Talk:Polarization (waves)/Archive 1

merge
This is three articles banged together -- can someone copyedit this?


 * Done. Excised text:

Polarization in telecommunications: Of an electromagnetic wave, the property that describes the orientation, i.e., time-varying direction and amplitude, of the electric field vector.

Note 1: States of polarization are described in terms of the figures traced as a function of time by the projection of the extremity of a representation of the electric vector onto a fixed plane in space, which plane is perpendicular to the direction of propagation. In general, the figure, i.e., polarization, is elliptical and is traced in a clockwise or counterclockwise sense, as viewed in the direction of propagation. If the major and minor axes of the ellipse are equal, the polarization is said to be circular. If the minor axis of the ellipse is zero, the polarization is said to be linear. Rotation of the electric vector in a clockwise sense is designated right-hand polarization, and rotation in a counterclockwise sense is designated left-hand polarization.

Note 2: Mathematically, an elliptically polarized wave may be described as the vector sum of two waves of equal wavelength but unequal amplitude, and in quadrature (having their respective electric vectors at right angles and &pi;/2 radians out of phase).

Source: from Federal Standard 1037C and from MIL-STD-188

-

From the article:
 * Individual photons are inherently circularly polarized; this is related to the concept of spin in particle physics.

Can someone fact-check this?


 * I think I just answered my own question: http://cse.unl.edu/~reyes/CPE.html

I have made a major overhaul of this entry, because it was somewhat incoherent and repetitive (presumably due to merges of several sources), missed some important points, and probably left a lot of people scratching their heads trying to visualize things without any diagrams. I hope my attempt is an improvement. I have tried to retain material from the previous version, or adapt it somewhat to fit in better. Some parts I omitted completely because they seem too specialized or they probably belong in other entries. The completely removed text appears below. Possibly some of the stuff I have added should also be ripped out and put into other more specific entries but it will take a little thinking over as to what is the best way to do that without reducing the article to a series of facts and links presented without explanation or continuity... so for now I've just put it all in here. Hacked out text:

For circular polarization, it is also useful to consider how the direction of the electric vector varies along the direction of propagation at a single instant of time. While in the plane the vector rotates in a circle (as time advances), along the propagation axis (at one instant) the tip of the electric vector describes a helix. The pitch of the helix is one wavelength, and the helix screw sense is either right handed or left handed. Visualizing this spatial variation in the direction of the electric field is useful in understanding how circularly polarized light can interact differently with helical molecular conformations, depending on whether the electric field and the molecule helix sense are the same or opposite. This is part of the phenomenon of circular dichroism.

[...]

As described by Maxwell's equations, light is a transverse wave made up of an interacting electric field E and a magnetic field B. The oscillations of these two interacting fields cause the fields to self-propagate in a certain direction, at the speed of light. In most cases, the directions of the electric field, the magnetic field, and the direction of propagation of the light are all mutually perpendicular. That is, both the E and B fields oscillate in a direction at right angles to the direction that the light is moving, and also at right angles to each other.

(In optics, it is usual to define the polarization in terms of the direction of the electric field, and disregard the magnetic field since it is almost always perpendicular to the electric field.)

[...]

A quarter-wave plate is constructed from a birefringent material, that is, in the plane of the plate there are two orthogonal axes and light passing through it propagates at a different speed along one axis than on the other. The thickness of the plate is adjusted so that the net difference in propagation speed is one quarter of a wavelength. If this plate is oriented so that the fast axis is forty five degrees to the direction of linear polarization then the light emerging from the other side will have two components of equal amplitude and a ninety degree phase difference, creating circular polarization. Rotating the quarter wave plate ninety degrees in the plane will reverse the sense of circular polarization.

Birefringence can be created by straining a normally uniform material. A properly arranged and controlled mechanical oscillator coupled to a strain-free window can convert linearly polarized light of a single color impinging on the window into alternating left and right hand circularly polarized light emerging from the other side. That is, the window can operate as an oscillating quarter wave plate. If this light is then passed through a material which has a circular dichroism at that color, the emerging light will have an amplitude modulation that varies with the frequency of the oscillator driving the quarter wave plate. This amplitude variation can be detected and used to measure the amount of circular dichroism exhibited. This amplitude will depend on the intrinsic property of the material, and upon the amount of material the light passed through, which in turn depends on the concentration of the absorbing substance and its thickness. Although the phenomenon measured this way is delta-absorption, the results are customarily reported in degrees of ellipticity through a simple algebraic conversion.


 * Theorie mathematique de la lumiere, Henri Poincar&eacute;, Gauthiers-Villars, Paris, 1892. The original description of the Poincar&eacute; Sphere.

Rkundalini 15:11, 2 Jun 2004 (UTC)

Polarization in elastic waves
Since elastic waves may have transverse components, they may be polarized. They also exhibit many of the properties of electromagnetic waves (e.g. birefringence, aka "shear-wave splitting"). But I don't know offhand how to incorporate that into this article. It's definitely related, but the structure of this article would make it hard to add. Gwimpey 06:02, Mar 5, 2005 (UTC)


 * Are they mathematically equivalent to electromagnetic waves? If not I'd suggest a separate article, which refers to this one for concepts that are related. Something similar should probably done for gravitational wave polarization and any other types of waves with transverse components. -- Rkundalini 06:37, 10 Mar 2005 (UTC)

Observing polarization effects in everyday life
"All light which reflects off a flat surface is at least partially polarized."

