I saw on the news this morning that 3D televisions for the home will be demonstrated at the upcoming Consumer Electronics Show. That got me wondering about the technique involved.
Supposedly the technology used in Avatar-3D relies on circularly polarized (CP) light: images meant for the left eye use clockwise-polarized photons, while images meant for the right eye use counterclockwise-polarized photos (or is it the other way around in reverse?); the glasses are CP filters whose left and right lenses have opposite “spin:” they only allow the appropriate photons through for the left or right eye as required.
Soooo, some questions about polarization:
-Is it possible for a light source to preferentially produce photons of a given polarization, either CW, CCW, or linear at some particular angle?
-Do CP filters only allow photons through that have the right direction of rotation AND the right helix angle (i.e. the E-field is rotating at the correct rate), or do they let through any photon whose polarization is rotating in the correct direction?
-If the light source used in 3D projection systems does not preferentially produce the desired polarization - meaning filters must be used - isn’t that 3D projection system going to be monumentally inefficient (i.e. lots of power input, lots of heat dumped into the filters, before any light even gets out to the viewer)?
I’ve never heard of a polarized-light emitter so I’m guessing that filters are used. You would get light energy absorbed by the filters but my WAG is that only a small fraction of a projector’s or TV’s energy consumption goes into light energy anyway.
If I remember correctly from my college waves course, the rate of rotation of the E vector is just the frequency. So filtering by helix angle is another way of saying filtering by color. I’m sure any filter would have some color-dependent loss. If it’s significant in real life I’m sure the sets would compensate by adjusting the intensities of the red/green/blue pixels.
Yes, it is possible to make a light source that is inherently polarized. The easiest way to do this is with a laser. The light source in a movie projector is not going to be polarized.
As Edward stated, the helix angle is fixed for a given wavelength. The direction of the electric (or magnetic) field rotates 360 degree in one wavelength of propagation. Circular polarizers work only over a range of wavelengths, so yes, there will be some loss and crosstalk as the color of light deviates from the center of the polarizer’s bandwidth.
Efficiency is indeed an issue. At least half the photons are wasted as they go through the polarized eyeglasses. If you use filters, you lose another factor of at least two in the projector. The latter can be avoided by using a beam splitting polarizer, like a Nicol prism, which can separate the light into two beams of opposite polarization.
The answers have pretty much been given already, but I’ll add a few comments
Lasers aren’t inherently polarized – it’s certainly possible to buy unpolarized lasers. But It’s particularly easy to make many gas lasers polarized by building them with internal Brewster windows, which minimize reflection losses without expensive anti-reflection coatings.
It is possible to make polarized light without using polarizing filters. The Zeeman effect, for instance, produces light with differrent wavelengths having different polarization states, and there are other sych effects:
It’s certainly not a practical way to create such light for a movie, however – you’re much better off creating unpolarized light and running it through a polarizing filter.
You can describe light with any polarization in terms of a combination of Right Hand Circular Polarized Light (RHCP) and LHCP. The circular polarizer will transmit whichever portion matches it (RHCP, say) and block the rest. The phase angle of the E-field vector is unimportant. If you throw linearly polarized light at a circular polarizer, half of the light will come through (as circular polarizeer light) and the rest will be reflected or absorbed (assuming perfect filters)
As stated above, yes. You’re throwing away half the light with the filter at the projector, so you need twice the intensity. (At the viewing end – you – all of the RHCP goes into one eye and all the LHCP goes into the other, so in a way you’re not losing at that end.)
Right, but the one eye that gets the RHCP gets all of it (and none of the LHCP light), and the reverse is true for the other eye, so in each case one eye gets all of the light from the projector intended for it (after, of course, the projector has thrown away half the light in the first place), so there’s no loss there. You don;'t want the RHCP eye to get any light from the LHCP image, since that would be double-imaging.
The original IMAX 3D films used polarized light; two projectors (or was it 1 projector and rotating filters?) to produce a verical and horizontal polarized image projected on the screen at once. The glasses are horizontal and vertical polarized one for each eye, so each eye sees a different picture. Take them off and you see a doubled, blurry and overbright picture. (actually, you see 2 pictures - hence the bright - and both eye views superimposed.)
