does different observer affect double slit particle experiment

Cthugha

I suppose you mean two photons (one signal, one idler) and repeatedly sending these two photons down the experiment such that there are never more than these photons present in the setup. Even at arbitrarily low count rates you will never see an interference pattern looking at the results of one detector alone under DCQE conditions.

You can have an interference pattern in the single photon count rates at one detector under spatially coherent illumination. That, however, breaks entanglement.

Or you can have a coincidence pattern in the coincidence counts only without seeing one in the single count rates. That requires entangled particles which means incoherent conditions.
You cannot have both, so these requirements are complementary.

Your earlier post sounds as if there were interference patterns spontaneously appearing or disappearing. That is not the case. There are no direct interference patterns visible at any of the detectors used. You really need to have a look at the pattern that arises when detector A is placed at some position, detector B is spatially scanned and the coincidence counts are evaluated. If you now place detector A somewhere else and move detector B around again, you will see another interference pattern in the coincidence counts which is different from the first one. You can continue like that for any position of detector A and will get many different coincidence count patterns. The sum over all of these patterns gives no interference patterns at all - these are only visible in conditioned coincidence detection.

shahzk

thanks... i meant one photon that is split into two... so what would you see at detector A where the signal photon hits in a dark room and you place detector b a light year away?

This physicist john cramer doesn't seem to understand this as an inherent limitation to coincidence counting... i can't find any results of his experiment online, but seems from this that avoiding coincidence counts is a practical obstacle to be overcome:

http://www.seattlepi.com/local/artic...st-1219821.php

here it also described just as a way to distinguish entangled from non-entangled:
http://en.wikipedia.org/wiki/Coincid...nting_(physics)

Cthugha

It would just be detected somewhere at detector A. Nothing special happens.

The reason why there are no published results is most likely that avoiding coincidence counts is not just a practical obstacle to be overcome. Cramer provides a strange interpretation which is not backed up by any facts, but sounds spectacular. Show me published results supporting his speculation and there is something to discuss.

Of course you can also use it also as a means to distinguish entangled photons from noise. However, doing so is not sufficient. If you want to read the whole story in a thorough manner, have a look at http://arxiv.org/abs/1010.1236, which is the ArXiv version of Physics Reports Volume 495, Issues 4–5, October 2010, Pages 87–139. This explains the whole topic better than I could. Especially the chapter on spatial correlations and conditional interference is a good starting point. Also, most of the good references are cited within that paper.

The situation in DCQE is basically the same and also a situation where the exact shape of the interference pattern seen in coincidence counting depends on the position of BOTH detectors. Having just one detector present is equivalent to averaging over all possible positions of the other detector which cancels any interference pattern. Wikipedia is also not really a good source for scientific debates, by the way.

shahzk

This is the latest news i can find of cramer's experiments on his website from 2010... toward the end of the ppt he explains that practical issues (e.g. noisy detectors) have not allowed the retro-causality experiments to work yet, but he doesn't believe these are inherent show-stoppers.

http://faculty.washington.edu/jcrame...n_20100403.ppt

Cthugha

He is free to believe what he wants to. That does not make his view which is based on Wheeler's point of view any more correct.

Even more recent publications explicitly oppose that point of view.
From Phys. Rev. Lett. 107, 230406 (2011):

"Discussing the delayed-choice experiment, Wheeler concludes: ‘‘In this sense, we have a strange inversion of the normal order of time.We, now, by moving the mirror in
or out have an unavoidable effect on what we have a right to say about the already past history of that photon’’ [5]. We disagree with this interpretation. There is no inversion
of the normal order of time [...]"

