3D Glasses: Diffraction or Not?

[cragar]

Do 3d glasses use diffraction to work , because the ones i read about did not say anything about diffraction , but i shed a laser through my 3d glasses and it diffracted it perfectly on the wall , it had a center bright spot and then equally spaced dots from the center out .

 

[berkeman]

Do 3d glasses use diffraction to work , because the ones i read about did not say anything about diffraction , but i shed a laser through my 3d glasses and it diffracted it perfectly on the wall , it had a center bright spot and then equally spaced dots from the center out .

AFAIK, 3-D glasses use polarization for separation.

 

[Claude Bile]

3D glasses work by filtering out elements of a stereoscopic image. Old-school 3D glasses use color filters, more modern ones filter polarisation. The newest generation that come with 3D televisions use what one might call temporal filtering. 3D TVs rapidly switch between "right eye" and "left eye" images; the glasses block the left or right eye in sync with the television.

Claude.

 

[nasu]

Maybe they are not 3D glasses but the ones made with diffraction grating. I have a pair like that. We use them in the planetarium and the student astronomy lab so you can see the spectrum of various light sources by just looking at them in a dark room.
Where did you get the glasses from?

 

[cragar]

I got them at an indoor mini golf place they had black lights and stuff , and he was like do you want 3d glasses .

 

[K^2]

What SORT of 3D glasses? There are all kinds.

First of all, it's all stereo, not true 3D, just to make it clear. If you move around, you still see the picture from the same point of view, unlike with true holography, for example.

The way all stereo glasses work is by letting you see objects with one eye slightly different than the other eye.

But anyways stereo glasses generally fall into following categories.

1) Refractive "3D" glasses simply split the light into spectrum. Objects of different pure spectral colors will have different "depth" to them. If you ever saw the 3D glasses that work with colored chalk or pencils, that's the type. It's not real 3D, but it can look like it. Fun to play with. (Sounds like these are the kind you got.)

2) Anaglyph. These are the classic red-and-blue 3D glasses. They let each eye see only one color, blue and red respectively. Two images are superimposed, again in these two colors, one for each eye. Since each eye sees only its own picture, you can get the stereo effect.

3) Shutter glasses. These are usually built with pair of LCD cells. Not like a full screen, but just a single "pixel" for each eye. It can turn transparent or opaque. Only one is transparent at a time, and it is synchronized to flickering images on the screen. These work very well with CRT screens and projectors. Not so well with LCDs. Apparently, that's what the "3D TV" is going with, so they either figured out how to make LCD flip images fast enough, or something else is going on. I haven't really caught up on that yet.

4) Polarizer glasses. These are a lot like polarizer shades you'd buy for driving or sailing, except that polarization on two halves is different. So again, it can allow only part of the image through. These are usually used with projectors. Real 3D, which has become real popular in the theaters, uses circular polarization, which takes care of many issues these used to have in the past.

5) VR Goggles. These are the real deal. Two individual mini-displays, one for each eye, generate a virtual 3D screen in front of you. Since each screen is controlled completely separately, and there can be head tracking, this is the best option to create true 3D with a help of the computer.

As far as effects on the laser beam, the first one will refract it, so the beam will be "bent" as it passes through them. And both the shutter glasses and polarizers will have effect on laser beam because laser beam is polarized. You should see the difference by simply rotating the laser pointer around its length axis with these.

 

[nasu]

I got them at an indoor mini golf place they had black lights and stuff , and he was like do you want 3d glasses .

OK, I think they are not actually 3D glasses. It makes more sense for an indoor place with all kind of lights. It will give nice effects.
Do you see rainbows when you look at a light source? If you have any gas discharge source (neon tube, sodium streetlights) you may be able to see the lines in the spectrum.
Actually you said that you tested with the laser. Do you see a rectangular pattern of dots on the wall? If yes, this is a 2D grating (periodic along two directions) and those are used for diffraction glasses (also called holographic grating). These glasses are sold by the same places that sell all kind of cheap real 3D glasses so it may be just a confusion.

 

[cragar]

ok here is the link to the 3d glasses
http://www.3dglassesonline.com/3d-chromadepth-glasses/

sorry i should have put this up when i made the thread .

 

[nasu]

OK, then they are 3D in some sense. See the link below, the Chromadepth is close to the bottom of the page.
They say that there are some micro-prisms (it can be a diffraction grating of sorts) on one of the lenses, but not on the other one. You can test this with your laser.

http://en.wikipedia.org/wiki/Stereoscopy

 

[mgb_phys]

@K^2
There is another very promising type which is a variation on the old anaglyph (red/blue) goggles.
These split the red,green and blue bands into a number of smaller wavelength ranges and block alternative ranges in the left and right eye.
So for instance red from 630-640nm would be left, 640-650nm would be right, 650-660nm left again and a corresponding comb filter on each camera.
The number of bands depends on how expensive you want the filters.

