Lunar Surface hologram

[Martin]

The problem I'm having is: Did Denisyuk think he was making a hologram of a mirror or a holographic mirror. In one case, it's a display hologram, in the other, it's a HOE.

Maybe, at that moment of time making a distinction between "HOE" and "display" simply made no sense. How many display holograms had been recorded at that moment?

But, the paper doesn't make it clear since it talks of a "scattered field". However, scattering implies diffuse scattering from a diffuse object, reflection of light from a mirror (curved or not) is not "scattered radiation". If you tried the geometry in the paper, with a series of curved mirrors, I think you would get pure reflection and so a holographic mirror and no scattering.

Johnston (p.71) points out: "Strictly speaking, the scatter in Denisyuk's implementation is more properly described as specular or mirror-like reflection. He had not attempted to record diffusely reflecting objects."
Is that so? Denisyuk seems to have used a filtered mercury vapor lamp only. So the coherence length wasn't quite large.

But, if you coated the mirror with a fine coating of dust, then the dust would scatter and you'd have a hologram of a mirror.

Right. That really opens a Pandora's box. E.g. if there had been some dust particles on the reflector (not particularly unlikely), do those Lippmann photos morph into reflection holograms (are there "Denisyuks" before Denisyuk)?

Perhaps, there could be a little uncertainty in the translation from the Russian?

Yes, that would be interesting to know.

The interesting thing is that Denisyuk himself was not consistent. When I asked him how he thought of the technique, he said that he had come back from vacation and saw a paper on Lippman photography. As he read the paper, he realised that the author did not completely understand Lippman photography. As Denisyuk was thinking of a response, it struck him that he could use the principle of Lippman photography to create white light reconstruction in holography. To place the conversation in context, I was talking to him about how strange it was that the laws of physics are so simple and he answered, "Yes, simple, but so difficult to prove". However Hans Bjelkhagen has stated that he (Hans) thinks it was because of a book that Yuri had read. Jeff Weil has told me that Yuri gave him yet another explanation of the origin of the Denisyuk technique.

Johnston writes:
"The inspiration for Denisyuk's ideas is unclear. (...) He recalls having been inspired for his thesis work by the writings of the Russian science fiction author Ivan Antonovich Yefremov."

[Martin]

... I apologize for derailing the thread into semantics.

I don't think it's just semantics. There are real problems there.

[Dinesh]

Maybe, at that moment of time making a distinction between "HOE" and "display" simply made no sense. How many display holograms had been recorded at that moment?

Johnston (p.71) points out: "Strictly speaking, the scatter in Denisyuk's implementation is more properly described as specular or mirror-like reflection. He had not attempted to record diffusely reflecting objects."
Is that so? Denisyuk seems to have used a filtered mercury vapor lamp only. So the coherence length wasn't quite large.

Exactly! Ever since I reads that paper, I always believed that his paper describes the recording of a specular reflection and not diffuse scattering. Denisyuk's later writings implies that he realised that an image of diffuse objects were possible, but his initial work was done with a specularly reflecting "object beam", despite the title of the paper. From a practical point of view, it may just be semantics today, but as a historical fact, it is curious. I think he used a specular reflection as an object beam because it was simple to do so. In order to prove such a far-reaching hypothesis, it is important to remove all superfluous phenomena. It could be argued that a plane mirror would be even simpler, but I think that a plane mirror could not have been distinguished from specular reflection of the plate itself. A diverging mirror, would have given definite proof that the principal worked. In the modern sense, he had actually made a HOE, not a "Denisyuk", but at that time, he was trying to prove a principle, rather than create a nomenclature.

One other possibility exists. The crux of the hypothesis is that there needs to be standing waves within the medium. However, in order to create standing waves, the beams in both directions need to be of equal magnitude. If they are not, there is a travelling wave component, the planes would move slowly, depending on the beam ratio. Depending on the sensitivity of his plates, any such motion would result in loss of efficiency. Therefore, using specular reflection would have been fairly important to ensure that the beam amplitudes were equal.

Right. That really opens a Pandora's box. E.g. if there had been some dust particles on the reflector (not particularly unlikely), do those Lippmann photos morph into reflection holograms (are there "Denisyuks" before Denisyuk)?

I very much doubt they would have been recorded holographically. The reflection from dust particles would have a very low amplitude due to weak scattering, especially as the Fourier component in the forward direction would have been the only important component. This weak forward reflection component from the scattering of the dust particles would have caused a travelling component to the planes which may not have been recorded, as the turn-of-the-century emulsions were not particularly high density. Has anyone discovered what the spatial frequency of Lippman's plates were?