I do not have any knowledge of the physical principles involved, but from my photographing days I seem to remember that a polarizing filter has a dramatic effect in suppressing light reflected from water or polished non-metallic surfaces, while the effect on light reflected from metals seems insignificant. I have used this effect for taking pictures from a mirror - a polarizing filter removes doubled contours caused by reflection in the glass, while the image reflected from the silver layer remains clear. Can someone knowledgeable explain? --Georgius 16:55, 6 Jun 2005 (UTC)

Yes, except at normal incidence, reflections from metal are a little bit polarized, but maybe not enough to be included in the article. Metals tend to reflect both polarization well, except for oxides and surface irregularities. Dielectrics have a Brewster angle at which vertically (for a horizontal surface) polarized light is completely transmitted into the medium. A glance at that article looks as though the Brewster angle would be complex, and therefore not correspond to any plane wave, for a lossy material like a metal.

I think "circular polarizer on the camera" in the figure caption should read "linear" or perhaps vertical, but I am not entirely sure. If no-one is sure, maybe we should just delete the word "circular". I know that the sky tends to be linearly polarized and that my camera filter is linearly polarized, like my sun glasses, but I can't entirely rule out the possibility that the polarization could become elliptical somehow.
 * Circular polarizers used in photography start with a linear polarizer and then add a second birefringent layer to create circularly polarized light from the linear polarized light. This is done because many modern cameras have beam splitters (for focusing and metering) that don't work with linear polarized light. -Steve Pucci | talk 03:43, 16 May 2006 (UTC)

The blue sky is polarized because the scatterers are electric dipoles that are polarized (the charge is displaced) perpendicular the direction of the incident light. This is called Rayleigh scattering.

Stressed materials such as eyeglasses where they are held by the frames and tempered rear windows of cars rotate polarized light by transmitting the components differently, so one can see a pattern when looking with polarized glasses through the rear window at a reflection from a windshield, shiny paint or asphalt. David R. Ingham 22:49, 9 March 2006 (UTC)

imax passive polarized 3d glasses
Can someone explain how the imax passive polarized 3d glasses work here?

They are not like the old red-blue paper 3d glasses that give me a headache. They don't give me a headache at all and make things 3-d. Since they are called "passive polarized" I would think they should be listed here, or at least linked to here. I cannot seem to find them anywhere else.


 * They're probably simply two orthoganally polarized lenses, with the movie being projected onto the screen twice, each channel being polarized so it's viewed by one eye. See stereoscopy and polarized glasses for more. --Bob Mellish 20:25, 11 November 2005 (UTC)

The screen has to be made with glass beads, not white material like paper or white paint. David R. Ingham 23:01, 9 March 2006 (UTC)


 * That's one way of doing it. The RealD system (Beowulf, etc, when in digital rather than Imax projection) uses two lenses of left and right hand circular polarisation. This is somewhat better as you don't have to maintain the glasses horizontal with the screen. Confusingly, the dichroic coatings on these glasses look red and green when viewed at about a 45-degree angle, leading people to incorrectly assume it's an anaglyptic method. —Preceding unsigned comment added by 68.123.239.154 (talk) 12:21, November 23, 2007

unclear (transverse polarization)
I think this part of the article is confusing: "A plane wave is one where the direction of the magnetic and electric fields are confined to a plane perpendicular to the propagation direction."

I dug up a bit and found this on a previous version of the article: "In optics, it is usual to define the polarization in terms of the direction of the electric field, and disregard the magnetic field since it is almost always perpendicular to the electric field.". And in every physics book they only seem to take in account the electric field when describing polarization.

So, is the current version saying the same thing? I think it needs to either be reworded or further explained, because it can be confusing to people with few knowledge of electromagnetic waves.

nehalem 11:22, 15 January 2006 (UTC)

The sentence you quoted is not saying the same thing. It is defining what a "plane wave" is. Plane waves are usually assumed in elementary descriptions of polarization. Essentially a plane wave is a uniform light wave travelling in a single direction. All the wavefronts of the light are flat, and the electric and magnetic fields are therefore perpendicular both to each other and to the direction of propagation.

When one considers polarization of a plane wave, it's convenient to just talk about what the electric field does, since for a plane wave the magnetic field is always perpendicular to the electric field, and is proportional to it. There is nothing fundamental in this&mdash;polarization is as much a magnetic field effect as an electric field one. It just makes for a simpler description to only deal with one of the fields. --Srleffler 15:02, 15 January 2006 (UTC)


 * $$\vec H \perp \vec E \perp \vec k$$ should hold for all waves, not just for plane waves. From the Maxwell equations we have $$\vec H \perp \vec D \perp \vec k$$ and in isotropic media $$\vec D // \vec E$$. I therefore see a point in nehalem`s comment and propose to change the article`s definition of plane waves from "A plane wave is one where the directions of the magnetic and electric fields are perpendicular both to each other and to the propagation direction." to "A plane wave propagates everywhere in the same direction, and like all electromagnetic waves has the electric and magnetic fields perpendicular to the propagation direction." I will change it if I don`t hear objections. --danh 02:33, 16 January 2006 (UTC)

It is not true that Maxwell equations require that $$\vec H \perp \vec D \perp \vec k$$. Only H must be perperndicular to D. For instance guided modes of a slab waveguide have longitudinal components (TE modes have a longitudal component for H, and TM modes have a longitudinal component for E). As Gnixon indicated on this page also polarizations in Fibers have longitudinal modes. The article intorduction is simply wrong to define polarization as perpendicular to the direction of propogation. Electric field Polarization (Usually simply called polarization, as via Maxwell equations it uniquely set the magnetic field polarization) is simply the direction of the electric field vector in space and time. All the polarizations of a system can be spanned by a basis of just two vectors because of the so called "transverse condition" of maxwell equations which puts a 1 dimensional constraint on three dimensional space. The name is misleading as the polarization is not always transverse to the direction of propogation.Eranus 14:36, 30 November 2007 (UTC)