The TV will likely work differently. Instead of using polarized light, you have a pair of glasses synchronized to the TV somehow. (Bluetooth?) The TV flashes a LH frame, and the glasses block the RH eyepiece. Flash RH frame, block LH eyepiece. The blocking is done with liquid crystal shutters; apply a current, glasses lens goes opaque. This technique was available for IBM PC’s over 10 years ago. However, speed of LCD shutters and TVs is much improved since then; those new LCD TV’s that do 120HZ refresh rate can probably do this trick best. Theoretically you don’t nee a new TV, just a box that syncs the glases with the signal fed to the TV.
Polaroid sunglasses were useful because light reflected from a very oblique angle from a horizontal surface tends to be horizontally polarized. The up-down light is more likely absorbed, the left-right waves bounce. You can see the difference by rotating polaroid glasses while looking at a reflective surface and an angle. This is also why polaroid filters are good when trying to take pictures into display cases - you don’t also get the reflections off the glass.
I suspect they are pushing the technology now so people will rush out to buy it to see Avatar when available on Bluray.
Cal, help me out here, as I don’t really know, but I get the impression that this isn’t correct. That is, they sell “polarized” and “unpolarized” lasers, but in the case of what they call “unpolarized” lasers, don’t they really mean that the polarisation of the laser is uncontrolled and often fluctuates?
To be more specific, I think that “unpolarized” gas lasers whose elements are all parallel, no Brewster angles, typically are fairly strongly polarized at any moment, but the polarization angle might flutter around, or might hang up at one orientation for a while and then adopt another, or for all I know might spin around for a while.
Doesn’t the radiation being coherent imply that there is stable polarization of some kind for at least durations comparable to the period over which coherence is measured?
So, if you really need a source of light that has no preferential polarization, say perhaps if you want to do scattering distribution measurements with asbestos fibers sitting on a slide in order to model the statistics of their orientations, you couldn’t use an “unpolarized” gas laser, could you?
I wonder if the same could be true of ViCSEL (not edge emitting) laser diodes, because they can have rotational symmetry, but know little about them.
Well, in a sense what you ay is true, but it’s true about any sort of light – not just lasers.
A beam of light always has a particular state of polarization (which might be linear vertical, or linear horizontal, or linear at an oblique angle, or circular, or – the most general, since it covers all bases – elliptically polarized). This state tends to change randomly over very short periods of time (the coherence time of the light) unless you impose a particular polarization state (e.g., by filtering out other possible states with a filter).
In the case of gas lasers with internal Brewster windows, those windows act to ensure that only linearly polarized light comes out. If you use anti-reflection coated windows instead, or put the laser mirrors right on the ends of the gas discharge tube, though, the light that emerges is unpolarized. If you place any sort of polarizing filter in front of it and rotate it, you wont change the intensity of the light.
But is it truly “unpolarized”, or simply changing its polarization state from one state of elliptical polarization to another at random and at frequent, short intervals. Well, it’s doing the latter. But so is light from the sun, or from a flashlight, or just about any source. The main difference is that the coherence time in a laser tends to be longer, but it’s still pretty damned short, unless you work to make it longer. In other words “unpolarized” really means “changing its polarization state from one state of elliptical polarization to another at random and at frequent, short intervals”.
To my knowledge, there is NST as “circularly polarized” light… but that polarization can occur, & be modified, through such things as filters, or by passing the beam of light through at an angle to transparent planes. Polarized glasses, for instance, have a “micro-fence”-like arrangement allowing a particular movement of wave pattern (for instance, fishing glasses allow only vertically-polarized light, which is typically reflected in far less a quantity than horizontally-polarized light, which is reflected at an angle by the surface of water, a transparent plane; hence objects below the surface being much more visible). Try wearing a pair, then tilt your head from side to side. You’ll see the difference immediately. A simple example of using a beam-splitter to produce oppositely-polarized laser light is to use a laser pointer & send it through a window at an angle - the beam that is reflected will be polarized parallel to the plane of reflection & the beam that is transmitted will be polarized 90 degrees apart.
Another method of producing 3-D is old-school: using 2 different colors through two different filter-glasses lens. Another method is switching the timing pulse of each different side alternately, matching the alternating LCD from clear to opaque in the glasses lens, producing the 3-D effect w/out polarization.
