Wikipedia

Talk:Afshar experiment

This is an old revision of this page, as edited by Afshar (talk | contribs) at 18:30, 30 April 2005 (→Reply to Unruh question). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

Learn more about this page

OK, page created. I consolidated information which was spread (in many case, as multiple verbatum or near-verbatum copies of the same information) across many pages. The pages I edited to consolidate this information include:

I also removed red links to other suggested names for this page.

Next: Get rid of some links from Double-slit experiment which only apply to the Afshar experiment. Samboy 10:26, 21 Dec 2004 (UTC)

Done. Samboy 12:30, 21 Dec 2004 (UTC)

Afshar Experiment: A Dissenting Opinion

The Afshar experiment demonstrates that a coherent path distribution must be assumed to predict the number of photons incident on measurement spaces where path information is being observed, but it does not demonstrate that the coherent path distribution is actually realized across the surface of measurement spaces where path information is being observed.

All Figure number references in this article are references to Afshar’s Figure numbers in his paper “Sharp complementary wave and particle behaviours in the same welcher weg experiment”: http://www.irims.org/quant-ph/030503/

In Afshar’s experiment, the two holes can be considered emitters in the sense that he pumps photons through these two holes with his laser, and these photons have to go somewhere. Assuming the Copenhagen Interpretation of different path distribution functions (coherent versus decoherent) for different type of measurements (visibility versus path information), one might ask the question how do these photons get distributed in experiments involving complementary measurement spaces (i.e. how many photons get mapped into each measurement space). This question is also relevant in an experiment that involves only a single measurement space where path information is being measured. The answer of course is that a global coherent path distribution function across all measurement spaces must be assumed to predict the photons incident on each measurement space. The number of photons incident on a measurement space is predicted by integration of the global coherent path distribution function across the surface of the measurement space. But the Copenhagen Interpretation indicates the predicted number of photons incident on a measurement space where we are measuring path information don’t actually exhibit a coherent distribution locally.

So when Afshar says that he has visibility of interference on the photons incident on the lens, if he means that a coherent distribution predicts the number of photons incident on the lens surface, then this is correct. But if he is saying that the photons incident on the lens surface actually exhibit a local coherent distribution across the surface of the lens, then this is incorrect.

In Afshar’s experiment there are 13 measurement spaces. There are 6 measurement spaces where visibility information is being observed corresponding to the 6 wires. There are 7 measurement spaces where path information is being observed corresponding to the 7 portions of the lens segmented by the wires. The number of photons incident on each of the wires is very low because the wires are very thin and Afshar has placed them at the point of maximum destructive interference of the coherent path distribution function. The number of photons incident on each of the lens segments can be predicted by integration of the coherent path distribution function across the face of the lens segment. If the results for all of the lens segments are added up, the result is very close to the value for the whole lens when the wires are absent, since only the portions of the total lens where there is maximum deconstructive interference in the coherent path distribution function are omitted in this integration process.

Afshar’s interpretation of his experiment seems to be that since the wires are very thin and he has carefully placed them at the points of maximum destructive interference, he can treat his experiment as a single measurement space. Since he is observing interference in the photons incident at the wires, he concludes that there must also be interference present in the photons incident on the lens, since he believes both sets of photons are part of the same measurement space.

Even though the wires are very thin, the photons incident on the wires must be treated separately from the photons incident on the lens, since different types of measurements are being made in the two cases, and therefore the photons exhibit different behaviors for the two cases. For the photons incident on the wires, visibility information is being observed, and therefore the photons incident on the wires go through both holes with a corresponding coherent path distribution that exhibits interference in the vicinity of the wires. For the photons incident on the surface of the lens, path information is being observed, and therefore the photons incident on the lens only go through one hole with a corresponding decoherent path distribution.

The argument that the local path distribution across each of the 7 lens segments must be decoherent consists of the following four points:

1) The photons only go through one hole and are resolved at the detectors. Therefore any interference pattern in front of the lens cannot be explained by self-interference.