 * I agree with you and Gnixon. --Danh 23:26, 2 December 2007 (UTC)

I agree that that was an improvement. To be even more picky than usual, I could suggest "propagating electromagnetic waves", because evanescent fields may be called waves but have no real direction of propagation. An example is the field on the outside of the dielectric, in total internal reflection. David R. Ingham 23:15, 9 March 2006 (UTC)


 * I agree with danh's edit of the sentence, except that E&M waves may generally have longitudinal components. An example is E&M fields confined to circular waveguide, for which either the electric or magnetic field may be longitudinal.  (The description of polarization in this case is virtually identical to that of free space E&M waves.)  The property of E and H being perpendicular to the direction of propagation holds for (infinite) free space or a uniform material.  I'll make a slight change to the sentence.  --Gnixon 17:15, 24 July 2006 (UTC)

The important point about a plane wave is that a plane is two-dimensional. Light is a transverse wave, the wave fields are orthogonal to the direction of propagation, but in general they have both x and y components (if propagation is in the z direction). Most descriptions, including the transverse wave article, talk as if light can be understood by exact analogy to a water wave. However, a water wave is transverse in only one dimension. Phenomena such as circular polarization can only arise when the transverse wave is two-dimensional.

AJim (talk) 03:23, 9 July 2008 (UTC)


 * Your comment about light compared to water is worthy. You, and the current article version (See former comments) are wrong in stating that polarization must be perpendicular to the direction of propagation. Polarization is a two dimentional space because of the transverse condition (One of Maxwell equations)


 * $$\nabla \cdot \mathbf{H} = 0$$


 * which is a 1D constraint on 3D space. This 2D space is not a necceseraly a plane. It is the plane perpendicular to propogation for plane waves, it is not for guided modes.Eranus (talk) 12:26, 23 July 2008 (UTC)


 * I'm familiar with the fact that guided electromagnetic waves can have longitudinal components. I think it is actually true that this cannot occur in free space or a homogeneous medium of infinite extent, but I'm relying on dim memory from a long time ago and don't have a reference handy. If true, this is an important fact to include. Plane waves don't actually exist. If we can say something about general waves in free space, we should.


 * Your first change to the intro is not technically correct, because the intro defines polarization (as discussed here) as a property of transverse waves. That definition does not limit the discussion to EM waves, but it does exclude guided modes with a longitudinal component. I'm not sure how best to deal with this. --Srleffler (talk) 04:21, 24 July 2008 (UTC)
 * I took a stab at reworking the intro. For transverse waves, it is true that polarization describes the orientation of oscillations in the plane transverse to the propagation direction. The intro now mentions that there exist waves that are neither transverse nor longitudinal, and that these have polarization too. The wording could be improved.--Srleffler (talk) 04:52, 24 July 2008 (UTC)


 * OK this is a matter of definition, I was willing to accept that all E&M waves are transverse becasue they obey one of Maxwell equations also called the transverality condition:


 * $$\nabla \cdot \mathbf{H} = 0$$


 * I always thought that this is a bad name because the condition does not require generally that the polarization be perpendicular to the propagation. So my thiking was to continue to call EM wave transverse, because :$$\nabla \cdot \mathbf{H} = 0$$ is an important quality of all E&M waves which means that the vectorial nature of light add only 2 dimentions rather than three. Maybe you are right that we should stick to the literal meaning of transverse and simply not call EM waves transverse in general. Then it is important to clearly state the 2D vectorial nature, somehow. I don't know how to write this well because this an issue of history, I believe initialy people thought that all EM waves are literaly transverse and the name stuck. I'm not sure how to write this well, as a first stage we must keep it correct.


 * I'm very curious about whether all bulk or free spacee modes are transverse in the way you mean, asyou say it is an important fact to state. Specifically I remember some journal club lecture of radially polarized light being purely longitudenal at the waist (focus). I think in general higher order Laguerre Gaussian Modes have longitudenal components, but I'm not sure. Anyway this should be better checked and stated, I'm glad you did not reinset the statement until we resolve this. —Preceding unsigned comment added by Eranus (talk • contribs) 08:07, 24 July 2008 (UTC)

We are rightfully trying to make polarization simple and correct in the intro but it is not so simple in reality.Eranus (talk) 08:48, 15 May 2009 (UTC)
 * I double checked and indeed realistic beams measured in experiments can have longitudinal components even in free space with no sources. This is true for instance for radially polarized beams near the focus where the longitudinal componenets can be larger than the transverse components.

Combining waves
Disclaimer: I am not a physicist.

