I spent the better part of last week at the Society for Information Display (SID) conference in Los Angeles. At the opening press briefing, in part of the keynotes, and in many of the booths on the tradeshow floor, the message was that a lot of people were pinning their hopes on 3D as the Next Big Thing in consumer products.
Boiled down, the message was that TV manufacturers want consumers to buy a new television every three years, and game developers are out of ideas for new kinds of games, and are looking to 3D to enhance the user experience in games of the same sort that they are producing now. In fact, they're anticipating bringing back vintage games in 3D, with the hope that customers will pay for a version of Frogger that will put players in the middle of a busy freeway right alongside that hapless frog. (My example; no one was actually that specific about what old content they wanted to upgrade.) Helping both TV makers and game-developers was the way that display technology is blurring the line between passive entertainment boxes like big-screen TVs and interactive game consoles like Kinect.
The Missing Ingredient
That high level summary of some SID talks might make it seem like all one needs is hardware to develop a 3D immersive experience. The final keynote speaker on the first day of the conference demonstrated why that's a simplistic notion.
Phil "Captain 3D" McNally (He legally changed his middle name to "Captain 3D") is a major movie 3D guru, now at DreamWorks. Parts of his talk were illustrated with real 3D animation footage. In his most explicit demonstration of the art, a 5-second clip from a feature-length production, he showed what goes into a real Hollywood (well, Glendale) movie.
The scene was a dolly-in (not a zoom) past an object in the left foreground, on a character in the near background. In a series of outtakes, McNally showed how the director and the animators who control the rendering dealt with certain visual-perception problems where the simple application of 3D could make some viewers feel queasy.
To start with, McNally said, you have to realize that the theater screen represents a plane. As the virtual camera moved in, in the demonstration, the foreground object shifted position from behind the plane to in front of it. McNally showed how this results in a disturbing visual effect if the same stereoscopic point of view is maintained throughout the move.
This was a fairly involved demonstration, but the upshot was that the final scene was optimally rendered when, as the virtual camera moved in, the 3D effect on the foreground object was reduced (i.e., the apparent distance between the stereo lenses on the virtual camera was gradually reduced), while the 3D effect on the character in the mid-ground and the objects behind him was enhanced by "moving" the virtual camera lenses further apart). Essentially, the room got deeper as the camera moved into it.
From the point of view of the film's director, all this has an aesthetic component, related to shifting the viewer's attention from one character to the other, but that was coupled to the desire to avoid a disturbing 3D artifact associated with a fairly elementary camera move.
Let people get used to that in full-length movies, and they're going to want it in their ordinary 3D TV, and ultimately in "Frogger 3D."
On to the Holotank
Talking about routine 3D TV brings up sports, and that's where my yearning for a holotank comes in. (I'm talking about an as-yet fictional display device like a tropical-fish aquarium that can be viewed from any angle to display a true stereoscopic --"stere" is Greek for "solid" -- image from any point-of-view. Nothing I saw at SID resembled a Holotank. The only way I can think of to do one would be to take the concepts of a 3D printer and apply them to multiple lasers that would excite phosphors in some kind of a transparent matrix.)
Why a holotank? You'd think that ordinary 3D would be great for sports, because the extra dimension could provide a more realistic sense of the action than ordinary monovision. But stereo TV cams are all designed with lenses positioned roughly as far apart as the distance between your eyes. Now position those cameras up in the press level of a sports stadium and consider how little information about a play can be extracted from a three-inch separation fifty yards from the action.
That problem is why there have always been multiple TV cameras up in the press level -- so you can see the play from widely separated angles. But you can only look at those views sequentially. You can't integrate them within the proscenium-like range of points-of-view allowed by a flat screen, no matter how well it can handle conventional 3D. As long as we're stuck with that technology, it's like your pants are nailed to your seat.
But with a holotank, on instant replays, you could move your point-of-view through a wide range of solid angles. For every replay, you could place yourself in a different seat in the stadium.
Now that would sell some new TV sets. I can hardly wait. I wonder what we could use for the matrix.