2) The only other way to explain a coherent path distribution in front of the lens is for a single hole to be emitting a coherent path distribution function. Hopefully we would all agree that this is about as likely as hearing the sound of one hand clapping.

3) If a single hole is (somehow magically) emitting a coherent path distribution function, then it would manifest itself in an interference pattern at the detector image.

4) Since there is no interference pattern at the detector image in Figure 8a, a single hole must not (somehow magically) be emitting a coherent path distribution function.

The third point above is illustrated by Figure 8b where Afshar has induced some of the characteristics of a coherent distribution sourced by a single hole by placing the wires in front of the lens at the points of maximum destructive interference. The image in Figure 8b clearly indicates interference at the detector. The interference pattern is also still clearly visible in Figure 8c, but it does seem to be attenuated somewhat from Figure 8b. Afshar’s explanation for this attenuation is that less photons are incident on the wires in Figure 8c than in Figure 8b, which is certainly true. In summary, Figures 8b and 8c indicate an interference pattern, but this interference is the result of placing the wires in front of the lens. Figure 8a illustrates clearly that there is no interference pattern at the detector without the wires present.

So how does Afshar explain his claim of a coherent distribution across the face of the lens? His explanation seems to be a superposition behavior in the neighborhood of the holes (that the photon goes through both holes exhibiting interference and it also goes through only one hole exhibiting path information). There is a superposition of states before the wave function collapses. Before the wave function collapses, there is a possibility that the photon goes through both holes with a corresponding coherent path distribution, and there is the possibility that the photon goes through one hole with a corresponding decoherent path distribution. Afshar seems to be assuming a superposition of two states at the two holes, but only a single potential path distribution. When we make an observation, the wave function will collapse in a way that depends on the nature of the measurement we are making. If we are measuring visibility information, the wave function will collapse such that the photon goes through both holes with a corresponding coherent path distribution. If we are measuring path information, the wave function will collapse such that the photon goes through one hole with a corresponding decoherent path distribution.

Afshar attributes far too much magic to the thickness and placement of his wires. A couple of examples with different thickness and placement are provided below to illustrate this point. Both of these examples assume a peak-to-peak distance of u = 1.4 mm for the consecutive fringes similar to Afshar’s first experiment illustrated in Figure 1.

For the first example, replace the wires with thin strips with a width of 1.4 mm leaving these strips centered about the maximum point of deconstructive interference. These strips will block some of the portions of the lens where the value for a coherent path distribution is less than the value for a decoherent path distribution (i.e. some of the valleys of a coherent distribution). All portions of the lens where the value for a coherent path distribution is greater than the value for a decoherent path distribution will be left exposed (i.e. all peaks of a coherent distribution). Next determine the predicted attenuation of the radiant flux at the image based on a decoherent path distribution either by calculation, simulation, or experimental measurement. Finally open the second hole and measure the actual attenuation in radiant flux at the image and compare with the value predicted in the previous step. The measured results will be much less than the value predicted by assuming a decoherent path distribution.

For the second example, keep these same strips, but move them to a position where they are centered about the peaks of a coherent distribution instead of the valleys of a coherent distribution. These strips will now block some portions of the lens where the value for a coherent path distribution is greater than the value for a decoherent path distribution (i.e. some of the peaks of a coherent distribution). All portions of the lens where the value for a coherent path distribution is less than the value for a decoherent path distribution will be left exposed (i.e. all of the valleys of a coherent distribution). Go through all of the same steps as in the first example. This time the measured results for the reduction in radiant flux at the image will be much greater than the value predicted by assuming a decoherent path distribution.

Both of these examples demonstrate the same results as the Afshar experiment (i.e. a coherent path distribution function must be assumed to prediction the reduction in photons incident on the detectors). Therefore there is no magic in using very thin wires placed at the points of maximum deconstructive interference.

