Dec 6, 2010


I’m sure some of you are contemplating upgrading your tiny old TV to a nice big 3DTV this Christmas. But with all the different goodies available, it is no wonder that ordinary folks get confused. Even the experts don't always agree on what is what. Personally, I have worked in a somewhat related field for a number of years and I still have a hard time figuring out what the various competing manufacturers are talking about. One thing is certain, 3D is here to stay. LCD and LED display manufacturers have been ramping up research and production. They are churning out 3D screens like there is no tomorrow and anything that is not labeled “3D” is so last year! Let's dive into the subject a little bit shall we.

What defines 3D? Well, what defines 2D? Already here I need to apologize for probably stating the obvious, but to some people this bit of information might have gotten lost in translation. Speaking of information, let's assume that we are looking at a regular computer screen. The height and width would correspond with two 2 numbers. Those are the number of pixels that define the resolution of your screen. A standard XGA (VESA standard) display would have 1024 (width) x 768 (height) pixels. If you are good with numbers, you will realize that the relation between those numbers is 4:3. This is your aspect ratio or the relation between width and height! For the longest time this aspect ratio was common in ordinary CRT (Cathode Ray Tube) TV’s. You may already have upgraded to a flat panel TV years ago which would have meant that your old TV probably would have ended up like this.

If you were really fancy, you may have upgraded from standard definition video to a cool 1080 HDTV! To some people 1080 is just a number. However, it is in fact one of the many different HD formats. The native resolution of your awesome 1080 HDTV is 1920 x 1080! 1920 pixels wide x 1080 pixels high! Guess what the relation between those two numbers is? YES! 16:9. By the way, like the CRT screen, the 4:3 aspect ratio is gradually being phased out and almost completely replaced by 16:9. Anyway, you don’t have to remember all these numbers. I am merely trying to prove a point.

Now, back to 3D. Just for giggles, let's introduce a 3rd dimension to our XGA (1024x768) equation. This is when we venture in to the subject of stereoscopics. For a truly volumetric display, viewable 360 degrees, we would have to work with volumetric pixels or "voxels" Let's make it easy and say that our work space is standard XGA and that the image depth or 3rd dimension, is equivalent to the image height. The resolution of the image would then be 1024 x 768 x 768. The result would be slightly less than 604 million voxels or 768 x 8 megapixel images stacked on top of each other 60 times every second. Now do the same calculation in HD at 120 Hz. Of course the technology to accurately display what can only be described as a hologram at that resolution, hasn’t been invented yet. While we wait for science to figure that one out, let’s just try and create some depth.

Luckily, we really only need 2 angles to create depth in a 3D image and today, even the best systems only process frame sequential at 240 Hz. That means left and right frames sequentially 240 times a second. Even at that speed it is significantly less information than the huge amount of voxels needed for full volumetric view. Even so, there is still a lot of processing going on.

For years the industry has been developing processors that were increasingly better at crunching large numbers very, very fast while, at the same time, slicing and dicing the vast amount of information so that we could replicate the 3rd dimension that we have been capable of seeing instinctively with our naked eyes for thousands of years.

Things get even crazier when we start talking about stereoscopic vs. auto-stereoscopic. First of all, let’s define the terms that even the experts don’t quite agree on. It is commonly determined that a stereoscopic image is exactly that, stereoscopic. It has depth. An autostereoscopic image has depth seen with the un-aided eye. This means without the use of any kind of viewer eyewear and/or screens. Huh?

Let’s face it. Most of us have felt silly wearing 3D movie glasses while silently appreciating the dim light in the theater. Back in the day anaglyphic glasses were common. Those were the ones that had red and green lenses and they were based on a color separation system that really wasn’t that accurate. These days, it seems that the battle is between Frame Sequential vs. Retarder systems. In theaters, the most common are passive polarized glasses like linear polarized or in rare instances circular polarization glasses. This is the Retarder system for use with the glasses that you just put on and that don’t require batteries. For the home theater market the frame sequential or active polarization system is most common. Those would be the ones that would have an on/off button. There are different approaches. They all have benefits and issues, depending on who you talk to but all of them rely on the same proven idea.

Whether it’s one or the other system, content is generally produced in the same manner. Video is shot with a stereoscopic camera. Essentially two synced cameras mounted side by side with approximately the same distance between them as there is between your eyes. Some of them even have the ability to converge like eyes. By the way, if you are a camera geek, check the link for the camera.

