3-D TV

Ahhh - the bliss of a good night in front of the 3-D box

3-D broadcast of a basketball match - in Japan, and really field sequential, not anaglyph

Home Cinema gets a new meaning with 3-D projection

Introduction - Current TV Technology

The Golden nugget of 3-D display and distribution must be 3-D TV. Because of its immense reach, 3-D on TV should be able to bring 3-D to the masses and liberate it from its non-distribution imprisonment. But TV has serious technical limitations and this is also why there is no or hardly any 3-D on TV.
What does television mean, technically speaking? It means you have to transmit a signal that is compatible with the methods in use since 1953; namely the NTSC, PAL and SECAM interlaced image encoding through use of the Y Cb Cr algorithm. Plain and short this means even/odd scanlines (fields that make up a frame) at a rate of either 30 frames per second (coming down to 60 fields per second) - with NTSC - or 25 frames per second (50 fields)- with PAL and SECAM. Satellite, cable and other digital platforms all use MPEG-2 encoding, which is built upon this technology and does not improve any part of the technical equation. The new HD standard uses Mpeg-4 encoding, which in itself is better to retain the purity of anaglyph colour information, but because of the transmission with broadcast TV and the cabling of HD-DVD and Blue-Ray DVD players it still all goes wrong.

The world television audience is lazy and cheap by definition (bear in mind that most people don't even change out-of-the-box TV settings or even know how to), so you'll have to start thinking about a way of showing 3-D stereoscopic images without complicated or expensive extra's and procedures - like electronic glasses.

So that's the playing field. Now this does leave some options open for consideration. Let's look at some of the available 3D encoding techniques and their employability with current TV technology. Again, realism dictates that your ultimate choice depends on your means and goals; the bigger the audience, the cheaper the glasses need to be and the longer the film lasts, the better the 3D needs to be to prevent a massively headache with your audience.

The latest technology

But what about the latest Philips, Samsung and Mitchubishi 3-D (DLP) TV sets? It will be a while before they penetrate the average Eropean and American living rooms. If the 3-D revolution really takes off beyond 2010, it will be another 10 years before new technology such as native 3-D TV truely reaches the global living room - or even just the western world. What's worse, Philips and Samsung have just pulled out of the 3-d TV market again and Mitsubishi sets are not available in Europe. Hitachi has then just entered the 3-D TV market, but without proper 3-D broadcast content, sales will remain very steadily low and unprofitable for TV manufacturing giants. It's a bit of a chicken and egg situation, and right now the chicken is not laying enough eggs.

'Regular' HD sets ('HD Ready', not true 1080p HD) are estimated to penetrate 60% of living rooms in the Europe by 2011, compared to 8% in 2007. True 1080p TV sets will be seen in 30% of European living rooms by 2012. So when are all these people going to get rid of their brand-spanking new televisions to buy new 3-D-ready TV sets? It will clearly take time, so if you're wanting to produce 3-D for broadcast TV right now, read on below about the current options.

If you're planning for future 3-D broadcast TV production (something that will work and reach enough households in 2025), or if you're planning to produce for a small number of special 3-D TVs, the Philips and Samsung screens are indeed ready to display your Stereoscopic ideas - with or without the glasses! The best bet, what ever you do, is to shoot two separate videos for left and right and decide in post how it's going to be presented (lenticular or field/frame sequential).

Regarding the 3-D TVs that do exist right now: they are ideal for special presentations, advertising in public spaces and other promotional use. The freeview models work with Lenticular presentation (see below) and some of the newest screens work with Shutterglasses and Polarized glasses (again, see below).

Polarized 3-D on TV	Anaglyphic/ Colorcode 3-D on TV	Chromadepth 3-D on TV	Pulfrich 3-D on TV	Shutterglass 3-D on TV	Lenticular 3-D on TV

Polarized 3-D on TV

The working principle of polarizing light

A setup with two identical televisions and a semi-transparent mirror can be used, but the source (the 3-D on tape or from broadcast) has to be split or a special splitter box has to split side-by-side or field-sequentially encoded 3-D imagery.