In summary, the Afshar experiment demonstrates that a coherent path distribution function must be assumed to predict the number of photons incident on measurement spaces where path information is being observed. According to the Copenhagen interpretation, the assumed coherent path distribution function actually collapses into a decoherent path distribution function across the surface of measurement spaces where path information is being observed. The Afshar experiment provides no evidence that the Copenhagen interpretation is incorrect.

Originally contributed by User:63.226.32.16 to Talk:Double-slit_Experiment; I copied it here since it is appropriate to place here Samboy 19:38, 22 Dec 2004 (UTC)


Afshar's Response

The bottom line for the above Dissenting Opinion is the following false conclusion (quoted):

"According to the Copenhagen interpretation, the assumed coherent path distribution function actually collapses into a decoherent path distribution function."

Unfortunately, the major error in this argument is a misunderstanding of what the Wavefunction collapse actually means. Without going into mathematical details, a coherent superposition state can only collapse into an observable with a coherent distribution when a measurement is made. There is no "collapse" from a coherent wavefunction to a decoherent state because upon measurement a coherent wavefunation Psi is mapped into an observable wihtin |Psi|^2 via a projection, and not into some other decoherent distibution, say |Psi_1|^2 + |Psi_2|^2 . If the above quotation were true, then we could not observe an interference pattern in a double slit experiment, or immediately after the wires in my experiment, by direct observation, but we do! We cannot change the definition of wavefunction collpase to an arbitrary and mathematically inaccurate one, just to save Complementarity! Afshar 00:06, 23 Dec 2004 (UTC)

It is quite telling that the anonymous writer of the Dissenting Opinion (whose name is actually Alex), having now realized that his definition of the wavefunction collapse was indeed erroneous, has again attempted to “correct”his conclusion by replacing the pervious version with the following: "According to the Copenhagen interpretation, the photons predicted by the assumed coherent path distribution function actually exhibit a decoherent path distribution function locally across the surface of measurement spaces where path information is being observed." I'm afraid, the more he awkwardly tries to avoid self-contradiction, the more his lack knowledge on even the most rudimentary QM formalism and language becomes apparent. There is no such thing as "coherent path distribution function" in QM. I suggest that he formally study QM before making more clumsy statements. Need I say more?! Afshar 07:50, 23 Dec 2004 (UTC)

The school between Central and Porter squares

Of course one could miss the stop, and coming from Kendall sq end up say in Davis sq or Alewife. And even if by luck, one did get off on the right stop, and succesfully finding one's way out of the station one might think one ended up instead in an upscale urban mall complete with the Gap and a really fancy Barnes and Noble (which actually has real books on math).

But why is it necessary to mention this place in the very first sentence of the article? CSTAR 22:55, 23 Dec 2004 (UTC)

Are you sure this has to do with the Afshar experiment? Samboy 23:43, 23 Dec 2004 (UTC)
Yeah. it's a snarky comment at the abundant mentions of Harvard(on at leat two sides). Is it really necessary? CSTAR 23:49, 23 Dec 2004 (UTC)
Yes! Thanks for the graphics. Afshar 23:49, 23 Dec 2004 (UTC)

The pictures look great

This article looks great with the pictures and the mathematical formulas! Samboy 23:46, 23 Dec 2004 (UTC)

Definitions

The relevant definitions need to preceed (e.g. V , Imax) and so on, in order for the critique to make any sense to the non-specialist reader of the article. CSTAR 00:02, 24 Dec 2004 (UTC)

Where to start?

The article keeps saying that Afshar's results are incompatible with quantum mechanics, that Unruh challenged and led some to doubt Afshar's results, and that if Afshar's results are correct then they have far-reaching implications. As far as I can tell, this is not right. It seems to be Afshar's interpretation of the results, and indeed his understanding of quantum mechanics, which are in question. Unruh's write-up, for example, concludes as follows: "I think Bohr would have had no problem whatsoever with this experiment within his interpretation. Nor would any other interpretation of quantum mechanics. It is simply another manifestation of the admittedly strange, but utterly comprehensible (it can be calculated with exquisite precision), nature of quantum mechanics." The current wording gives quite a misleading impression. -- Reuben 03:03, 3 Feb 2005 (UTC)