Together with the really good reference I have given you before, it should be clear where he goes astray - it of course takes some time to read and grasp the topic, I must admit. However, it does not make much sense here to discuss bare claims which are not backed up by any experimental evidence. By the way Cramer's claim that coincidence counting is necessary only to overcome noise comes from the time when he noticed an experiment performed by Birgit Dopfer in the group of Zeilinger. These results also stressed the importance of complementarity and noted that single-photon and two-photon interference patterns are complementary. Just do the theory and eliminate the noise and integrate over all possible detector positions and you will notice that there cannot be an interference pattern under DCQE conditions without coincidence counting.

Spence

I have my best answer to the original question, plus 3 questions of my own.
I personally don't think the intended "animal" question posed at the beginning of this thread has been quite answered yet. Please allow me to restate my interpretation of the question:

In the double-slit experiment, assuming a generic non-human animal can accurately relay which-path information to a human, but no human is permitted to see the which-path info, will there be an interference pattern?
For example, the machine detecting which path each photon passes through commands a perfectly obedient dog to "sit" or "shake" based upon which slit the photon passes through, and humans observe the dog's actions.
(I am not asking about Schroedinger's cat per se, or about the background light "noise" in the experiment- those topics would be somewhat irrelevant to this question.)

Correct me if I am wrong, but from every delayed choice/double slit/quantum eraser experiment I've read about, the results generally seem to agree on the following conclusion:
It's not the act of observing per se, NOR is it the act of taking a measurement (detecting), that disrupts the interference pattern. Though these conditions likely must be met to accomplish disruption, they alone are not sufficient. The disruption occurs when an observer understands WHICH slit the photon travels through.


So getting back to the question- based on that last statement, I believe the experiment I described would result in the photons behaving like matter, rather than creating the interference pattern. I say this not because the animal itself is disrupting the interference, but because humans are gaining the which-path knowledge by observing which command the animal follows. In other words, the perfectly-trained animal is essentially just an extension of the which-path detectors.


So, building on that, I ask 3 questions about the double-slit experiment:
1. Am I correct with my prediction? If not then, in layman terms, why not?
2. Assume we have a chimp that can look at the which-path results and understand that something is passing through 2 possible paths. Does this disrupt the interference pattern? Does the chimp need to generally understand what is happening in the experiment in order for the disruption to occur?
3. How much about the experiment does one need to understand in order to disrupt the interference and make the photons behave as matter? For example, say we detect our which-path information, and before anyone else can see it we send the which-path info (whatever info the experiment conductors normally draw conclusions from) to Grandma, who is unaware such an experiment even exists. Assuming that Grandma continues to observe the data without knowing what it means, would the waves make an interference pattern?

I appreciate anyone's input!

DrChinese

Welcome to PhysicsForums, Spence! This is an old thread, kinda dead and buried. Next time it would probably make sense to start a new one instead...

I will pass on a few thoughts regarding interference vs no interference. The rules are:

a) If you know, or could have known, in principle, which slit was traversed, then there will be no interference. This has been demonstrated in every experiment, and is the master rule.
b) As far as anyone knows, the presence or absence of an intelligent observer makes no difference. Certainly you can record the event and look at it a week later, and rule a) above will still apply.
c) Notwithstanding b), you could always assert that the pattern changed precisely because an intelligent observer looked at the pattern. You could never prove otherwise, but I would call this an assertion which is not scientific in nature.
d) In DCQE experiments, the ordering of the events is immaterial to the outcome.

morrobay

What about this explanation for the disappearance of the interference pattern ?
http://www.physicsforums.com/showthread.php?t=662920 post#5 by Dali

Zzzfilesk1

Give 2 humans in separate rooms opposite results, at random, during each run of the experiment.

No one will know which observer is fed the correct results.

If both observers report the experiment behaved accordingly 100% of the time, you have proof that the observer controls the outcome.

Otherwise, you have proof that measurement collapsed the wave function.

If you get mixed results, perhaps one person's conciousness has more influence than the other.

rajeshmarndi

Ok, I haven't thought of that i.e what if wrong information is fed into the detector without the knowledge of the experimenter. Does the experimenter still get the result according to his information of the particle from the detector?