Polarized 3-D is best known for the big theme park films like 'Terminator 3-D', 'The Amazing Adventures of Spiderman' or 'Captain EO' (That one was replaced by 'Honey, I shrunk the Audience' in 1999) and, in DLP enabled cinemas, 'Chicken Little'. It is being used in these parks and DLP cinemas because it produces the best possible results in 3-D.

Besides the clunky solution illustrated below, 3-D Flatscreen Plasma TVs are now available that use two screens behind each other, with sheet polarizers, turning on and off at high framerates to give the glasses-wearing viewer a smooth and spectacular 3-D experience.

Polarized 3-D TV setup

Same principle, more elegant solution

Anaglyphic 3-D on TV Then there's anaglyphic stereoscopics. Like ColorCode, the only thing needed for this display method of stereoscopics is -besides the very cheap glasses- correct red and green (or yellow and blue for ColorCode) colour representation. This is where television fails to deliver because TV just can't display colours at full strength. As a result, the red images can be seen through the blue lens and vice versa. This results in ghosting and a 3-D image that is not separating the left and right eye images properly. Computer monitors, however, do work with anaglyphics and so do beamers with VGA input. Film works perfectly with anaglyphics, too, but we are talking about the application for TV here. Many distributors or 3-D films and shorts have tried encoding anaglyphics for video and DVD, but time and again this has resulted in disappointment with its consumers - and although the sale will have been made, 3-D will yet again get a bit more of a bad reputation. More on the technical limitations of TV in relation to colour representation and anaglyphic 3-D Film, video, and computer-generated imagery (CGI) all start with red, green, and blue (RGB ) intensity components. In video and computer graphics, a nonlinear transfer function is applied to RGB intensities to give gamma corrected R’G’B’. This is the native colour representation of video cameras, computer monitors, video monitors, and television. The human visual system has poor colour acuity. If R’G’B’ is transformed into luma and chroma, then colour detail can be discarded without the viewer noticing. This enables a substantial saving in data capacity ¯ in “bandwidth,” or in storage space. Because studio video equipment has historically operated near the limit of realtime recording, processing, and transmission capabilities, the subsampled Y’Cb Cr 4:2:2 format has been the workhorse of studio video for more than a decade. The disadvantage of 4:2:2 is its lossy compression. Upon conversion from 8-bit R’G’B’ to 8-bit Y’Cb Cr , three-quarters of the available colours are lost. Upon 4:2:2 subsampling, half the colour detail is discarded. Of the 16.7 million colours available in studio R’G’B’, only about 2.75 million are available in Y’C B C R . If R’G’B’ is transcoded to Y’Cb Cr , then transcoded back to R’G’B’, the resulting R’G’B’ can’t have any more than 2.75 million colours!
RGB Image	TV Image
DVD with its MPEG-2 encoding is completely based upon the Y' Cb Cr television signal and thus does not improve the situation one bit. Digital television transmission, like digital cable, digital satellite and digital antenna (freeview) all use MPEG-2 encoding - making anaglyphic broadcasts a problem. As you may have guessed, because most codecs used in computer video formats like AVI, Quicktime, RealVideo, Windows Media, DIVX and others are based on MPEG-like saving of colors, the same problem applies to these digital video formats. As a rule of thumb, the codecs that result in the largest file size are the ones that save most of the original colour values - and thus enable anaglyphics to work better. The latest versions of codecs like DIVX, XVID, 3IVX and H264 are more colour true and can work reasonably well with anaglyphic content. Again, the higher the bitrate, the better the results. MPEG-4 can work with anaglyphs as long as there is a true RGB component monitor connected to the playback device, where a computer monitor connected to a computer and a computer monitor connected to an MPEG-4 enabled DVD player will work, but not a television screen connected to a computer or a DVD player. This is, because YCbCr colour encoding and transport is used.