I was just reading that you can take any light source, split it into two orthogonally-polarized beams, and combine them again to get the original source. This is news to me (I would have thought that you were putting in an infinite variation of polarizations, and only getting two out), but it makes sense. In the article, it says: If I am understanding correctly, would this be better worded: In other words, they're just like complex sinusoids, where $$e^{ix} = \cos x + i\sin x \!$$ and $$\cos x = {e^{ix} + e^{-ix} \over 2}$$ are two different valid ways of thinking about things. (A sinusoid is a sum of helices or a helix is a sum of sinusoids.) — Omegatron 02:20, 18 July 2006 (UTC)
 * Plane waves of any polarization can be described instead by combining waves of opposite circular polarization, for example.
 * Plane waves of any polarization can be described by combining waves of opposite circular polarization, for example, or by combining waves of linear polarization rotated 90&deg; from each other.
 * That would not be better wording in the context, but it is true. You can decompose a beam of light of any polarization into any two polarization components that form a basis. There are many possible choices, of which perpendicularly polarized linear polarizations and counter-rotating circular polarizations are the most obvious. Your version is not better wording only because the preceding portion of the article describes light in terms of perpendicular linearly-polarized components, and the point of that statement is to indicate that this was an arbitrary choice, and that one can instead consider a pair of circularly-polarized components.--Srleffler 05:48, 18 July 2006 (UTC)
 * Oh. :-)  The article is worded a little poorly.  Overly technical.  — Omegatron 05:58, 18 July 2006 (UTC)
 * I thought it was very clear where it said: The "cartesian" decomposition of the electric field into x and y components is, of course, arbitrary. Plane waves of any polarization can be described instead by combining waves of opposite circular polarization, for example. It's a technical subject, so some skill in reading logical technical writing may be assumed. Dicklyon 06:40, 18 July 2006 (UTC)

Mud flats
The only water in the picture is a trickle at the bottom of the river channel, and the sea (which is the dark brown patch just about visible at the top right hand side, above the bank of seaweed and below the headland). The tide was very far out (I think I'm right in saying it was spring tide, and the Severn estuary has one of the highest tidal ranges anywhere in the world), and the reflections are sunlight off the wet mud. Originals are at image:Mudflats-polariser-1.jpg and image:Mudflats-polariser-2.jpg. --ajn (talk) 20:37, 18 July 2006 (UTC)


 * OK, I yield that point. I still think it makes more sense to talk about reflection off water however, since that's the relevant ingredient of mud that is reflecting. Dicklyon 20:43, 18 July 2006 (UTC)

Just to make things clear, "In the first picture, the polarizer is rotated to minimize the effect; in the second it is rotated 90° to maximize it: almost all reflected sunlight is eliminated". "First", "Second"? How about left and right? It is a little confusing...
 * I hope now it is clearer. --danh 00:02, 12 January 2007 (UTC)

eccentricity/ellipticity
I don't understand the preference for ellipticity over eccentricity in the description of the polarization ellipse. In terms of the Stokes parameters I, Q, U, V, with Ip=sqrt(Q^2+U^2+V^2), we have ellipticity = V/Ip and eccentricity = sqrt(Q^2+U^2)/Ip, so ellipticity represents the degree of circular polarization (0 for linear, 1 for circular), while eccentricity represents the degree of linear polarization (0 for circular, 1 for linear). (See, e.g., Stokes parameters.) As far as I can see, the physical interpretations of the two parameters are complementary. Is there some other motivation for preferring ellipticity?

[Sorry, those expressions are wrong. See below. Gnixon 14:26, 22 August 2006 (UTC)]

I'll hold off on an edit for awhile in case someone sets me straight.

Gnixon 21:35, 23 July 2006 (UTC)
 * I don't know the answer to your question, but policy forbids you from changing it on the grounds you describe. Per No original research, Wikipedia documents what has been published (and to some extent what is done) elsewhere. The article asserts that ellipticity is used in preference to eccentricity. This is a statement of fact, about the practice in the optics community. You may not replace such a statement with an argument based on your thoughts about what the practice should be. That would be "original research" as Wikipedia defines it. There are good reasons for this strict rule, and the policy linked above explains them so I won't reiterate. The only grounds for altering the statement would be if you thought it was an incorrect description of the practice, in other words if you have evidence that optical engineers and scientists do not always use ellipicity in preference to eccentricity.--Srleffler 02:48, 16 August 2006 (UTC)

I believe the answer is what the text was trying to say by "limited physical meaning", meaning that eccentricity as a measure is not very good because it is undefined for linear polarization, and infinitly sensitive to small amounts of ellipticity near linear polarization. So I put that in the article. Dicklyon 03:51, 16 August 2006 (UTC)

Never mind. I take it back. Looks like I was wrong about eccentricity. Dicklyon 03:56, 16 August 2006 (UTC)

I think of eccentricity in problems like orbital mechanics, where the problem is centered on one of the two foci of the ellipse (off-center). Ellipticity is better for problems which are symmetrical about the center of the ellipse, which applies to light polarization, and covariance, and spheroids. Pqmos 21:00, 14 November 2006 (UTC)

Discussion moved from srleffler's talk page
I seem to be interpreting the statement about ellipticity differently than you are. The article says "Ellipticity is used in preference to the more common geometrical concept of eccentricity, which is of limited physical meaning in the case of polarization."

You seem to read

"Ellipticity is used [by the field]."

whereas I read

"Ellipticity is to be preferred over eccentricity, which has limited physical meaning."

I believe I'm competent to disagree with the latter statement (or at least its second clause) without "original research." My question might have been better posed as: Is there a tradition (perhaps well-justified) of using ellipticity in the field of optics, (and if so, why)? In any case, unless I'm missing some point, I think the sentence should be reworded on NPOV grounds. Would it have been more appropriate for me to add a "citation needed" tag instead of posting to the talk page and planning an edit? Gnixon 21:35, 21 August 2006 (UTC)


 * Posting it to the talk page was entirely appropriate, and probably better than just putting a citation needed tag on it. Yes, you're right that I read the statement differently, focusing more on the first part rather than the second. There are really two separate issues here: Is ellipticity used in optics in preference to eccentricity, and if so, is this because the latter "is of limited physical meaning", as opposed to merely being due to some historical convention etc. In principle, WP:NOR prescribes that both questions should be settled by reference to published literature; you aren't actually supposed to use your own personal expertise in editing Wikipedia, unless you are certain that your contribution is backed up by verifiable external sources.