  • Dear Reuben, you say above: "The article keeps saying that Afshar's results are incompatible with quantum mechanics." I can only conclude that you have not read the article carefully. Where exactly is such an assertion made in the article?! As I have said many times in different interviews, my experiment indeed confirms QM formalism, but it disproves Bohr's Complementarity by showing that his "principle" does not follow from QM formalism at all. Sorry, but there is no misleading language here! --Afshar 03:54, 3 Feb 2005 (UTC)
  • Fine, it doesn't say "incompatible with quantum mechanics." To be precise, it says "the Copenhagen interpretation" instead of "quantum mechanics." It does say the following, all of which are misleading.
"If his results are verified, it has far-reaching implications for the understanding of the quantum world, and invalidates the Copenhagen interpretation."
"On August 7, 2004, Bill Unruh presented an argument in which he claims to disprove Afshar's results."
"He demonstrates that his experiment is consistent with the Copenhagen interpretation and, on that basis, argues that Afshar's results are incorrect."
It's not any experimental results that are in question, it's Afshar's understanding of them. -- Reuben 04:47, 3 Feb 2005 (UTC)
  • Naturally, you should read "results" as representing both novel experimental observations and novel theoretical arguments presented in the preprint. And no, Copenhagen interpretation is certainly not the same as quantum mechanics... Afshar 06:20, 3 Feb 2005 (UTC)
  • There's nothing particularly novel about the experimental results, since they are as you say exactly what quantum mechanics predicts. Only Afshar's interpretation and understanding is at issue. If that's what the article means, it should clearly say so. The current language is misleading. -- Reuben 06:38, 3 Feb 2005 (UTC)
  • You may wish to look up the adjective "novel" in a dictionary. Simply because a particular line of argument follows from an already accepted formalism, does not mean it is not a novel one! Einstein's Special Relativity was implicit in Maxwell's equations, but it was aptly considered as a novel result. Edison’s light bulb also followed from Faraday’s work, but it was a real invention… The non-perturbative measurement scheme introduced for the first time in my preprint is indeed a novel concept with no precedence in the literature. If you wish to dispute my claim, I would appreciate a reference please. Afshar 07:05, 3 Feb 2005 (UTC)
  • The precedent would be passing a vertically polarized beam of light through a vertical polarizer, and then measuring the final polarization along an axis 45 degrees from the vertical. Your setup is formally equivalent. Your scheme is no more "non-perturbative measurement" than passing a vertically polarized beam through a vertical polarizer is. But on a more important note, you don't seem very interested in my suggestion for making the text less misleading. Why is that? -- Reuben 07:53, 3 Feb 2005 (UTC)

I don't really see how this is supposed to violate wavefunction collapse. Perhaps you can try to clarify. As I see it, the photon travels through the two slits, interferes, the interference pattern travels through the carefully arranged dark-fringe wires, and then the photon collapses upon being detected at one of the detectors. Now, obviously if the wavefunction collapse occurred at the time of measurement, then that photon has reached that detector and so possibility of it having been stopped by the wires is erased. It seems that the argument is being made that because the path of the photon is being determined, that somehow it should go back in time and reestablish a chance to collide with the wire. But this does not seem to logically follow. In a standard delayed choice experiment, the path the photon took is selectively determined after the photon has already passed through the slits, but this at no point goes back in time and arranges something inconsistent with the measurement that was made. So since in order for collapse to occur, the photon has to strike the detector, that collapse already rules out the possibility that it struck the wire. The interference pattern at the detector is only determined by the past of the photon, not the future, because just like in the delayed choice experiment, the detector arrangement could be replaced by a screen. — Cortonin | Talk 06:40, 3 Feb 2005 (UTC)