Johnston writes:
"The inspiration for Denisyuk's ideas is unclear. (...) He recalls having been inspired for his thesis work by the writings of the Russian science fiction author Ivan Antonovich Yefremov."

I have heard this before. However, I am wondering if this referred to his thesis work for his PhD thesis, rather than the "scattering" paper. By 1962, Denisyuk was in his mid-thirties, so this was not his PhD thesis. I rather tend to accept his explanation of trying to form a reply to the paper he saw on his return from vacation. Scientific breakthroughs are usually caused by a response to a former work, but great breakthroughs are always associated in some romantic way with some anecdote - Newton and the apple, Einstein seeing someone falling off while trying to clean a clock etc. I remember that my professor, Abdus Salaam (sorry, name dropping, but there is a point!) was asked to write about his electro-weak theory in "New Scientist". I read the article and told Salaam that I thought it was a pretty good attempt at popular writing for what was such an abtuse work of quantum field theory. He mentioned that the "New Scientists" editors had asked if he could "dumb it down" a little and weave some sort of inspirational story around it!

[BobH]

One other possibility exists. The crux of the hypothesis is that there needs to be standing waves within the medium. However, in order to create standing waves, the beams in both directions need to be of equal magnitude. If they are not, there is a travelling wave component, the planes would move slowly, depending on the beam ratio. Depending on the sensitivity of his plates, any such motion would result in loss of efficiency. Therefore, using specular reflection would have been fairly important to ensure that the beam amplitudes were equal.

This is the first I've heard of this. Fringe motion as a function of beam ratio?!? Can you cite some references?

[Martin]

if there had been some dust particles on the reflector (not particularly unlikely), do those Lippmann photos morph into reflection holograms (are there "Denisyuks" before Denisyuk)?

I very much doubt they would have been recorded holographically. The reflection from dust particles would have a very low amplitude due to weak scattering,

So having a reflection hologram recorded in that case would have depended on:
the size of the dust particles, recording wavelength and the index modulation of the recording medium, right?

As an aside, I lately became aware that over the last 20 years or so there's been plenty of research on short-coherence holography. And that subject sometimes seems to converge with the early Denisyuk/Lippmann photography.

especially as the Fourier component in the forward direction would have been the only important component. This weak forward reflection component from the scattering of the dust particles would have caused a travelling component to the planes which may not have been recorded, as the turn-of-the-century emulsions were not particularly high density.

You mean AgX concentration?
By the way, not just AgX but dichromated and ferric recording media had been used as well.

Has anyone discovered what the spatial frequency of Lippman's plates were?

There's been some research by Fournier who had access to some tiny bits of AgX emulsion dissected from an original Lippmann photo (recorded by Lippmann himself). I think they easily reached 7500 lines/mm in the violet parts.

[Ed Wesly]

To get the facts straight, read Denisyuk’s own explanation in Leonardo Volume 25, No. 5, pp. 425--430 (1992), “My Way in Holography”.

For those who don’t have a copy, he does indeed relate being inspired by the story “Star Ships” by Yu. Efremov (sic, as compared to Johnston.) He started work in 1958, collimated Hg-vapor light source, initial test experiments recording convex mirrors with radii of curvature from 2000 to 300 mm to accommodate the short coherence. He laments that he should have used a coin as the object instead!

[Dinesh]

This is the first I've heard of this. Fringe motion as a function of beam ratio?!? Can you cite some references?

It's difficult to know where to start, I've got a lot of flack from display/art holographers when trying to explain that a hologram is a phase capturing system. The phase relationships inside the medium are not some kind of comb or shah function, as is popularly depicted in most books.

If you take two beams that are not co-linear, thus a reflection or transmission holographic geometry, then, so long as there is mutual coherence, the phase difference between the two beams is k*dx, where k is the wave vector and dx is the difference in distance. This is true so long as dx < [(lambda)^2]/d(lambda), where d(lambda) is the wavelength spread in the laser. At any rate, the intensity at any point is given by cos(k*dx) (omitting constant terms), creating a sinusoidal variation fixed in space. This is what allows the recording of fringes or Bragg planes in both reflection holograms and transmission holograms and this is not a Lippman system. Thus any reflection hologram where the object is on-axis, but the reference is at some angle theta, records planes of iso-phase, not the nodes and anti-nodes of a Lippman system.

