Nature, Science, or PRL.

 

[Demystifier]

Either you assume LHV or you assume QM or you assume neither. But how can you assume two supposedly conflicting approaches in single experiment and claim that one of them is ruled out by that experiment. It just seems such a crap that I am really confused.

I am quite used to the fact that experimentalists often keep mutually incompatible interpretations of QM at once. A good example is how experimentalists talk about delayed choice experiments, as something that "changes past" (which of course is a nonsense*).
So maybe we have a similar problem with interpretations-by-experimentalists here.

* See https://www.physicsforums.com/showthread.php?t=402497

 

[zonde]

Thanks for the tip, which does tell me enough.

Or maybe not quite. If both measurements are in the same basis, then how steering differs from "Bertlmann socks" (if you know what I mean)? In other words, how steering differs from classical correlations caused by local predefined properties of objects?

Predefined properties of particles in Bell sense determine measurement in any base. But in case of steering as I understand predefined properties determine measurement with certainty in only one base but in complementary base particle is simply not detected. That way we get perfect correlations at different bases.

EDIT: But wait, this was classical steering that they are falsifying. So quantum steering does not differ from "Bertlmann socks".

 

[Demystifier]

The more I read about steering, the more convinced I become that steering is no less proof of nonlocality than violation of Bell inequality is ...

In other words, the paper in New J. Phys. we discuss seems to be the first genuine loophole-free demonstration that nature is nonlocal. But then why New J. Phys.? Why not Nature? That I still don't understand.

 

[zonde]

The more I read about steering, the more convinced I become that steering is no less proof of nonlocality than violation of Bell inequality is ...

But in that case you see in steering more than authors are claiming: "An important sub-class of local realistic theories can be tested with the concept of 'steering'."
They talk about subclass only. Loophole-free Bell test on the other hand excludes all LHV explanations.

 

[Demystifier]

Loophole-free Bell test on the other hand excludes all LHV explanations.

Well, not all:
https://www.physicsforums.com/blog.php?b=3622

 

[Demystifier]

But in that case you see in steering more than authors are claiming: "An important sub-class of local realistic theories can be tested with the concept of 'steering'."

Does it mean that there is some type of LHV that COULD explain steering, but could not explain violation of Bell inequalities? If yes, what type of LHV it is?

 

[zonde]

Well, not all:
https://www.physicsforums.com/blog.php?b=3622

> - Copenhagen - nature is local, but objective reality does not exist (Bohr, Mermin, Rovelli-relational, Zeilinger, ...)
"Reality does not exist" is not an explanation it's more like antithesis of explanation.

> - many worlds - objective reality exists and is "local", but not in the 3-space (Everett, Deutsch, Tegmark, ...)
Simultaneous distant reality is not uniquely determined in MWI so how can you claim it's local?

> - superdeterminism - objective reality exists, it is local and deterministic, but initial conditions are fine tuned ('t Hooft)
Generally conspiracy theories are not considered.

> - backward causation - objective reality exists and is local, but there are signals backwards in time (transactional interpretation)
You can't propose oxymoron as an explanation.

> - noncommutative hidden variables - objective reality exists and is local, but is not represented by commutative numbers (Joy Christian)
There is no correspondence between noncommutative hidden variables based explanation and physical reality.

> - solipsistic hidden variables - objective reality exists and is local, but objective reality describes only the observers, not the observed objects (H. Nikolic, http://xxx.lanl.gov/abs/1112.2034 )
From solipsistic point of view you are now arguing with yourself.

 

[zonde]

Does it mean that there is some type of LHV that COULD explain steering, but could not explain violation of Bell inequalities? If yes, what type of LHV it is?

Bell's naive LHV model can violate steering inequalities.

 

[yoda jedi]

> - Copenhagen - nature is local, but objective reality does not exist (Bohr, Mermin, Rovelli-relational, Zeilinger, ...)
"Reality does not exist" is not an explanation it's more like antithesis of explanation.

> - many worlds - objective reality exists and is "local", but not in the 3-space (Everett, Deutsch, Tegmark, ...)
Simultaneous distant reality is not uniquely determined in MWI so how can you claim it's local?

> - superdeterminism - objective reality exists, it is local and deterministic, but initial conditions are fine tuned ('t Hooft)
Generally conspiracy theories are not considered.

> - backward causation - objective reality exists and is local, but there are signals backwards in time (transactional interpretation)
You can't propose oxymoron as an explanation.

> - noncommutative hidden variables - objective reality exists and is local, but is not represented by commutative numbers (Joy Christian)
There is no correspondence between noncommutative hidden variables based explanation and physical reality.

> - solipsistic hidden variables - objective reality exists and is local, but objective reality describes only the observers, not the observed objects (H. Nikolic, http://xxx.lanl.gov/abs/1112.2034 )
From solipsistic point of view you are now arguing with yourself.