  • Please post your question in my weblog. I will reply to it there and then make an addition to the Wiki article to include the wavefunction collapse issue. We are conducting another experiment which directly addresses your concern. I may be able to discuss it in April.--Afshar 06:49, 3 Feb 2005 (UTC)
Well if it doesn't yet address that wavefunction collapse issue, then maybe we should avoid saying it violates wavefunction collapse on Wikipedia until the further experiments are finished. — Cortonin | Talk 06:54, 3 Feb 2005 (UTC)
Please do take a look at my weblog, Cramer's article and my NPR interview. The mention of the collapse is quite justified, and I will improve the article further. I do not think it should be removed just because the article is a live one! --Afshar 07:14, 3 Feb 2005 (UTC)
I read every mention of collapse on the weblog, and no justification of the completed and reported experiment disputing collapse is made. Instead, promises are made for future work and a future experiment which would clearly show this. I think this is great, and I would love to read about it, but the fact remains that we shouldn't call things concluded on Wikipedia before the experiments which will clearly do so are actually completed and reported. It's the obligation of the article to document and reason out why said experiment says wavefunction collapse is incorrect. If it can't do that yet, then it shouldn't say that yet. — Cortonin | Talk 07:49, 3 Feb 2005 (UTC)

Weblog Entries

"Addendum: A brief response to Afshar's criticism can be found at http://www.physicsforums.com/archive/t-62460_Afshar_2_slit_experiment--peer_review.html (see rkastner's third posting)"

I moved the above entry by 70.21.61.81 to this page, simply beacuse I have not yet responded to Kastner on the main Wiki page. After I enter my rebuttal, we can have Kastner's response to my official rebuttal. Let's not replace Wiki entries with Weblog links. Thanks! --Afshar | Talk 04:28, 18 Mar 2005 (UTC)

A natural language restatement of Kastner's interpretation

The problem with asserting that this experiment rules out either the Many-worlds interpretation, or the Copenhagen Interpretation's Complementarity principle, is the assumption that, because with only one slit open the lenses and detectors provide welcher weg ("which way" or "which path") information, they will also do so with both slits open. This is the assumption of the principle of locality, and locality has been disproven (though perhaps not conclusively, according to some opinions) by Alain Aspect's implementation of the EPR experiment.

The photons remain in the entangled state until they encounter the detectors; at that point, they apparently make the welcher weg choice. However, this presumes that they had a path all along. This cannot have been the case, because otherwise they would have interacted with the wires. And in fact, in the one-slit-open case, they do interact with the wires, and detection at one detector does represent welcher weg information. But despite the apparently intuitive fact that when they choose one detector and not the other it must represent welcher weg information because of the measurement with only one slit open, it actually does not; this is due to a non-local interaction.

Consider the EPR experiment. Albert Einstein, Boris Podolsky, and Nathan Rosen believed that they had shown that either the particles must have had spins in particular axes all along, a violation of the Heisenberg Uncertainty Principle, or that the wavefunction collapse involved a violation of local realism at the time of measurement, and they maintained that the second was impossible. They were in fact correct in their first assertion, but they were incorrect in their second assertion. What they could not have foreseen was that John Stewart Bell would devise Bell's Theorem and show that the values could not have existed prior to the measurement, thus showing that non-local interaction actually takes place.

In the Afshar experiment, the non-local interaction is at the point of measurement, where the photon chooses one or the other detector. In this case, the non-local interaction causes the photon to manifest at one detector and not the other; this is called "wavefunction collapse" in the Copenhagen Interpretation. This is a common description in that interpretation of particle detection in general. The wavefunction propagates until a detection occurs. At that instant, a non-local interaction takes place, and the particle is detected at a particular location; this implies that the particle simultaneously cannot be anywhere else. If we include "the Andromeda Galaxy" in the definition of "anywhere else," this requires an instantaneous transmission of welcher weg information to the Andromeda Galaxy in violation of the absolute speed of light limit postulate of Special Relativity. 66.114.138.239