Solutions To avoid the problems with bad colour representation on TV, a similar setup to the polarizing two-TV one could be built. Again, willingness of your TV audience to do this and the need for split 3-D image encoding and decoding is needed. Other possible solutions include modifying the broadcast signal so that the available bandwidth or the analogue signal is used not for a Y'Cb Cr composition, but Y'b Y'r. You'll need the cooperation or the broadcaster for this and build a device to modulate the television signal. Also, digital and satellite broadcast remove this special advantage and turn the modulated signal back into the old problematic TV signal. Yet another solution is the shutterglass use of the TV signal to transmit the red and blue signal in field sequential order. This produces a heavy amount of flickering at 25 fps, though, and is a serious threat to epileptics and aesthetics.

Anaglyphic 3-D TV setup as suggested by Adam Ross

Chromadepth 3-D on TV

Chromadepth is a system that works with difference in hue; the visual spectrum- with colours running from red to blue.
Red objects appear in the front, yellow and green in the middle and blue ones in the back. This results in a coloured image in the strictest sense, but it's really monochrome when the colour is fixed to its position in depth.

When presenting moving images in Chromadepth, this results in something that resembles an LSD trip. If you can live with this colour design, then ChomaDepth is probably the best system for cheap, accessible 3-D television production.

(Click for Video example)

The transparant ChromaDepth glasses work like a spectrum-separating lens

Pulfrich 3-D on TV

The Pulfrich system is quite an interesting solution. It works just fine, as performed on a broadcast of '3rd Rock from the Sun' by NBC and 'Shark Week' on Discovery Channel. The problem is, as is indigenous with Pulfrich 3-D, the camera has to keep on moving for the effect to work. In short, it works on the combined premise of delayed image transfer to the brain -the eye with the darkened glass- and horizontal parallax -moving sideways, one eye will see equal imagery later than the other eye, creating the illusion of depth.

(Click for Video example)

Shutterglasses / Field-Sequential 3-D on TV

When people thought of the future of 3-D in the 80's and 90's, they thought of shutterglasses. Because they look futuristic and because the IMAX uses them for some of their cinemas (the others that are 3-D enabled use polarized glasses).

The shutterglass - or VR - system is also a great option for 3-D imagery on television. But your audience has to be willing to spend the money on expensive VR goggles or relatively less expensive shutterglasses. Tthat means one pair for every family member.

The new Samsung and Mitsubishi 3-D DLP TV sets use frame sequential 3-D encoding on high framerates of 60 fps - that's 60 progressive frames per second (120 Hz). The same field/frame sequential glasses as those used from the 80's onward are compatible with these new 3-D TV sets.

Field/frame sequential glases / VR goggles have an LCD display for each eye, and the shutterglasses work by opening and closing the left and right eye glass in sync with the field interlacing a television does (field sequential). This means that the viewer will see a substantial flicker - somewhat less on NTSC than on PAL or SECAM because NTSC works on 60 fields per second, while PAL and SECAM work on 50 fields per second. 100 Hz televisions do not work with the shutterglass system, nor do plain HDTVs.

The History of shutterglasses As early as 1923, shutterglass systems have been used. This first version, called the Teleview 3-D system, may have been electrical/ mechanic rather than electronic only, but its results were equally good. The flicker caused by this sytem is as good or bad as it is now in IMAX cinemas. The viewer units were cetainly more hygenic because they never touched the audience's face. Today's shutterglasses need to be cleaned after every performance. 1923 was the first of the 3-D boom years. It took another 30 years -1953- another 30 years -1983- and yet another 30 years -2003- to get to today's 3-D renewed wave of popularity. Most cinemas in the western world will install 3-D enabled DLP projectors before the end of the decade. This means a constant possibility for 3-D films to be shown in regular cinemas. For home-cinema use this is an irrelevant fact because cinema systems are technically totally unrelated to television systems.

Lenticular 3-D on TV

Ideally, a 3-D television presentation would be one where the viewer does not wear glasses at all. In other word a freeview solution. This sort of a television exists and it employs the lenticular method.

The surface of this television contains a specially ribbed lens that allow the left and right eye to see different angles of the image on the screen, which is parted into lots of small strips to match the ribs of the lens on the screen. This surface can also be a plastic sheet add-on.