When playing back the content in the movie theater, you will see both left and right at the same time but one side is polarized and the other is not. If you ever took a sneak peek at what that looks like without the 3D glasses you would think that the projector needed a serious tune up. However, when wearing the glasses your left eye sees what is intended for your left eye and the right sees what is intended…. Well, you get the idea! Our brain then makes up for the fact that we are in fact looking at a flat surface.

In that context, it is interesting to note that our stereoscopic field of vision is narrower than our overall, or 2D, field of vision.

(Footnote: Some people may experience visual discomfort. This comes from the fact that we are trying to focus on something that appears if front of the screen and on the screen at the same time. It can be very confusing and has been known to trigger photosensitive epilepsy attacks.)

While we are at the subject of visual discomfort, I want to mention one of the guru's in the field of stereoscopy. I know, I have mentioned him before in one of my previous blogs, but Michael Halle is kind of a household name when the subject is 3D. In his paper "Autostereoscopic displays and computer graphics" he writes about “window violations”:

"Photons must originate in, or be redirected by, some material. The material can be behind, in front of, or within the space of the image, but it must be present. All claims to the contrary violate what we understand about the world. Figure 1 shows the possible relationships between the image and the display. A corollary to this constraint is the observation that air, water, or smoke are, in general, very poor display media. Images appearing "in mid-air", called aerial images, will invariably have originated not in the air from some other medium. Technologies lavished with claims of mid-air projection should always be scrutinized with regard to the fundamental laws of physics.

A specific and practical result of the projection constraint is that no matter where a spatial image appears with respect to its display, the image will be clipped by the display's physical boundaries. If for instance, an image appears in front of its display, a sufficient translation of the viewer will cause part or the entire object to intersect and "fall off" the edge of the display. This condition, known as a window violation, is particularly disturbing for aerial images. Figure 2 illustrates a window violation.”


But I digress. Auto stereoscopy, the ability to see 3D with the un-aided eye, has been a challenge for a very long time. Already in the 1920’s Scottish engineer John Logie Baird, who is widely known as the inventor of the television set, first introduced the concept of the parallax barrier. The idea was to create a multipoint viewable 3D experience and the parallax barrier was the very first attempt.

Surprisingly, it did actually work. It probably wasn’t pretty, but it was a start. Mr. Baird was way ahead of his time and I wonder if the world just wasn’t ready for it! Since then there have been many attempts to create other and better solutions. The most recent is a lenticular solution that deals with horizontal ego-motion only, as opposed to covering both vertical and horizontal. Ego-motion is the ability to look around objects in a 3D image when you transition from side to side or move your head. In reality, objects that are closer to us travel at a different rate than objects that are farther away. This obviously applies to both vertical and horizontal head movement. However, when trying to create an auto-stereoscopic effect on a display system, it is much more practical to disregard the vertical ego-motion entirely.

In general, lenticular screens have 9 horizontal angles over the width of the screen. One manufacturer claims to have a whopping 64. Usually, it doubles with each generation of prototyping. I have seen lenticular screens with 18 angles and the technology is definitely getting there. Obviously, with each step towards perfection, the industry will have to deconstruct, decode and reconstruct content to make it work on the fancy new screen systems. More numbers to crunch….

So, there you have it. Currently the consumer market seems to settle on eye wear 3D while the business to business markets seem to push auto stereoscopy.
Exactly like the projector market, it is only going to be a matter of time before we will see a cross over to the consumer market. Of course, the auto stereoscopic displays have their own set of issues. Less brightness, reduced resolution and sweet-spot viewing issues (ego-motion) are among a few.

Generally, the industry is really pushing the idea of 3D hard. Already content providers and networks are delivering content in 3D. Granted, not so often, but major sporting events are transmitted in HD3D. Now, it’s just a question of what standard and which delivery system will survive. Remember VHS and BETAMAX? VHS won that one. Not because it was the best but because it was affordable and available. I am old enough to remember going to the little neighborhood video store (before there was Netflix and YouTube) wondering who would rent BETA tapes when there were literally only 3 choices among hundreds of VHS. Finally, studios stopped releasing movies on BETA all together and that was that.

In the end, this raging evolutionary war will be governed by the old trusted mechanics of supply and demand while competing manufacturers are building the next infrastructure we just can’t live without. So it has been and so it will be in this case. With any new technology, someone will always be first but whichever manufacturer delivers a system that makes content production and distribution affordable and practical will most likely become the standard.

1 comment:

  1. This is interesting. I never really understood why we had to wear the glasses. I think it is interesting that the consumer market is headed in this direction! Thanks for your enlightening post. I am going to pass it along to all my geeky friends! :)