Your eye is very good at working with missing colors, many of us are slightly color blind to some extent but your onboard software takes care of it.

 

[cesiumfrog]

The number of bands depends on how expensive you want the filters.
Your eye is very good at working with missing colors

Perhaps it is (would you give an example?). But also, we only have three types of colour receptor, which is pretty much why RGB displays can replicate any colour in the first place. I don't see why you would pay for a system using more than six bands in total.

 

[denni89627]

Perhaps it is (would you give an example?). But also, we only have three types of colour receptor, which is pretty much why RGB displays can replicate any colour in the first place. I don't see why you would pay for a system using more than six bands in total.

I'm thinking you could have more depth gradients than the two you would see on what you describe.

 

[mgb_phys]

Perhaps it is (would you give an example?). But also, we only have three types of colour receptor, which is pretty much why RGB displays can replicate any colour in the first place. I don't see why you would pay for a system using more than six bands in total.

The response of your colour receptors overlaps so it's not so simple.

The problem with just splitting each R G B into a short and long is that if an object in the scene is monochromatic it would only appear in one eye which would destroy the stereo.
Ideally you would have a comb filter with say odd nm in one eye and even nm in the other.

 

[cesiumfrog]

The response of your colour receptors overlaps so it's not so simple.

The problem with just splitting each R G B into a short and long is that if an object in the scene is monochromatic it would only appear in one eye which would destroy the stereo.
Ideally you would have a comb filter with say odd nm in one eye and even nm in the other.

What?

I thought we used digital these days, so if the problem you describe ever existed, we would simply program the outputs to automatically compensate for typical retinal response.

But the way the colour is captured in the first place is by using normal video cameras in stereoscopic configuration. Not by having one camera collect two feeds from disjoint bands of the spectrum.

 

[mgb_phys]

But the way the colour is captured in the first place is by using normal video cameras in stereoscopic configuration. Not by having one camera collect two feeds from disjoint bands of the spectrum.

It depends on the system the basic idea is for the left eye to only see the image captured by the left camera and v.v.
With a polarizer (Real3D passive or NVidia active glasses) system - color is captured by both cameras and isn't the major issue.

The color band system is based on the old anaglyph red/green glasses - the problem with those is the 'purple cow' effect. You can't film strongly colored objects because a red flower would be white in the right eye (red filter) and black in the left (green filter). Hence you need purple objects - which are rare in nature!

The improvement is to split the RGB filter on the camera into a short and long half of each red,green,blue band. One on the left camera and one on the right. With a similar filter on the glasses you see R_s,G_s,B_s in one eye and R_l,G_l,B_l in the other.
This means you can have stereo full color imagery.

But if you had a red monochromatic object it might appear only in one of the red bands and so would appear bright in one eye and dark in the other- just as a red flower with the red/green glasses.

One solution is a lot of software to post-process the footage and work out which objects appear different in the left/right feeds because of their position (which you want) and which appear different because of camera angle and ligthing effects - and adjust the image.

Disadvantage of this is that it takes a lot of time, money and skill - this is why avatar took so long to edit and cost so much.
Advantage is that they pay me lots of money for developing the software

The alternative is to split the color bands into many more than two parts so the chances of anything other than a laser being in exactly one band is small.

 

[cesiumfrog]

mgb, your posts paint a confusing picture. On one hand, you imply that you are personally working in this industry. On the other hand, you give other signals of knowing very little of this industry (like repeating incorrectly the name of the big player, RealD), and it seems strange that you consider the shutter system as a subset of polariser systems (since the polarisers in the LCD shutter are incidental) and mention the brand used for PCs rather than cinemas.

Here's how the actual colour band system works:
"Dolby 3D uses a Dolby Digital Cinema projector that can show both 2D and 3D films. For 3D presentations, an alternate color wheel is placed in the projector. This color wheel contains three additional color filters to the three color filters found on a typical color wheel. The additional set of three filters are able to produce the same color gamut as the original three filters but transmit light at different wavelengths."

In other words, you record the two video feeds just like normal (either using CGI or stereo configured identical normal colour video cameras). If an object is monochromatic, it's still going to look the same in both left and right feeds (except for orientation). I take it the projector sequentially displays the six component images in synchronisation with illumination filtered through the six sections of the spinning colour wheel. So although (just like an OLED TV) the final output (even after filtering through either side of the glasses) may be nearly monochromatic, it closely reproduces full spectral perception.

But if you had a red monochromatic object it might appear only in one of the red bands and so would appear bright in one eye and dark in the other- just as a red flower with the red/green glasses.

