DCQE - how does/can the pattern change?

SpectraCat

First off, there's no call for any kind of abuse, and especially not calling people who are discussing things in good faith "idiots" or "crackpots". I am not a crackpot, and neither is Cthugha. That is quite an accusation, and you should be very careful before throwing it around so freely.

You and Cthugha have tangled before, and I have read those threads ... in my opinion he made his points much more clearly than you did, and supported them better with literature references as well. Everything he mentions is consistent with a QUANTUM explanation, not a simple classical one .. where do you think the well-defined phase of the two-photon state comes from, if not the fact that it is a pure quantum state, and therefore has a coherent phase-relationship between the photons? Furthermore, his explanation is completely consistent with the experimental results. He is not denying any of the claims or interpretations made in the papers themselves, he is only pointing out that there is less mysticism associated with the observed effects that is generally attributed to them by laymen and in the popular media.

Finally, it is worth pointing out that no one claimed that "coincidence counters do phase matching", so I don't know where you came up with that phrasing ... rather, Cthugha's point is that the interference that is observed arises from the well-defined phase relationship between the entangled photons. Unless you can come up with a detailed rebuttal to his arguments, I will remain convinced that he is correct, and you are the one who is confused.

zonde

Cthugha's analysis is clearly correct at least about one thing.
Postselection by coincidence counter has a key role in appearance of interference pattern.

That can be easily seen if you replace polarizer in idler beam with polarization beam splitter. Then you will have fringe and antifringe pattern at the same time just by looking at coincidences between signaling detector and one of the two detector at different outputs of PBS.

unusualname

Well you should read the threads again. When Cthugha first suggested the coincidence counters were to ensure classical (spatial) coherence between entangled pairs I thought he was being too ridiculous to argue with. You see you can't argue clearly with someone who has a wrong understanding of QM. And the fact that you think my arguments aren't made clearly is probably due to you not understanding QM either.

These sort of debates had merit maybe in the 1930s but in 2011 to have people still bewildered by QM is just tiring. Go and read a popular discussion of QM like Gribbin's "In Search Of Schrödingers Cat", it's a non technical explanation of what is accepted by all correctly thinking scientists today.

unusualname

and don't falsely state that peer reviewed references were provided to support an argument that the DCQE can be explained by classical phase relationships, there were none. There may have been some links to irrelevant results from quantum optics and an obscure german phd thesis (which has since gone offline), but that doesn't hide the basic fact the the DCQE has NO classical explanation. And no amount of obfuscation will fix that.

If you don't think QM is correct then you will have a hard time understanding the DCQE, and it's fruitless to argue with such people. There is no simple "explanation" of what is "happening", there is Quantum Mechanics and there are the various interpretations of it, and they are the best explanation you CAN have.

SpectraCat

Of course I think QM is correct, and so does Cthugha as far as I can tell. Neither of us has said anything that would suggest otherwise. Cthugha's analysis doesn't refute QM .. it requires it, as I said in my last post (which you didn't address). Specifically, it requires that entangled photons maintain a well-defined phase relationship over large (possibly spacelike, although that is not specifically addressed) separations. No attempt is made to explain that phenomenon, or interpret it, or rationalize it in terms of classical or local realistic arguments as you have implied, it is just accepted in the spirit of "Shut up and calculate", and then used to explain the experimental observations.

I am tired of seeing you post blanket "refutations" of this stuff without a shred of supporting detail. You just say things like, "Well, if you think <X>, then you're never going to understand <Y>, so I won't bother explaining it." You don't seem to understand Cthugha's arguments at all, because you insist on characterizing them as classical, and you have never said why any specific aspect of what he proposes violates any particular physical principle. Also, if you think spatial coherence somehow necessarily implies a classical explanation, perhaps you should review the double-slit experiment.

Finally, the "obscure Ph.D. thesis" was from Anton Zeilinger's group, and quantum optics is hardly irrelevant to what we are discussing. The dissertation in question also has not disappeared from the web .. it can be found here. It is true that Ph.D. theses do not undergo peer review in the same fashion as journal articles, but they have certainly been reviewed by the student's advisor and Ph.D. committee. Perhaps you think Anton Zeilinger also doesn't understand the DCQE?

unusualname

I'm not going to further argue with you, I could make up uncountable numbers of convoluted analyses of the DCQE and demand that you refute them. If the argument appears even vaguely in any peer reviewed literature then can you post a reference, thanks.

Drakkith

Ok, at this link: http://arxiv.org/PS_cache/quant-ph/p.../0106078v1.pdf which unusualname gave, I can see in figure 1 that the detector is moved to the left and right. I thought they were moving it back and forward. And you were right, the slit to the detector is only .3 mm wide.

Drakkith

Alright, let me see if I have this correct.

Initially you have two entangled photons travelling to seperate detectors, each linked to a coincidence counter. One detector we call S. The other we call P. Without the slits, the number of photons detected is at max with detector S right in the middle of the "beam". Moving it left or right results in a steady dropoff of detections.