I’ve been in contact w/ a laser-technology & manufacturing firm in Canada on another project… I’ll pass this Question along to him & get back to you on it, if he’s got anything more on it. Also, my uncle was one of the people key in development of the Bell Labs first ruby laser. He’s still alive & enjoys his hobby as a gemologist.
In the meantime, I’ve got a laser-type Question to post; it should be stimulating (pardon the pun!)
I think you’re expressing yourself badly here. You don’t seem to be saying that circularly polarized light doesn’t exist (which is what your words, taken at face value, appear at first to say), but that it doesn’t exist in nature. Most light in nature is either unpolarized or, if it has a polarization state, is elliptically polarized. By chance alone, some of that will be circularly polarized. Strictly speaking, your statement isn’t true. It is true that sources of circularly polarized light are going to be rare, and that it’s much easier and more reliable, as I say above, to take an unpolarized source and pass it through a filter (A circularly polarizing filter, or a linear filter followed by a quarter wave plate).
Just for the record, the first ruby laser was built by Ted Maiman at Hughes Research Labs in California, not at Bell Labs. Other lasers were originated at Bell Labs, including the uranyl salt laser (the second laser to be built) and the Helium Neon Gas Laser. An interesting aside – Bell labs insisted on called lasers “Optical Masers”, and they were so designated for years in both internal publications and in papers originating at the :Lab. It was years before they dropped the clumsy term in favor of “laser”. I’m sure there was a “first ruby laser” built at Bell Labs, but the thunder was already taken by Hughes. Are you sure your uncle wasn’t involved in one of the other types of laser that was first built at Bell?
This is a hotbed of controversy. There is an article in this month’s “Physics Today” (“Bell Labs and the Ruby Laser”) about this very subject, told from the point of view of the Bell Labs people. Maiman and the Bell Labs team were both racing to be the first to make a ruby laser. Maiman had a press conference in which he claimed success, which caused the Bell Labs team to accelerate their work on ruby. They made it work and it became clear that, while Maiman had achieved optical amplification, he had never gotten the thing to do what we would now call “to lase”, i.e. the gain was not high enough to sustain an oscillation. So it becomes a matter of semantics. Now days, no one would call it a laser if it didn’t reach threshold. In those days, the nomenclature was not established and no one seems to disagree that Maiman was the first to measure gain and the first to establish “pink ruby” as a viable material, instead of the “dark ruby” that the Bell Labs team started with.
The whole field of the original laser inventions has been controversial.
I think you actually do end up with losses at both ends. There’s some intensity of light entering the human eye which is considered acceptable brightness for a movie. In an ordinary movie, you project a single movie-bright image onto the screen, and the same image is seen by both eyes. In a 3D movie, you project two different images onto the screen, and since each of them will only be seen by one eye, each of those two images will have to be movie-bright. And then, the simplest way of producing each of those two movie-bright images is to start with light that’s twice as bright and throw away half of it in a filter.
You’re projecting two images, A and B onto the screen
You’re saying that because of filtering on the viewer’s head half of A gets to your left eye and half of B gets to your right eye.
But that implies that the other half of A, filtered out before the left eye, would pass the filter in front of your right eye, defeating the 3D scheme.
Half of the total incoming light is being discarded at each eye, because it’s intended for the other eye (the whole image on screen is a superposition of the two channels), but if half of each channel is being filtered out at the viewer’s head, what does that consist of?
Sorry guys – this is a matter of semantics. Assuming that you have perfectly lossless filters, then I maintain that the right eye gets 100% of the light projected in its polarization, so all the loss is taking place at the projector, while the left eye gets 100% of thew light intended for it.
Of course, your right eye only sees 50% of the light that hits the screen, so you could say that you’re losing 50% of the light there. But it’s all in your definition. If you go around a tree while a squirrel on the other side goes around the tree, keeping it between you and him, have you gone around the squirrel?
I’ve got my copy, but haven’t read this yet – I’ll have to have a look. Laser history is, indeed, full of such nomenclature inexactness. Most “nitrogen lasers” aren’t – having little or insufficient feedback (they’re essentially ASE sources). In my own work, I maintain that I built the first laser of one gain medium (having a true resonator cavity), while a previous paper claiming lasing really observed Superfluorescence, with no mirrors present.