I am here in Berlin to attend IFA, but also to attend a small 3D event focused on stereoscopic 3D (S-3D). So far, I have learned some interesting points and seen some interesting demos. I am becoming more and more bullish on the 3D markets, but the show has also reminded me how difficult doing 3D can really be and the dangers the industry faces if it does not do it right this time.

Chris Chinnock
Senior Analyst and Editor
for Insight Media

First the bullish part. Big players are starting to realize that 3D technology is maturing and that viable applications and market opportunities exist. Texas Instruments is using the upcoming CEDIA trade show to highlight its 3D DLP rear projection TVs to the home theater crowd. That’s pretty mainstream. At our upcoming 3D BizEx event, just look at who is speaking, sponsoring or supporting the event: Sony, Mitsubishi, Philips, Dolby, RealD, Paramount Pictures and Dreamworks Animation. These are important and major players. 3D digital cinema is here, TV is coming and a host of consumer-oriented 3D products are likely in the next few years. 3D may indeed be the next big wave in consumer electronics.

That’s the bullish part, now for the bear’s view. First and foremost is the danger that we will see another 3D "bubble" that will flounder on poor implementations amid high expectations. This happened in the 1950’s in cinema as Hollywood rushed to embrace 3D only to see it abandoned very quickly due to shortcomings in the technology and resultant image quality.

Today’s 3D cinema technology is vastly superior and image quality is quite good. But creating a good 3D movie takes a lot of effort and special skills to add just the right dimensionalization on a scene-by-scene basis. There are skilled people to do this, but there are also more than a dozen studios gearing up to do 3D movies. Is there enough talent and training available to ensure these projects produce high quality content? A bad 3D implementation of a first run movie can have a disproportionate impact on the impression it creates with the viewing public about 3D.

Another area of concern is content conversion from 2D to 3D. Games and other computer generated graphics content with a 3D database behind them could fare well in conversion to stereoscopic 3D. But so far, gaming content developers have merely converted 2D games to 3D games. What is needed are 3D games that take advantage of the ability to hide or reveal objects that may not be visible in a 2D version. This will create a real reason to own a 3D game.

Converting 2D still images to S-3D is not too hard, but converting 2D video is tough. Doing it offline where professionals can adjust and tweak is the preferred, and expensive, approach, but there is a huge pull to do it automatically in real time. Most of real time demos I have seen of 2D to 3D conversion have not been very compelling. In fact, some are not good at all. We need to be careful in how fast we roll out these solutions so as not to create bad impressions of 3D that will take years to reverse.

And let’s not forget the hardware implementation of 3D display systems. Even at trade shows dedicated to 3D, I have seen demos that are of poor quality or even set up incorrectly. If the people who are trying to sell 3D can’t configure it properly or create compelling demos, that’s a problem.

Case in point. Autostereoscopic 3D displays are ones that do not require glasses to see the 3D effect. To do this, the technology requires that you trade off image resolution in order to enable multiple viewing zones across a fairly wide field of view. One clear lesson with using such displays is to limit content to low resolution images such as icons or larger graphic elements.

At S-3D, I saw one demo that was showing SD resolution video on an autostereoscopic display. As expected, the video was so compromised it looked like it was out of focus. Also, the viewing zones were so narrow that it was difficult to find and keep the image in full stereo.

Other demos showed large rainbow patterns and similar difficulties in visually acquiring the image. Another common mistake is to reverse the left and right images when coupled to the polarization filtering glasses. This creates a stereo image, but it looks funny and will create eye strain. How can manufacturers ensure this doesn’t happen? There are no standards or methods that I know of.

And, 3D needs to recreate, as much as possible, the way we see the world. You can not see stereo pairs when looking at objects beyond 50 feet or so, so don’t try to add dimension to these long distance shots - it looks wrong. And, when moving your head laterally around a stereo display, don’t maintain the same object orientation as you move. That’s not how it works in the real world. Finally, making objects jump out at you may work in a theme park 3D experience, but not if you want to use the 3D display for extended periods.

The bottom line: while we have to solve the technology part, we can’t sell the technology to the consumer. It’s about the application. Let’s stop being obsessed with the technology and focus on making the applications for the technology work. Once it is easy to use and offers a clear benefit over 2D, 3D will be adopted. But let’s not be too over-anxious to roll out 3D systems either. Bad implementations create a poor impression and a backlash that could take years, maybe decades to reverse.

Reply to the author