The recent Dimension3 stereovision conference in Paris was a meeting of doers and market makers, not hobbyists or fans. Jon Peddie assesses the state of the business.
By Jon Peddie
The Dimension3 conference in Paris covers everything being done in stereovision on the PC, TV, mobile, and cinema. Unlike many S3D conferences, this brings together the doers rather than the fans and hobbyists, and it has a distinctly international flavor. 
For three days in three tracks leaders in all disciplines of S3D showed their work, described the technical issues, and gave their forecasts.
The cinema is the leader in S3D. In the U.S. there are more box office releases of 3D AA titles than there are of 2D titles. S3D is also making more money at the box office, which of course has everyone scrambling to get in on the ride and creates the risk of a lot of bad S3D movies being made and giving the technology a bad name.
The PC, games in particular, is the second most popular and successful S3D platform, followed by 3DTV, signage, and mobile. Mobile will unquestionably become the most popular viewing device for S3D because of its glasses–free display, intimacy and immediacy. But like the PC (for most situations) it will be a solitary experience.
3DTV is a long way from getting any kind of a transmission standard. As a result some companies, like SkyB, and ESPN have just gone ahead and put together their own systems. There are choices of over and under, side by side and alternate frame transmission schemes.
Glasses a red herring
The issue of glasses still haunts the technology, and it’s a red herring. In recent and repeated surveys users were asked if glasses were a problem and always the majority said no. They seem to be a problem for reporters and analysts trying to come with something new to talk about.
Nonetheless glasses-free efforts continue. The choices (like almost everything in S3D) are plentiful. There’s multi-view where eight to 22 cameras are used and displayed on a lenticular screen. There are parallax barrier filters, mostly used and useful for smaller screens (although we know of one company in Silicon Valley that is experimenting using them in large screen signage application). You can get eye tracking, but it’s a single user experience, and there are light-field quasi-holographic approaches that use special screens.
In the glasses arena there are multiple choices, including polarized, shutter, or anaglyph. Polarized is limited to plasma screens and cuts the resolution in half (horizontally). There are multiple suppliers of shutter glasses and no standard for how they sync up with the screen, so Nvidia glasses won’t work on an HP Envy 17 3D, or a LG TV, and vice versa. This is probably one of the major road blocks to getting widespread adoption of S3D in the home.
Capture
From a content creation aspect new S3D cameras are coming out every week. The Nintendo 3DS has dual cameras on the back side, and we think mobile phones are next with either dual cameras on the back side, or a 3D sensor, or panoptic lens array, and probably all combinations of them.
From a broadcast and cinema point of view the jury is still out what is the best camera setup—above (or below) and straight, side by side, or some combination with half mirrors and/or prisms. There is also the issue of convergence—where should the focal plane be? Directors want to have it variable and determined by them, which is easy but time consuming.
And there are some situations where a multi-camera rig just can’t (or shouldn’t) go—imagine a 22-camera array in an underwater scene.
More 3D content please
Content is and always will be king and here S3D has opportunities and challenges. The cinema has primarily relied on 3D capture, with Tim Burton’s Alice in Wonderland being a notable exception of 2D to 3D conversion. Much more 2D-3D conversion of previously released films will take place in the coming months and years. The techniques for doing it successfully are known now, and yet there will be cheap hacks. As one of the speakers at Dimension3 said, we don’t have automatic colorization or automatic editing in film, why should we expect to have automatic 2D/3D conversion?
The other two platforms that will make great use of 2D to 3D conversion are signage and 3DTV. Slimmed down versions for mobile devices will also use such content. PCs don’t need it since games are designed and built in real 3D and so extracting the S3D aspects is algorithmic.
Slow down
When considering S3D it is important to remember it is a trick to fool the brain into thinking it is seeing 3D. We use 30% of our brain to process visual information. That means S3D content has to move a little slower if we’re going to get it; it’s like listening to a foreign language, for most of us we need the speaker to talk slower so we can translate; in the case of S3D our brains need time to translate the sensation. The sensation of S3D is analogous to the sensation of physcoacoustic sound that tricks you into thinking you are hearing five speakers through headphones.
S3D isn’t only about entertainment, movies, TV, games, adverts, sports and concerts, etc. S3D, especially on a computer, was in use in scientific applications long before it was adapted to entertainment. GIS applications make great use of S3D as do CAD, molecular modeling (one of the first S3D applications), and MRI to name a few.
This is the fourth wave in visual entertainment technology, going from black and white moving pictures, to moving pictures with sound and then to color (some might want to include widescreen as one of the steps but that was not a major technological achievement). Resisting this development and finding excuses or reasons to disprove it is the way of the Luddite.
What’s next?
To further development on glasses-free concepts, a standardization committee has been formed to standardize the IR signal for glasses. Transmission standards are being hammered out, albeit slowly. Cinema can’t decide on active vs. passive glasses so both are being offered and its unlikely a standard will emerge in theaters for three to five years. Signage suppliers seem to have decided on lenticular, while handhelds seem to be going for parallax barrier, and 2D- to-3D conversion has not only become acceptable but techniques are being refined and classics like Star Wars are being converted.
Five years from now all active glasses will be compatible. Passive glasses for 3D TV will slip into a minority position due to their resolution penalty, and yet probably become the de facto standard for cinema, and we’ll still be talking about a TV standard (remember how long it took to get the miserable ATSC specification in place?). Mobile will have standardized on parallax barrier without it being an official standard and head-tracking will be a challenging second choice. Signage scientific, and engineering will continue to use active shutter, and most likely RF-based rather than IR. Projectors will abandon the resolution, cutting checkerboard and go to 120 to 240 Hz with active shutter glasses. Blu Ray and HDMI 1.4 will be the only hard medium solution and use active shutter glasses.
Augmented reality will lean toward head tracking solutions for mobile and shutter glasses for PC (entertainment, industrial, and science, medicine, and engineering). There will be no universal all-platform solution.
What we think
If you listened to the pundits and reporters, people not actually involved with creation or selling of S3D, you’d think the most important issue was glasses and lenticular lenses were the best solution. The general public has little appreciation for content creation, mechanical and transmission issues of S3D, and are being misdirected by the wannabe experts into looking at the lens and not the image.
Criticizing glasses is like criticizing headphones to listen to music. Both are media delivery systems and like psycho-acoustic sound our brain is being tricked into seeing something that isn’t really there. The focus, no pun intended, needs to be taken off the apparatus and applied to content—is S3D making the story more interesting, more compelling, and more enjoyable?
The Dimension3 conference helps the actual participants in the industry deal with those issues. At the conference the presenters and exhibitors who are making their living solving these problems and bringing us enjoyable content come together to share their findings and developments. It’s the place to be if you want to know what’s going on and what’s coming in S3D.
Dr. Jon Peddie is president of Jon Peddie Research, the publisher of GraphicSpeak and TechWatch.
