State of the State of 3DTV


Now that 3DTV is inching ever closer into living rooms — whether we want it or not — it’s probably worth taking a verbal snapshot of the state of the state right now, in the afterglow of summer, before the holiday buying season and prior to the 2011 Consumer Electronics Show.

Let’s start with how consumers who have 3DTVs experience 3DTV right now. So far, there aren’t many: Two hands went up when attendees of the recent Mid-America Cable Show, in Kansas City, were asked if anyone owned or planned to own a 3DTV. (A first. It’s usually zero.)

Later that same week, I drove up to CableLabs to talk with David Broberg, cable’s 3DTV technology ombudsman (real title: VP of video technology), about the new 3D encoding specification that came out on Sept. 1.

In his office was a beauty of a Sony 3DTV. Fiddling with the remote, he aptly characterized the 3DTV state of the state so far as “an experiment.” The remote had its own 3D button, for instance, except that it wasn’t the way to the 3D settings. Finding the 3D settings from the TV’s menu got involved enough to want for a paper and pen, to write down the sequence.

It was equal parts amusing (that someone so advanced in class would struggle) and alarming (if it’s hard for him, what about regular people) to see what an ordeal it was to go into and out of 3D mode. (Default 3D mode is “simulated 3D=ON,” by the way, in many 3DTVs.)

The good news is, it’s all about to change for the better. That’s partly because of the new encoding specification and partly due to the emerging set-top landscape, which will send additional HDMI data that tells the 3DTV when and how to automatically activate the correct 3D format.

But why a 3D encoding spec now, if the “frame compatible” method of tucking the left- and right-eye images into the same channel-width as a 2D HD stream is already a foregone conclusion?

Because no two 3D “panelizers” work the same right now. Yes, the over/under and side-by-side encoding methods exist in market. But they work differently enough to create confusion, especially about total bandwidth usage of a 3D stream. Stipulating basics (like a low-pass filter after the sub-sampling process, for advanced readers) will go far in streamlining the amount of bandwidth needed for a 3D stream.

Think about it: Somewhere along the short timeline of 3DTV, a rule of thumb emerged: That a 3DTV signal will fit into the space of a 2D HD stream, but a 1.2 to 1.3 bump in bandwidth is better.

That doesn’t have to be the case, Broberg and others submit. “Panelizing” is smooshing, and sloppy smooshing can cause what’s called “spatial aliasing” - a type of noise. Noise requires more bits to compress. Filter out the noise before it gets to the encoder, use less bits.

So, part of the state of the state of 3DTV, in mid-September 2010, is a clarified bandwidth chapter. What remains to be seen is what consumers put under the tree this holiday season. iPad or 3DTV? We’ll keep you posted.