 * In this case, I think the first question definitely would need to be settled by reference to published sources, if you dispute it. I have several books I could consult, but unfortunately they are all at work right now. The statement that eccentricity is of limited physical meaning could probably just be removed if you are sure it is wrong, but give it a few days. That claim was added by User:Russell E several years ago, separately from the claim that ellipticity is the preferred parameter. Russell is still around, and I left him a note to check out this discussion. He may well look at the issue and say that he made a bad call in writing that, or he may have some explanation or citation to back up his statement. --Srleffler 04:14, 22 August 2006 (UTC)


 * I don't really know what I was thinking when I wrote that phrase ("of limited physical meaning.."). I guess I was thinking that, seeing as the polarisation ellipse is a phantom concept (except in the unusual case of coherent monochromatic waves where the inequality I^2<=Q^2+U^2+V^2 becomes the equality Gnixon quotes above) then it makes sense to work with a parameter that is more closely tied to the Poincare sphere representation of statistical polarisation instead (since the ellipticity angle, the arctangent of the ellipticity, is equal to half the "latitude" of the QUV-vector). But in retrospect, saying as much doesn't seem necessary at that point. We could just remove the phrase. We still need to say that ellipticity is used in preference to eccentricity, but I don't see any ambiguity; to me it is clear that it is descriptive ("is preferred") and not prescriptive ("is to be preferred"). However I don't see the harm in changing it, to, say, "is commonly preferred". As to the issue of personal expertise versus familiarity with literature, I did consult literature extensively when I made my major edits to this article. Rather than cite every sentence, which is just too cumbersome and really mainly suited to controversial or cutting-edge topics, I listed them at the bottom. I will admit though that I cannot recall whether I drew the reason for the convention from a specific source or from my memory. If we really want to say why, we should indeed try to find it in the literature.. unfortunately I don't have access to any texts at the moment but if I did I'd start with Born & Wolf. --Russell E 04:29, 22 August 2006 (UTC)


 * I'm used to everyone just working either in the coherence matrix Ei*Ej or the Stokes parameters, which of course define the Poincare sphere and can be used to define a mean ellipse for the polarized component. I personally find it useful to visualize the ellipse.  As for references, Hecht doesn't mention ellipticity and I don't think Jackson does, either.  Tomorrow I'll flip through B&W and perhaps a couple others.  Anyway, I can see the argument for using ellipticity since it's a coordinate of the Poincare sphere.


 * If there's someone who's familiar with the common practice in optics, I don't think a reference is necessary, but it would be nice if the sentence could briefly explain the reason. Can you take a shot at it, Russell?  Gnixon 06:02, 22 August 2006 (UTC)
 * I wouldn't like to take a shot at it as I'm now feeling unsure of it myself! And I don't have any texts on hand to check... sorry.--Russell E 07:31, 22 August 2006 (UTC)

P.S. I think math above is incorrect. What messes this all up is the factor of two relating angles in physical space vs QUV space. It is indeed true that V/Ip = sqrt(1-(U^2+Q^2)/Ip^2) and also that eccentricity = sqrt(1-ellipticity^2), it isn't true that ellipticity = V/Ip... V/Ip is the sine of twice the ellipticity angle, the latter of which is the arctangent of the ellipticity. sin(atan(epsilon)/2) is not equal to epsilon. --Russell E 05:25, 22 August 2006 (UTC)


 * Yikes, I think you're right. I may have to retract any claims of competency.  But it's late; I'll look at this again tomorrow.  Gnixon 06:56, 22 August 2006 (UTC)


 * On the other hand, if eccentricity = sqrt(1-ellipticity^2) and eccentricity = sqrt((Q^2+U^2))/Ip (according to Stokes parameters), doesn't it then follow that ellipticity=V/Ip !? Gawd, it's not late here but I'm suffering children-induced sleep deprivation!--Russell E 07:34, 22 August 2006 (UTC)


 * Sorry, you're absolutely right (and my face is red). With L=sqrt(Q^2+U^2) I calculate ellipticity=V/(Ip+L) and eccentricity=sqrt(2L/(Ip+L)), so neither parameter has the simple relationship to Stokes parameters that I claimed.  Stokes_parameters is no excuse for me since I put those statements there.  I really have no idea what I was thinking---you're probably right that I made a 0.5 error in some angle.  It would still be nice if that sentence explained why ellipticity (or ellipticity angle) is used.  Can the Poincare sphere argument be put concisely enough?  Gnixon 14:23, 22 August 2006 (UTC)


 * Ah good, glad I'm not going nuts. Those relations are at least a bit less ugly than mine with forward and inverse trig functions but you're right, they've no obvious geometric meaning in terms of the Poincare sphere. I've added words to this effect in the article... how's that?--Russell E 00:10, 23 August 2006 (UTC)


 * Thanks, I like it. Also thanks for catching my math error.  Actually, I'd be fine with not mentioning eccentricity at all, but what you have makes a useful point.  Gnixon 01:05, 24 August 2006 (UTC)

I wrote that sentence: "Ellipticity is used in preference to the more common geometrical concept of eccentricity, which is of limited physical meaning in the case of polarization."