"reality" in the case, argumenting that the term reality is defined by "properties".
.

 

[Demystifier]

Bell's naive LHV model can violate steering inequalities.

That seems interesting and important, but requires additional explanations.
1. What is Bell naive LHV model?
2. Where in the Bell's "Speakable ..." book is it described?
3. Is there a reference where it is explicitly shown (or at least claimed) that Bell's naive LHV model can violate steering inequalities?

Thank you in advance!

 

[zonde]

That seems interesting and important, but requires additional explanations.
1. What is Bell naive LHV model?
2. Where in the Bell's "Speakable ..." book is it described?

I don't have this book so if you don't mind I will use this link:
http://www.drchinese.com/David/Bell_Compact.pdf
Bell says: "..., it follows that the result of any such measurement must actually be predetermined."
and then this paragraph starting with "Let this more complete specification be effected by means of parameters λ."
Basically it says that experimental outcome is predetermined for particular angle of analyser by λ and nothing else. λ is shared between particles.

3. Is there a reference where it is explicitly shown (or at least claimed) that Bell's naive LHV model can violate steering inequalities?

?
Bell's naive model is based on the need to provide perfect correlations for matching angles. It starts from that point. Steering inequalities consider only matching angles. What else do you need.

Besides steering inequalities appeared rather recently so it would be up to authors of these inequalities to compare them with Bell inequalities, don't you think?

BTW have you seen any free access reference for that steering?

 

[yoda jedi]

non-locality from uncertainty principle.

..Likewise, nonlocality was traditionally thought of in terms of something called a Bell inequality, which also obscured the link between nonlocality and uncertainty...

... strength of a property called "steering'', which determines which states can be prepared at one location given a measurement at another...

http://arxiv.org/abs/1004.2507

 

[zonde]

Thanks yoda jedi, but the key person seems to be Wiseman so I suppose that most relevant papers should be in these search results:
http://arxiv.org/find/quant-ph/1/AND+au:+wiseman+ti:+steering/0/1/0/all/0/1


Say from this paper http://arxiv.org/abs/quant-ph/0612147
"In this Letter, we address the following questions: Can all entangled states be used to demonstrate steering? Does a demonstration of steering also demonstrate Bell-nonlocality? We prove that in both cases the answer is again: no. Thus, steerability is a distinct nonlocal property of some bipartite quantum states, different from both Bell-nonlocality and nonseparability."

Seems to answer Demystifier's question.

 

[harrylin]

Thanks yoda jedi, but the key person seems to be Wiseman so I suppose that most relevant papers should be in these search results:
http://arxiv.org/find/quant-ph/1/AND+au:+wiseman+ti:+steering/0/1/0/all/0/1


Say from this paper http://arxiv.org/abs/quant-ph/0612147
"In this Letter, we address the following questions: Can all entangled states be used to demonstrate steering? Does a demonstration of steering also demonstrate Bell-nonlocality? We prove that in both cases the answer is again: no. Thus, steerability is a distinct nonlocal property of some bipartite quantum states, different from both Bell-nonlocality and nonseparability."

Seems to answer Demystifier's question.

I'm still completely lost here... Can anyone explain in clear terms what the difference is between "Bell-nonlocality" and "nonlocal"? :uhh:

 

[Demystifier]

My current understanding of it (open to further refinements), compatible with the zonde's explanations, can be summarized as follows:
Steering can be explained by local hidden variables, but it cannot be explained by local reduced density matrix alone.

 

[zonde]

I'm still completely lost here... Can anyone explain in clear terms what the difference is between "Bell-nonlocality" and "nonlocal"? :uhh:

I would say that "Bell-nonlocality" is correlation that can not be explained by common cause.

 

[harrylin]

My current understanding of it (open to further refinements), compatible with the zonde's explanations, can be summarized as follows:
Steering can be explained by local hidden variables, but it cannot be explained by local reduced density matrix alone.

I would say that "Bell-nonlocality" is correlation that can not be explained by common cause.

As I understood all discussions until now, "nonlocal" means a correlation that can not be explained without something "spooky" such as instantaneous "action at a distance". And I think that with "Bell-nonlocality" people essentially refer to the same, but more specifically to "Bell" type of experiments and his arguments about those.

If so, then with the statement "steerability is a distinct nonlocal property", the Arxiv letter http://arxiv.org/abs/quant-ph/0612147 by Wiseman claims that steering implies "nonlocality" in the sense of EPR - so that steering is claimed to be unexplainable with "local variables".

 

[Demystifier]

As I understood all discussions until now, "nonlocal" means a correlation that can not be explained without something "spooky" as instantaneous "action at a distance". And I assumed that "Bell-nonlocality" is essentially the same, but with precise arguments that lead to that conclusion as presented by Bell.

Yes, that's correct.