h a l f b a k e r y
What's a nice idea like yours doing in a place like this?
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
or get an account
Certain aspects of this Idea have popped up in other places. I'll add some links, but I haven't seen this particular combination before, so....
First, we need a high-resolution LCD. See the IBM link, although I'm pretty sure that Sharp's 3D monitor, which has a sheet of prisms on its surface, is
equivalently high-resolution. Expensive, but necessary here. Let me assume an "initial" monitor resolution of 2560x2048 pixels, having 4 times the pixels of a lower-resolution monitor of the same size (1280x1024, a standard).
However, due to what is described below, we could reduce the cost of the monitor if it was made high-resolution in only one of the two dimensions. That is, 2560x1024 pixels would be fine.
Next, this particular LCD needs to be constructed in a somewhat special way. See the "How LCDs Work" link. Note that by-default they work with polarized light. Normally, every cell in the LCD works the same way, using polarized light the same way.
The key to this Idea is that we want to make alternate columns of cells, in this LCD, use polarized light in opposite ways. We have in effect divided our monitor screen into thin vertical slices, 1280 of which polarize light one way, and the other 1280 polarize light the other way. Vertical resolution is unchanged in every slice (that's why it would be OK if it was 1024 instead of 2048).
Now we set the video system to display two images at once on this monitor screen. One image goes to one set of 1280 vertical slices, and the other image goes to the alternated set of 1280 vertical slices. If you looked at the result directly with the bare eyes, you would probably see something like the Idea in the "Blurred display" link.
In this Idea, though, you would wear polarizing glasses (with clear lenses, not for use as sunglasses). These are cheap and don't affect color-perception. They will ensure that your left eye sees only half the vertical stripes of the LCD screen, while your right eye sees only the other half. Note that the higher-resolution is the display (in the dimension we slice it) the less-obvious it will be that each image has been sliced up. Your brain will easily merge the two images for the full 3D effect.
This Idea can also work for other monitors that emit ordinary light (CRTs, plasma, OLEDs, etc). We would place a "sliced polarizer" over the surface of the monitor screen. It will have the effect of only allowing polarized light through it, very similarly to that which LCDs generate by design. The downside is that total light output can be significantly reduced, and also that exact positioning of the polarizer on the screen is very important, if we want the slices of the two-image video signal to be properly aligned with it.
IBM press release
This was when LCD monitors finally began to approach CRTs in resolution. [Vernon, Jan 08 2009]
How LCDs work
As mentioned in the main text [Vernon, Jan 08 2009]
Blurred Display Idea
As mentioned in the main text [Vernon, Jan 08 2009]
A variety of 3D display techniques
It should be easy to see that various aspects of this Idea are already out there. [Vernon, Jan 08 2009]
Likely THREE sets of vertical slices!
This variation of Sharp's 3D display (with prisms at the surface of the screen) allows 3 separate images to be seen, depending on the direction from which one is looking at it. [Vernon, Jan 08 2009]
3D LCD display
Not the same process. [phoenix, Jan 08 2009]
slanted lenticular lens on screen
there was a more recent release around jun/jul last year. [4whom, Jan 09 2009]
epson last year
[4whom, Jan 09 2009]
In the oven...
This implentation apparently wants to adjust pixel polarization individually, and doesn't quite make it. Perhaps a good reason to prefer twice as many pixels along one axis of the monitor, as described in the main text here. [Vernon, Jun 08 2009]
Looks very good, too... [cowtamer, Jun 10 2009]
Definitely in production now
So far none are using double-resolution, though. This LG model, however, doubles the scan rate in order to compensate. I don't know that LCD TV's can actually handle that fast of a rate for pixel-changes, though, so I'm suspicious. That is, I think the behind-the-scenes tech is workable in theory, but the up-front part, the actual screen, may not be good enough for it to work as well as Theory indicates it should. [Vernon, Mar 27 2012]
This is more like it!
Pretty much what I've been wanting for a longish time. Now to figure out how to afford one.... :) [Vernon, May 31 2016]
displays that are better than vision
[beanangel, Jun 04 2016]
||Welcome back [Vernon], I like the idea**. Sounds like it should be baked (seems like a nice solution).
||An immediate query that springs to mind is that I am wondering what the image to be projected would look like. By that, I mean the actual video stream (assuming we're talking digital video) and how the L and R streams would be mixed* and decoded and all that stuff. I'm just pondering out loud.
||I don't understand how this is different from the current polarizing glasses solution.
||[Jinbish], it would probably be most efficient if the two images were generated and processed separately, and then sliced up, in the penultimate stage, and then combined in the last stage for sending to the screen. I have a suspicion that horizontal interlacing could work as well as the vertical interlacing described in the main text, provided the monitor had appropriate high resolution in that dimension, for slicing.
||[phoenix], the 4th link here describes using polarizing glasses in theaters, where two projectors project two polarized-light images onto the normal reflective screen. Here the polarized light is an output from the screen.
||Seems like it would be MUCH easier to use a twin DLP system? Probably cheaper, brighter and higher resolution too.
||[MisterQED], a single DLP mirror might be fine, if it was used in conjunction with lasers (one for each primary color) that could have their polarizations changed quickly on the fly. Laser light is always polarized light. So, as each laser scans a row of DLP mirrors, you would want to switch polarization from one direction to the opposite, as each beam goes from one mirror to the next.
||On the other hand, of course, two DLP systems using lasers and projecting onto the same screen would be easy, if the lasers in one system had the opposite polarization of the lasers in the other DLP system. But then, we would practically be mimic-ing the thing done in 3D movie theaters, and then this Idea wouldn't be so original...
||Would a CRT with extra electron guns work? It would be interesting to try.
||I like this idea, although probably better to have alternating pixels diagonally rather than rows to give better visualisation
||[cowtamer], yes, that link does appear to be very similar to my descriptions here. I see they are using the horizontal interlacing approach, but I also see they are not doing one thing I specified, which was to double the number of pixels in the appropriate dimension. According to the brochure, the monitor resolution is 1920x1200, typical for that monitor size. If they are using alternate horizontal lines for two images (one for each eye), then each eye only gets 600 horizontal rows, rather coarse (remember the days of 800x600 resolution). They need a monitor resoultion of 1920x2400, with horizontal interlacing, for each eye to see 1920x1200.
||Something that some movie theatre 3D glasses use, that this could benefit from (though I'm not sure of the implementation) is circular polarisation, instead of linear. It means you don't have to keep your head/glasses perfectly aligned with the screen (about the screen normal axis).
That said, this idea seems to simply be "use double resolution in 1 direction to give a full-res 3D display". I thought (possibly incorrectly...) that the TVs that use LCD shutter glasses used the whole screen res, just at a faster 'flicker'?
||I read that there was a handyphone app that was able to use a high resolution (perhaps Retina display like) display to make it so people with blurry vision saw better images. I imagine that astigmatism (doubled images) could be addressed with partial forms on the retina display, the thing is to have an IR laser view the shape of the cornea (just like PRK automechanisms) then figure out the persons visual acuity then adjust the retina display to compensate producing higher than natural resolution at handyphone displays, then of course the same thing on computer monitors
||that is a slightly laseriffic version of the MIT thing [link]