I learned the concept of eccentricity in high school (part of analytic geometry, I think). I only encountered the much more esoteric concept of ellipticity when I began to work with circularly polarized light, over 20 years later. Circular Dichroism is, for instance, commonly reported in millidegrees of ellipticity. So I had to look it up. Once I compared them, it seemed obvious why the one that did not blow up would be preferred. The most important thing I wanted to accomplish was to help other people when they encountered ellipticity for the first time to realize that this was not the method they (most likely) already knew about for describing the shape of an ellipse. In other words, to raise a red flag for them, so that they would notice that this was something they did not already know about. --AJim (talk) 21:49, 19 March 2009 (UTC)

incoherent / noncoherent
Is anyone alse bothered by the use of the term incoherent? I think that means without logical thought, and that the term we want here is noncoherent, or perhaps non-coherent. tim 14:21, 16 August 2006 (UTC)


 * Incoherent is the more common term here. It applies to light just as it does to your thoughts (sorry, I couldn't resist).  Try googling "define:incoherent". Dicklyon 15:09, 16 August 2006 (UTC)


 * Yes, "incoherent" is the correct term for both thoughts and light: "incoherent: lacking cohesion, connection, or harmony;".--Srleffler 16:03, 16 August 2006 (UTC)

Introduction
I think this article's introduction needs expansion, and some clarification. It also (most importantly) needs to tie down the subject matter more specifically. If I gave you this piece of text on its own:


 * In electrodynamics, XXXXXXXXXXXX (also spelled XXXXXXXXXXXXXX) is a property of waves, such as light and other electromagnetic radiation. Unlike more familiar wave phenomena such as waves on water or sound waves, electromagnetic waves are three-dimensional, and it is their vector nature that gives rise to the phenomenon of XXXXXXXXXXXX.

You wouldn't know for certain what it was talking about, no matter how knowledgeable in the subject you were. However, if I instead gave you this:


 * XXXXXXXXXX is the economic theory holding that the prosperity of a nation depends upon its supply of capital, and that the global volume of trade is "unchangeable." The amount of capital, represented by bullion (amount of precious metal) held by the state, is best increased through a positive balance of trade with other nations, with large exports and low imports. XXXXXXXXXX suggests that the ruling government should advance these goals by playing a protectionist role in the economy, by encouraging exports and discouraging imports, especially through the use of tariffs. 

If you knew enough about the subject, you could determine that this was the introduction to Mercantilism. As I understand it, this should be the primary goal of the introduction - to define in simple terms the subject of the article so that a) readers who were looking for something else don't get bogged down in the rest of the page; and b) readers wade into the very nasty maths which comes later on armed with some knowledge of the subject matter. Correct me if I'm wrong on this.

Happy-melon 12:32, 30 May 2006 (UTC)


 * Actually, someone knowledgeable in the subject would immediately think of polarization. It's the most obvious distinction between electromagnetic waves and water or sound waves, which depends on the 3-dimensional (transverse) nature of the EM wave. I understand your point, though, and you're right: the introduction fails to give any sense of what polarization is.--Srleffler 14:23, 30 May 2006 (UTC)

I have omitted any reference to the fact the transverse waves must have the oscillation perpendicular to the direction of propagation. This is simply not true, my comments on the talk page about this have been ignored in the main text for the past year, thus I corrected it myself. I don't think I changed the intro especially well, and in general agree with happy-melons comments. I'll be glad if anyone rewrites the intro better, but please do not reintroduce the false concept of polarization always perpendicular to direction of propagation. This is true for plane waves and is stated in that section.Eranus (talk) 14:52, 23 July 2008 (UTC)

I reverted the intro to an older version (with minor revisions), July 24th. I don't understand why it is important to continue and write the misconception that polarization must be transverse to the direction of propagation. I don't understand what's wrong with the simple concept of polarization being the direction of the wave oscillation? Why do we need to say that the sound does not have polarization rather than saying that sound waves have longitudinal polarization. The important difference between sound and electro-magnetic waves, is that for sound waves the direction of oscillations is determined uniquely (It must be longitudinal)for any wave distribution, while for electro-magnetic waves the direction of oscillation is not unique, but limited to two dimensions by the transvesality condition div H=0.Eranus (talk) 12:37, 2 December 2008 (UTC)
 * Sorry Eranus. I was focused on trying to provide a simpler explanation, and didn't take the time to refresh my memory about this discussion. Your new text is better. I am still concerned about the intro though. Polarization is an important concept that comes up in everyday life and in high school (elementary school?) science classes. The intro to this article ought to be both technically correct and comprehensible to a smart 13-year old. At the least, the latter ought to be able to get an understanding of the typical case: a transverse EM wave in free space, which can have linear, circular, or elliptical polarization. That's not all there is to polarization, of course, but it is such an important special case that it needs to be handled prominently and in plain English. I'm not sure how we do that without losing accuracy, but I think we need to.--Srleffler (talk) 23:19, 2 December 2008 (UTC)

Photographic Polarizers
The picture showing how a polarizer on a camera enhances the appearance of clouds and sky has an error in the caption below. 

The caption claims that "... on circular polarizers because they emit circularly polarized light". This is wrong for two reasons. Most obviously, it can only "pass" or transmit any light, not "emit", or it would be a source. The other more important error is that these polarizers are never in my experience truly "circular" in the sense that they select one circular polarization. In my experience they are always linear polarizers, which can be rotated so the linear axis is at any orientation. They incorrectly advertise themselves as "circular" polarizers, but only because they are circular in shape and motion, but not in polarization.

I fixed this in the article, but then someone reverted it. So I am offering an explanation here. I'm a noob at Wikipedia, so correct my protocol if you can.