Wrong, that isn't how the video is captured. At worst, it might appear slightly more crimson in one eye and slightly more flame or vermilion coloured in the other, and only if the projection wavelengths are different enough to produce a noticeably different response from the typical eye. (But since the whole process is digital, it would be absolutely trivial to run a transformation at some point to rebalance the colour levels and cancel this out if it is noticeable. I don't know whether the system involves an approximate frequency comb, but if so it is not for quite the reason you suggest.)

 

[mgb_phys]

On one hand, you imply that you are personally working in this industry

I do

it seems strange that you consider the shutter system as a subset of polariser systems (since the polarisers in the LCD shutter are incidental) and mention the brand used for PCs rather than cinemas.

The industry generally refers to them both as polarizer systems. But yes the use of polarizers in shutter glasses is incidental, it's just that mechanical shutter glasses are a little bit more inconvenient.
There are shutter glasses that use a switched retarder which is a bit more polarization-ish.

Here's how the actual colour band system works:

That's how Dolby's split-color system projector works. It's a good idea and avoids some of the head position issues of a polarizer system.
I work on the camera side - let's just say that there are alternatives coming soon to a cellphone store near you that use a different approach.

I take it the projector sequentially displays the six component images in synchronisation with illumination filtered through the six sections of the spinning colour wheel.

Some use a single projector, some a pair - it depends on the size of the screen. There are limits to the power of a single digital projector,

But since the whole process is digital, it would be absolutely trivial to run a transformation at some point to rebalance the colour levels and cancel this out if it is noticeable.

If you have a two camera system. Although getting lighting/color levels exactly the same over a stereo system is surprisingly difficult.
The problem is that it's more psychology than physics, some optical effects that are huge in terms of MTF or proper physics measures you don't notice at all - while some really subtle effects totally destroy the stereo effect.
It's not always clear on paper exactly what perfromance you need to achieve for what paramters.

 

[Zarqon]

hmm, I'm curious, how good frequency comb filters can be made today? Considering that this technique would offer passive glasses and a lot less sophisticated TV than other 3D approaches, it feels like the major thing holding it back would be that it's hard to actually do the glasses sufficiently well, i.e. with narrow peaks and a flat spectrum. Is it possible to do them with within 1 nm, i.e. even nm in one eye and odd in another? Is this narrowness enough to make it look good?

 

[mgb_phys]

As cesiumfrog pointed out - with a projector/TV system you don't need a comb filter - 6 colors pretty much works for most people (who aren't color blind).

But for a limited viewing angle polarizers work pretty well and are cheap and easy to make, not sure precisely what advantages Dolby's system has on the projector side

 

[Zarqon]

yeah, sounds reasonable. Though I was wondering about the home TV setup, where you can't use the polarization teechnique. It seems to me that if one only needed twice the amount of pixel segments (6 colors in stead of 3 for every pixel) to give any TV 3D-capability then it would have been done already because it seems like a simple thing, so there must be something else that is difficult with that approach? Maybe the exclusion of colorblind people is enough that you don't want to go for it as the end solution.

 

[JDługosz]

As cesiumfrog pointed out - with a projector/TV system you don't need a comb filter - 6 colors pretty much works for most people (who aren't color blind).

But for a limited viewing angle polarizers work pretty well and are cheap and easy to make, not sure precisely what advantages Dolby's system has on the projector side

The screen that reflects polarized light is expensive (made with real silver!). I suppose that plain colors would not have that problem.

 

[mgb_phys]

The screen that reflects polarized light is expensive (made with real silver!). I suppose that plain colors would not have that problem.

The silver screens are also necessary for reflectivity, since the polarizers cut out half the light.
If you half the color band you also reduce the amount of light so would probably still need a color screen.

The big financial driver is to have a technology that works at home as well as the big screen and at home a system that can seamlessly show 2D is important.
The nice thing about the polarizers is that you don't lose any resolution with a 2D image

 

[cesiumfrog]

The screen that reflects polarized light is expensive (made with real silver!).

Is this strictly necessary? At the cinemas where I've seen the circular polarisers used, the blank screen has appeared white and ordinary.

 

[cesiumfrog]

Any word on holographic video?

The current 3D technology is really merely stereoscopic: both eyes still keep their focus fixed on the screen. On the one hand, this provides better-than-natural clarity (since sharpness of different layers is not limited by the depth you select to focus your eyes at), but on the other it is probably the cause of eye complaints (and apparently is implicated in developmental eye problems). With sufficient resolution it should be possible to create true 3D video (on a big screen you would see slightly different angles by walking across the cinema), but so far I've only seen digital holographic techniques http://sciencewise.anu.edu.au/articles/Holographic_Neurone_Stimu". Or are there any projects to imitate true 3D using, say, retina projection glasses?

 

[mgb_phys]

At the moment it's difficult to do moving holograms - even static digital ones require a serious amount of computational power.

These guys are about the best http://www.zebraimaging.com/