In the Lippmn system, there is a mirror at the back of the medium. This ensures that the incident waves and the reflected (retroreflected?) waves are of equal magnitude. In such a situation, if the incoming waves are E(01)*cos(kx+wt) and E(02)*cos(kx-wt), then the disturbance inside the medium is:

E = E(01)*cos(kx+wt) + E(02)*cos(kx-wt)

Where E(01) and E(02) are the amplitudes of the two counter propagating waves. If these are equal, ie if E(01) = E(02), then the waves within the medium are

E = E(01){ cos(kx+wt) + cos(kx-wt) = E(01)cos(kx)*cos(wt)

This creates nodes where the total disturbance is zero, and anti-nodes where the total disturbance is the E(01) and positions in between where the amplitude is some fraction of E(01). In other words, the classic standing wave. This is the principle of Lippman photography.

However, when E(01) does not equal E(02), then you do not get pure nodes or anti-nodes. The maximum amplitude is E(01) + E(02) and the minimum amplitude is E(01) - E(02). This difference in amplitude causes the phasor of the sum to rotate as an ellipse and creates a slowly moving traveling companion.

In the description of a Denisyuk, the planes are supposed to form in the same way as a Luippman structure. This would be true if the two beams are counter propagating, ie the ref and object are co-linear. However, if the beams do not have the same amplitude, ie the ratio is not 1:1, then this traveling wave component results, causing both a node and anti-node set of planes, which record in the emulsion, and a traveling wave companion, which does not. This traveling wave companion will cause a degradation of the fringe modulation. So, If Denisyuk wanted to exactly copy the Lippman technique with two counter-propagating beams in line with each other, ie on axis, then if the two beams did not have equal amplitudes, the traveling component would cause a loss of efficiency.

But, and here's the important thing, a Denisyuk made in the "standard" off-axis geometry is NOT a Lippman photograph. A Denisyuk captures phase information, which is a complex set of convoluted two-dimensional curves (by the way, I got a lot of flak when I mentioned that a Denisyuk is not a Lippman because a lot of art/display holographers seem to think that it is with almost religious fervour!). A Lippman does not capture phase information, it "captures" the image in a set of nodes and anti-nodes. The mirror in the back ensures E(01) = E(02). In a Lippman "hologram" the planes will be parallel to the face of the medium, separated by lambda/2. However, they are still described by sin(kx), not a shah function. A curious fact is that all the papers on reflection holograms all state that the efficiency of the hologram is a tanh squared function. However, in the Kogelnok paper, it's clearly stated that the tanh squared function is the efficiency of unslanted planes with c(s)=0, ie a Lippman geometry but with E(01 not equal to E(02). In this situation, Kogelnik does not take into account the traveling wave component, which would decrease the efficiency. However, a display hologram cannot have unslaned planes. The actual efficiency of a display reflection hologram is a complicated sinh squared function. I don't know where this error started, the first record I have of this is in Rallison's papers, but everyone else simply copied the previous paper's efficiency term!

There is, however, a further subtlety. If the object of a Denisyuk hologram propagates light along the direction of the reference beam, ie the ref comes in at say 60 and the object throws off light at 60, then you do have two counter-propagating waves. If now these counter propagating waves are not of equal amplitude, then a traveling wave component comes in, but only in the direction of the reference beam.

The important thing is to recognise the difference between a standing wave and interference

[Dinesh]

Here are phasor diagrams of standing waves when E(01) does and does not not equal E(02). It's from Hecht's "Optics" 4th edition.

[Dinesh]

So having a reflection hologram recorded in that case would have depended on:
the size of the dust particles, recording wavelength and the index modulation of the recording medium, right?

Yes

You mean AgX concentration?

Sort of. In the direct line of the Lippman the return beam from the dust particles would be very, very low. In this situation, there would be a weak interaction. If the emulsion were sufficiently slow, it might be captured. In the off-axis scattering, it would depend on the spatial frequency response of the medium, which would, in turn, depend on the size and concentration of the AgX grains.

[Sergio]

"A Lippman does not capture phase information, it "captures" the image in a set of nodes and anti-nodes."

So a Lippmann photograph will never "offer" any multidimensional image.

But in the case of E(01) does not equal E(02), in a Lippmann photograph that is recorded without a mirror but using only the refraction index difference from the gelatin medium and the air interface you may get E(01) not equal E(02)? So you will get in this case difference in amplitude that causes the phasor of the sum to rotate?

I may conclude that the the travelling component in Lippmann photographs is one source of image degradation and is not related to the phase information (that is not related to Lippmann photos)?