Discussion of circular polarization: Sunlight gets linearly polarized as it scatters from air and fine dust to make the blue color in the sky. So a linear polarizer can cut most of the blue from the sky and make it seem darker, so the clouds stand out. There wouldn't be any such effect from a circular polarizer: everything would appear as before, just darkened by 1/2 of the intensity or power. Of course reflected glare is often strongly linearly polarized, so a viewer or photographer can benefit from a linear filter.

But very few phenomena give rise to circular polarization. I have read that the light reflected from the carapace of certain scarab beetles is partially circularly polarized. You could only verify this if you had a circular-polarizing filter (not a linear filter cut in the shape of a circle). So I've been on the lookout for circular polarizers for years, but have yet to find one. I've read that you can make circular light from linear with a "Fresnel Rhomb", from which I think you could make a circular filter using that and a linear filter. But I have yet to do that. Pqmos 21:30, 14 November 2006 (UTC)

I reverted you. The way I interpret that caption and what I have read elsewhere, is that a so-called "circular polarizer" is neither a device that selects a single circular polarization, nor merely a linear polarizer cut in the shape of a circle. Rather, it looks like it is a device that selects a single linear polarization, and then converts that linearly polarized light to circular polarization. The polarizer thus emits, but does not pass circularly-polarized light, and it behaves when rotated the same as a linear polarizer, but the output, being circularly polarized, does not cause polarization-dependent changes in reflection off of the mirror in an SLR camera. One can make a device that does this for a single wavelength by combining a linear polarizer with a quarter-wave plate. I have no idea how they would make an achromatic version for photographic use.

Now, I am not 100% certain that my interpretation is correct, but I am pretty sure that a "circular polarizer" is not just a linear polarizer that can be rotated. There are lots of photography websites that discuss the choice between a linear polarizer and a circular polarizer for photographic use. The distinction seems to be that the latter is more expensive, but works with auto-exposure SLR cameras, while the former does not. This is consistent with my explanation, but not with yours. Hopefully someone else here can give us a more definitive explanation.--Srleffler 02:21, 15 November 2006 (UTC)


 * You are correct that a so-called circular polarizer is a linear polarizer sandwiched with a quarter-wave plate. It will select a linear polarization state from the scene just like a linear polarizer, but then mixes both linear states behind it so that the partially-reflective mirror will still direct about half the light to the AF sensor.  As for things in nature making circular polarization, optically-active molecules have that property, don't they?  If you really want to select a circular polarization state for the scene, like for the beetle, just hold the filter backwards, quarter-wave plate out.  But I agree that "emit" is perhaps the wrong word. Dicklyon 02:55, 15 November 2006 (UTC)

Very exciting. It sounds like they did the same thing I wanted to do with the Fresnel Rhomb and linear analyzer, but in a compact plate. I'll go check it out in a store somewhere and get back to you here. Thanks!!

Optically active molecules are supposed to rotate the plane of plane-polarized light, not select one circular polarization over another. (Also see the Wiki on Optical Rotation.) It seems to me that chiral molecules should glow circular if they fluoresce, and absorb linear and emit circular for transmission. The first may happen in chirally pure samples; and the second is probably precluded by phase-matching problems over extinction distances. If I can buy a circular analyzer, I'll try to check that out too, but it may be tough to find chirally pure substances -- they'll have to be bio-molecules. 199.46.200.233 17:40, 17 November 2006 (UTC)

I checked out a camera "circular polarizer", and indeed it appears to be more than just a linear polarizer cut in a circular shape. Looking one way through it (proper for the camera) it works for the eye just like an ordinary linear polarizer, like polarized sunglasses. But turned around, it only changes the (apparent?) color of linear polarized light without blocking it. I say apparent color because the faint shades of blue and yellow I saw correspond to the appearance of linear polarized light to the human eye as a very faint blue-on-yellow hourglass (stare at a spot on a blank white lcd monitor, and slowly tilt your head, watching for a blue and yellow pattern about an inch in diameter).

I'm convinced this is not just a linear polarizer, and that the wiki article has no error regarding the circular polarizer as it is written. Should I delete this sub-discussion, or condense it for the benefit of others like me who may follow? Thanks to all who helped me get this right. 199.46.199.231 18:58, 2 January 2007 (UTC)


 * The word "emits" is probably not correct. See my description above.  There's no need to remove complicated discussions that may be of value to someone who is trying to understand what's up to improve it. Dicklyon 20:06, 2 January 2007 (UTC)


 * In case you don't already know this, the blue and yellow pattern you describe is known as Haidinger's brush, and appears due to a weak polarization sensitivity in the human eye. As Dicklyon indicates, the discussion should remain as it is. Talk pages are a permanent record. When they get too long, we archive them.--Srleffler 23:41, 2 January 2007 (UTC)

The classical way to produce circular polarization is, indeed, to place a quarter wave plate after a linear polarizer. There is an additional requirement that they be in a particular relation to each other. The axis of linear polarization should be half way between the fast and slow axes of the quarter wave plate. Think of the linear polarization as the vector sum of two orthogonal components, one parallel to the fast and one parallel to the slow axis of the plate. After passing through the plate their relative phases have changed so that their vector sum is now circularly polarized. The direction of passage matters; if the light enters the quarter wave plate and exits through the linear polarizer, the exit light will be linearly polarized. That is, turning a circular polarizer around converts it into a linear polarizer. --AJim (talk) 01:58, 20 February 2009 (UTC)

Poincaré sphere is a 3-manifold not a 2-manifold
Something is wrong in the discussion of a Poincaré sphere in the present version of this (polarization) article. Here the Poincaré sphere is presented as a 2-sphere, a 2-dimensional objected embedded in three-dimensional infinite Euclidean (flat) space. This is wrong. See Poincaré sphere.

i can believe that the genuine Poincaré sphere is useful for understanding polarisation and the Stokes parameters. However, describing it incorrectly is not going to help anyone to understand anything. If someone understands this, then please correct the text.