The resolution of the screen will drop dramatically, because of the left and right image next to each other (even/odd vertical lines) on one screen. For example, a 1024x768 image on a computer monitor will be reduced to 438x256 (less than 1/4 of the screen). So one will need a screen of very high resolution to maintain normal resolution 3-D imagery. Standard Definition TV currently works with a resolution of 720 x 480 for NTSC and 720 x 576 for PAL and SECAM. True HDTV increases the horizontal resolution to 1920x1080, resulting in a final image of 820x360.

In other words: freeview plasma 3-D TVs can result in a bit of a blurry image.




Fresnel lenslets

Another version on this theme implies the use of a fresnel lens, dispersing the TV or computer monitor information by directing left eye image information to the left eye and right eye information to the right eye by means of positioning the viewer in a exact position in front of the TV screen (and, indeed, the fresnel lens). Again, a low resolution 3-D image will result as a cause of the low image resolution that all TVs start off with. The fresnel lens is a very popular item with manufacturors and inventors of autostereoscopic 3-D displays because of its light bending properties.

Fresnel lenslets	The manner in which the fresnel lenslets refract incoming and outgoing light
Example of a 3-D image of a pumpkin as seen through a fresnel lens slab	Left eye and right eye see different image portions

Holographic 3-D TV

For people who've seen holographic technology featuring in Star Trek, Star Wars, Logan's Run, Minority Report, A.I. and countless other movies there's an important reminder of the kind of films they are: science fiction. The theory that a holographic imaging system may become reality in 100 years time doesn't mean its technology is actually around the corner right now. True, there are ongoing experiments with electron beams that make the electrons flare up at a certain point in space, with a certain colour, but sticking your hand in such a beam would mean definite electrocution. Besides, these are highly limited, theoretical lab tests.

Another kind of freeview 3-D technique is the spinning surface approach. By projecting onto a spinning surface with multiple viewing angles, a walk-around image can be generated in the containing cube. Again, this technique is still in its very infancy and practically speaking, is totally unrelated to Television. A future application of this technique will look like a big glass cube in the middle of your living room, requiring the complete bandwidth of your current satellite, cable or digital antenna to transmit one broadcast channel. The end result will be a bit like miniature theatre in your living room.

Also, reports in newspapers of holographics using 'dry ice' as a means of displaying a 3-D image are highly exaggerated and often journalists hungry for news wrongly interpret these experiments as the arrival of the future. A practical application of dry ice in our living rooms or in a cinema theatre is difficult to imagine.

A third common misunderstanding is that the technology behind picture holographics can be used to produce a holographic 3-D movie. The technological restraints of the production of a holographic print actually make this completely impossible with current technology. Required bandwidth for holographic display is so insanely big, that current and next generation technology is not able to transmit a moving holographic image larger than a pinprick.

The well marketed DepthCube As you can see it will take more than just 2 cameras to shoot for this puppy

As you can see, making a hologram requires perfect lab conditions, while displaying it on TV would require an insane amount of bandwidth

Implications of making Holographic Video

One has to make a clear distinction between stereo 3-D and holographic 3-D imagery. They are different mediums. Why? Because holographic 3-D (including the 360-Degree Light Field Display imagery) has no framing, as opposed to stereo 3-D.
3-D is filmed with two cameras, while holographics are recorded with light hitting an object from a very wide array of angles. A real or virtual actor on these displays will be looking in one direction, while you, the viewer, may be looking at his side, his front or even his back. So there is no controllable eye-line and your fellow-viewer will always see a different image from you. In this way, cinematography cannot be employed for these displays; rather these moving images have to be approached as you would theatre or a stage show. This is a very important distinction to make, and once you recognise this, you can create the best possible entertainment for these displays rather than doing something that equals a ‘radio show on TV’.

Contact 3-D Revolution Productions


Tel +44 1179 441 449

3-D Revolution Productions | 51 Sefton Park Road | Bristol BS7 9AN | United Kingdom

© 2009 3-D Revolution Productions