  • Moved this entry by 66.114.138.239 here because it contains major errors. i.g. photons are not in a entangled state, as there is only one photon at a time in the system. Locality is not the reason why we consider the images as provinding which-way information, it is the conservation laws. Too many errors too little time to discuss them...--Afshar | Talk 02:54, 08 Apr 2005 (UTC)

AFSHAR'S EXPLANATION ARISES FROM IGNORANCE

If one has time to study the introductional mathematics of Fourier optics s/he will immediately find out where Afshar's error is! Afshar simply ignores the fact that converging lens action is just to convert the spherical frontwaves into plane frontwaves. Actually the image at the Fourier plane is nothing but INTERFERENCE PATTERN. Afshar's error comes from improper comparison of fig.1 and fig.2 in his IRIMS paper. Well it would be obvious that fig.2 is interference pattern if he had provided the intensity distribution of the higher order maxima surrounding the two central Airy discs. The diference between fig.1 and fig.2 in Afshar's paper is that in the first case the Rayleigh criterion is not satisfied so there is only one central maximum, while in the fig.2. Afshar has splitted with the use of the lens the central maximum into two maxima.

For reference and detailed explanation one may read this paper: http://philsci-archive.pitt.edu/archive/00002281/

If one has time s/he coul update the Afshar's entry in the Wikipedia about his experiment. Simply the Wikipedia article is biased, the neutrality of the entry is questioned, and what is more - Afshar's reasoning is clearly based on ignorance.

  • I will not remove your rediculous comment, if only to keep the record of fringe statements, but:

Dear Danko Georgiev MD, You have certainly outdone yourself this time! I am sure the copying and pasting of ill-understood concepts and graphics from different sites must have been time consuming, but sadly it shows your complete incompetence in the topic, and it is quite frankly disgraceful. You have certainly overstepped your personal boundaries by statements like: "We suppose that Cramer's error does not result from ignorance of mathematics but it is due to psychological factors..." I would suggest you have yourself checked for BPD (I'm sure you know what that is Dr.), as you certainly seem to project your own shortcomings unto others!

Just like your telediagnosis, the technical merits of your argument is nonexistent (again due to the fact that you have no idea what you are talking about!) In page 13 of your non-paper you say the lens converts the spherical wavefronts into plane waves. That is true if we use a 1f-1f system as shown in Fig. 10. But in my experiment the lens is placed well beyond 1f from the dual pinholes, and thus produces images in the image plane. Another sign of your lack of understanding of even the basics of the subject matter is the fact that you naively ASSUME that the image plane is the SAME as the focal plane; it certainly IS NOT! If the distance between the lens and the object is more than f (which is the case in my experiment), the image plane falls at a distance always grater than f from the focal plane (which is the Fourier plane if the light illuminating the object is collimated), unless the object itself is placed at infinity, in which case the focal plane and the image plane become the same thing (due to the fact that light from an infinitely far source is a plane wave)...

This Blog is not an educational site (as outlined at the top of page), but since you seem to be fond of using internet resources, I would suggest you play with the applets in this link: http://webphysics.davidson.edu/applets/Optics/intro.html to get the hang of how a lens works in the geometric limit, and where the focal and image planes are.

To wit, the images in the IMAGE plane (read my paper carefully) provide highly reliable which-way information. [I can’t believe I wasted 15 precious minutes of my life on this nonsense!!!!]

P.S. I am normally not too critical, but you have certainly earned this frank counsel: Please stick to your own profession. There is nothing more embarrassing than a non-specialist making irrelevant remarks in a field well outside his domain.--Afshar | Talk 10:25, 28 Apr 2005 (UTC)


  • Dear prof. Afshar,

I will "close my eyes" for the fact you have personal attitude to my personality, not to my critique. Now straight to the topic:

You simply do not understand what the Fourier transform is, so I will clarify the issue using the exact values [data] from your experimental setup. I have been able to use the online calculator of the Rayleigh criterion at: http://webphysics.davidson.edu/mjb/SESAPS2000/rayleigh3.html in order to verify that your results are not manipulated. What is good is that you really have provided true results and I express my regards for providing all details about your setup in the IRIMS paper.