Boud 14:08, 9 February 2007 (UTC)

i put on the two sections where Poincaré sphere is discussed rather confusingly here. Feel free to remove them once the confusion is sorted out. If optics people use the term in a way totally different to the way mathematicians use it, then that would need disambiguation. Boud 14:21, 9 February 2007 (UTC)

Careful - Instructor at work
I'm a student at the K.U.Leuven, Belgium. We're currently participating in a project concerning polarization in collaboration with another college. Apparently one of the instructors there has deliberately entered erroneous information in articles concerning polarization. There is no specific information available as to the precise identity of the instructor or to the nature of the information he has slipped in, but I thought people should be aware of this. —Preceding unsigned comment added by The Akulamatata (talk • contribs) 10:27, 17 May 2007


 * Did anybody notice this? It should be possible to look for IP addresses from Belgium and check out the edits made around the time this comment was posted. 140.247.79.217 (talk) 16:50, 18 September 2008 (UTC)
 * I noticed it at the time. I took it as likely a hoax, or an attempt to encourage students to check other sources and not treat Wikipedia as "gospel". Deliberate errors tend to get fixed pretty quickly, anyway. I don't see any sign of unreverted harmful edits in this article between 31 Dec 2006 and 23 May 2007, but of course the warning above doesn't say that the edits are in this article, but rather that they are in "articles concerning polarization".--Srleffler (talk) 23:41, 18 September 2008 (UTC)

Wording
''The discussion below is copied from User talk:Srleffler. It is in regard to this edit.''--Srleffler (talk) 03:43, 14 December 2007 (UTC)

I admit that "movement" is not perfect, but the word "evolution" is simply wrong in this context. This is a scientific article, so when you use scientific words, they should be used properly. —Preceding unsigned comment added by 74.134.251.147 (talk) 03:35, 13 December 2007 (UTC)


 * I'm not sure if I missed your point, or if you just don't understand the meaning of the word "evolution". As far as I can see, the word is used properly.


 * Excerpt from dictionary.com:
 * ev·o·lu·tion, noun
 * 4.	a process of gradual, peaceful, progressive change or development, as in social or economic structure or institutions.
 * 5.	a motion incomplete in itself, but combining with coordinated motions to produce a single action, as in a machine.
 * 6.	a pattern formed by or as if by a series of movements: the evolutions of a figure skater.
 * 6.	a pattern formed by or as if by a series of movements: the evolutions of a figure skater.


 * --Srleffler (talk) 04:40, 13 December 2007 (UTC)

If somewhere in this article was a section about a "theory" involving polarization, but the word was used with the common meaning of "guess" instead of the scientific meaning, it would quickly be changed. "Evolution" has the same problem. It has an exact scientific definition as well as common definitions (like the definitions you would expect to find in a dictionary after the better, more scientific ones have been given). In a scientific sense to evolve means the gradual change due to a selective pressure, and only that. Your use matches the COMMON usage definitions, and would be fine to go with in a COMMON article. But this is a scientific article, and evolution is simply not the right word here. —Preceding unsigned comment added by 74.134.251.147 (talk) 06:27, 13 December 2007 (UTC)


 * If this were a biology article, I might agree with you. The usage seen here is not unusual in physics, however. Something that changes gradually and progressively from one state to another is properly described as "evolving" from the first state to the later one.


 * This discussion really should be at Talk:Polarization. I will copy it there. If you want to reply, please reply there. I'll see your response.--Srleffler (talk) 03:43, 14 December 2007 (UTC)

Quantum?
MattSzy pointed out an error in the following text, in User talk:Rkundalini. I have cut it out and put it here until someone can correct it (or until I get around to reading up on the topic)... excised text follows

Since photons are spin-1/2 particles, mathematical descriptions of polarization states are closely related to spinors. Individual photons are inherently circularly polarized, and the coherency matrix is equivalent to the density matrix of quantum mechanics, if expressed using a circular basis. The quantum mechanics version of the Poincar&eacute; sphere is the Bloch sphere.

Rkundalini 15:07, 25 Jun 2004 (UTC)

Yes it was right to cut that out: Photons are spin one particles. Lasers would not work with spin 1/2 particles. The satement that photons have spin one is equivalent to the statement that the electric field is a vector field. The lack of a third spin direction has something to do with the photon's zero rest mass. Indevidual photons can have linear polarization. "Density matrix" is a statistical mechanics concept and does not appear in the description of pure quantum states. I don't know about the Bloch sphere, but I don't think it needs to be mentioned here. David R. Ingham 21:59, 9 March 2006 (UTC)
 * Polarization is the intensity or the intrinsic magnetic momentum(the spin)? statement like individual photon is linear polarized, any reference? Jackzhp (talk) 03:30, 21 October 2013 (UTC)
 * Your question is not clear. Individual photons are certainly not linearly polarized, however. They are circularly polarized.--Srleffler (talk) 05:58, 21 October 2013 (UTC)
 * MathPages: The relationship between photon spin and polarization

External links modified (January 2018)
Hello fellow Wikipedians,

I have just modified one external link on Polarization (waves). Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20050405221855/http://www.colorado.edu/physics/2000/polarization/index.html to http://www.colorado.edu/physics/2000/polarization/index.html

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 05:43, 22 January 2018 (UTC)