Now we place a double slit between detector S and the source. Now, since the photons travelling to detector S interfere, moving the detector left or right slightly will show the interference pattern at each point as it would appear if you had a large detector capable of detecting the entire field of the beam instead of very small slit which only enables the detection of a small part at a time. (Meaning that the greater number of detections corresponds with a bright constructive interference band, and a lesser number of detections corresponds to a darker band)

So now we place a polarizer in each slit. One circularly polarizes them +45 and the other -45 depending on their initial linear polarization. Each slit results in the same interference pattern, only offset by 90 degrees. The built up inerference pattern is approximately the sum of the two different patterns, which looks like there isn't any interference pattern at all. (Page 5, http://arxiv.org/PS_cache/quant-ph/p.../0106078v1.pdf)

Alright, so now we place a polarizer between detector P and the source. If we set this polarizer to +45, then because of the entanglement the only photons on path S that can be detected at the same time as the ones at P are the photons that are originally polarized oppositely of P. If P is linearly polarized in the Y direction, which gets absorbed by the +45 polarizer, then it cannot get through to the P detector. The S photon that is polarized in the X direction gets through, strikes the detector, but because his twin didn't get detected, S is ignored by the coincidence counter. (I'm guessing that due to the doubled detection time once POL1 was placed in path P. Page 5, same link as above)

This results in only Y polarized photons striking the S detector at the same time as the P detector detects photons. So the resulting interference pattern observed at is what you would see if only Y polarized photons were allowed through the double slit at all.

If this is pretty much correct, I see absolutely nothing weird here. It looks like you have pre-selected which photons can or can't get to detector P, and via the coincidence counter determined which photons will be counted at detector S.

Edit: Spelling

Drakkith

According to http://en.wikipedia.org/wiki/Coincid...nting_(physics), and various other sites I have looked at, the counter is required not because of the probabilistic nature of QM, but because of the background noise and to make sure that the two particles you detect are indeed the ones that are entangled. Is this incorrect?

unusualname

Yes but what goes through POL1 ( the eraser) is probabilisitic. You have to double the detection time because only 50% of photons (approx) get through (Malus' law).

And there is no way to "know" which p photons go through and which don't.

And you can place POL1 and the p-photon detector distantly so that the corresponding s-photons are already measured (detected) before the p-photon can have even reached POL1 (so you might think the presence of POL1 should now not affect the s-photon measurements)

unusualname

it's partially correct, but you can reduce background noise and use efficient entangled pair sources almost to make these effects negliglible.

The main reason for the coincidence counters is that only ~50% of a randomly polarised source of photons will pass through a polarizer (Malus' Law) and you can't know which will and which won't, so you must use coincidence counters. (similar probabilistic laws apply to wave plates and other apparatus)

People get confused because classically we couldn't measure individual photon detections.

Drakkith

Maybe I'm missing some key concept here, but I don't see why knowing or not knowing which individual photons go through POL1 and into the P detector has any bearing on this. If the only time there is a count added by the coincidence counter is if the times for both detectors match up, then simply not letting Y polarized light through POL1 immediately means that you will never count any X polarized light at detector S. The coincidence counter doesn't get a count from detector P, so it ignores any count from S that were initally X polarized, correct?

Edit: Spelling

Drakkith

If you can get the background down to such low levels, why do you even need a coincidence counter? Wouldn't it be simpler to just record every detection? I understand that only 50% of the photons will get through a polarizer. (Ugh, been spelling polarizer wrong for a bit now) Why wouldn't you just count all the photons going into the detectors and look at the timestamp of each detection?

SpectraCat

That is not really correct. Coincidence counting is required because you have to correlate specific detection events at separate detectors with precise delay times in order to know with a high degree of certainty that two photons were generated as an entangled pair. Say for example you are sending one photon to Alice, who is 3 m away in the lab where the pair is generated, and the other photon to Bob, who is 150 m away in another building at the end of a fiber optic cable. So for a given entangled pair, the photons will arrive at the two detectors at different times ... you must have some way of knowing how to properly pair the separate detection events, and this is called coincidence counting.

The point about Malus' Law is something of a red herring, since most modern experiments with entanglement use polarizaing beam splitters (PBS's). A PBS sends photons with one polarization (say |H>) along one path to one detector, and those with the opposite polarization (say |V>) along a separate path to a separate detector. Both of those detectors are hooked to the coincidence counter, so both the |H> and |V> detection events can be captured in a single apparatus.

unusualname

Well that's what coincidence counters do! They measure all photons, and if two are detected within a certain time window the electronics signal a "coincidence". You could just use separate detectors and compare timestamps afterwards to recover your interference pattern if the technology was accurate enough, but you'll have to investigate the practical limits of the technology yourself to see if this is possible.

But, actually this type of coincidence counter will be needed to do experiments where the eraser and p-photon detector is placed so distant (much beyond the 2 meters in the Walborn experiment) that the time between p-photon and s-photon detections may be a second or more (if we can use satellites for example)

Drakkith

Ok, I understand and agree with that. I don't understand why any of this is called "Delayed Choice Quantum Eraser". It looks to me like you are preselecting which photons can and can't be detected. To me it all comes back down to the coincidence counter. No matter when the first photon is detected, if the counter doesn't receive a count from both detectors the detection doesn't matter. It is thrown out. The choice seems to be made by the detectors and counter after the fact, not before by the photons. Nothing is being "erased", it just isn't even being counted.

unusualname

No Malus' law is not a red herring it is the main reason why coincidence counters are needed.

Where is the PBS you describe in the DCQE experiments?

Even in the case of large distances for the p-photons (which hasn't been done in practice beyond a few meters for the DCQE btw) you will still have ~50% of p-photons not reaching the detector, even if you use fibre optics to reduce background noise and an efficient entangled photon source.