So now I will show you that I am very well acquainted with the action of the converging lens and that I do make distinction about the focal and image planes. However I didnot consider that the details of so great importance for my PhilSci paper - they are obvious.

Let me first list some relevant info from your setup: photon wavelength - lambda = 650 nm; distance between pinholes - d = 2 mm; pinhole (aperture) diameter - a = 0.25 mm.

1. Using the Rayleigh criterion you may easily see that the Airy disc of each pinhole "grows" along the optical axis. At distance up to ~0.5 meters the two Airy discs have not yet significantly interfered with light coming from the other pinhole.

2. You first put screen sigma_1 at distance L from the double-pinhole L = 4 m. So at this distance the double-slit pattern is clearly seen. Also at this place the image is essentially Fraunhofer diffraction image - i.e. the wavefronts can be approximated as plane ones - the usual criterion is d << L [well, the most accurate d^2/lamda << L is not satisfied, but actually how good is the appoximation to plane wavefronts is not essential at all for my argument].

3. So you can conclude that the image at sigma_1 is essentially a Fourier transform [F] of the two pinholes (with good approximation). Actually F implies that the image at sigma_1 screen is INTERFERENCE pattern, but you have said that already in your paper, so there is no dispute about that.

4. Now let us focus on your wrong conclusion about the case when you put lens. Yes, you put the lens with focal diameter f = 1 m at a distance L from the double-pinhole L= 4.2 m.

So one can easily find out that at its focus [L = 3.2 m from the double-pinhole] is located again interference image of the double pinhole [see my $1 in this reply]. So the lens observes a Fourier transform of the double-pinhole. It will act to implements inverse Fourier transform [F^-1] and will resolve the two pinholes at its focal plane located at L = 5.2 m from the double-pinhole.

Actually the inverse Fourier transform action of the converging lens can be understood as time reversal of the case described in my PhilSci paper, where is described 1f-1f scenario where double-slit image at one focus is transformed into F at the other focus. You may thus consider that the plane wavefronts observed at L = 3.2 m are converted into converging wavefronts in the region between 4.2 and 5.2 m from the double slit. After the focal plane the converging wavefronts are converted into diverging ones. You have in this second experiment put detector sigma_2 at L = 5.58 m from the double-pinhole. Actually you see that from the focal plane where the F . F^-1 transform produces again two pinhole image, there is a second round of "growing" of the Airy discs. For the distance of only 0.38 m from the focal plane of the lens till the final detector sigma_2 the two Airy discs have not yet "fused" into a single maximum.

You should see that Fourier optics explains your setup step by step, so you should be aware that the applied F and F^-1 before the final detection at sigma_2 have fully erased the "which way" info.

If you try to object to this last claim, then you should prove that there is reliable way to extract "which way" info from Fourier transformed image [understood here the classical interference image that you have obtained with the sigma_1 detector]. Whether F^-1 restores the double pinhole-image is not important, it acts on F, and that is why the "which way" info cannot be restored.

I will soon upload a second more technical paper with step-by-step computations [simulations] on Wolfram's Mathematica.

p.s. the fact that you have adopized the wave functions [i.e. "you have erased the outer interference maxima surrounding the central Airy discs"] just will make F^-1 with worse quality [i.e. worser reconstruction of the two pinhole image], but does not change anything of my argument.

You may easily see that I have studied carefully your paper. I hope you will finally realize that my professional interests in medicine are based on careful study of physics. Actually I am working on biophysical modelling, where a detailed usage of geometry is needed [as well as Wolfram's Mathematica programming to generate the final outputs].

I hope you will be able now to realize where your error is, and if so I will not pretend for the 1000$ award announced by you for disproving your interpretation :-)

Cheers,

Danko

Reply to Unruh question

Here is the question posed by Unruh. He asks "Yes, the lens acts as a "Fourier transform" device, but again so what?"

From: Bill Unruh [1] Sent: Friday, April 29, 2005 1:33 PM To: Danko Georgiev

Again, your interpretation I agree makes no sense. The "which way" infomation IS there in the images, just as it is there in the original pinholes. I would certainly not advise him to give you the $1000 based on this argument. While he is wrong, he is not wrong in the way you have outlined it.

The fact that IF you were to place a screen before the lens, you would see an interference pattern, is irrelevant to whether or not the photons falling on the detectors convey "which way " information.

Yes there is a Airy disk. but so what. Yes, the lens acts as a "Fourier transform" device, but again so what. The natural propagation of the light also acts as a fourier transform. And the lens can be set up to compensage for the natural propagation transform to give you back the original image.


Bill Unruh

My reply is:

Dear Bill,

You simply don't understand the essence of my argument.

The existence of light "wave" implies that there is no "which way".

What I have shown is that application of F . F^-1 returns the SAME image at the focal lens plane as one can expect from "classical straight ray" optics.

So IF you cannot decide whether you have F . F^-1 or "clasical straight ray image" THEN THERE IS NO WHICH WAY INFO!

Also I can show you that indeed F. F^-1 existence NEEDS "wave interference" and also I CAN PROVE that the actual implication of F . F^-1 is seen in Afshar experiment, because the image plane is 0.38 m behind the focal plane, so I expect higher order diffraction maxima, while the "classical scenario" CANNOT predict these outer maxima, but just expects enlargement of the central Airy discs.

Cheers, Danko

  • [Bill Unruh's response, from e-mail CC/ed to afshar@rowan.edu by Bill]

Since you decided to post an answer to a private reply to you to this whole list, I will respond in the same way.

a) EM waves are linear. Thus the image at the detectors is precisely the sum of the images from the waves emitted at the two slits. Those independent images are that if the wave came from pinhole 1, the image (yes, with diffraction pattern) falls entirely on detector 1, and the image from pinehole 2 falls entirely on detector 2. This is an elementary result of fourier optics. Yes, there is interference in the intermediate region, interference which comes from precisely the linearity of the waves, but that is irrelevant.

On the basis of your fourier optics, the Green's function of the EM field is precisely such that the image at detector 1 depends only and solely on the amplitude at slit 1, and that at detector 2 depends only and solely on that at slit 2.

b) What has "classical straight ray" arguments got to do with anything. Who discusses them but you? The argument is not based on "classical straight ray" arguments. It is based on the wave nature of light throughout.

You may be getting confused when the term "which way" is used. It is NOT which way the photon went between the slit and the detector. That is NOT what "which way" refers to. It refers to whether or not the photon came through slit 1 or slit 2. That is all. It has nothing to do with what the photon did in the space between the slits and the detectors. You may have gotten misled by the use of the word "way" in "which way" and thought that it referred to a complete path. It does not. In this context it would better be phrased as "which slit". It is ONLY which slit the light came out of that is being determined, not anything about how it travelled between the slit and the detector.

    • Dear All,

Above is an exchange between Bill Unruh (a veteran physicist) and Danko Georgiev (AKA The Fake) in which Bill rejects his utter nonsense. Although Bill disagrees with me, I applaud his defense of truth, which in this case happens to be the defense of my arguments regarding which-way info in the images. The likes of Georgiev MUST be rejected and marginalized by all real scientists in the interest of scientific advancement, all personal differences aside. I have no problem with considering unorthodox ideas (after all some may consider me as unorthodox,) but after one realizes an individual is fake, and it takes no more than a couple of exchanges for a pro., the individual must be warned, and if stubborn, BANNED from professional forums. Therefore, I shall make recommendations to PhilSci and arXiv archives, as well as Wikipedia to put his name on their blacklist. I would also encourage you to do so. Sorry Danko, but you brought this unto yourself after my repeated warnings...--Afshar | Talk 15:46, 30 Apr 2005 (UTC)

Retrieved from "https://en.wikipedia.org/w/index.php?title=Talk:Afshar_experiment&oldid=13485611"