The Stereoscopic Displays & Applications conference turned 21 years old this week. This is an event where cutting edge 3D display and application technologies are discussed and demonstrated. I like this event as there is always a chance to meet old friends and pioneers in 3D and to see where the latest research projects are headed. These usually provide a compass to the direction commercialization will take in the years ahead, so this is the kind of show we like to attend.

Chris Chinnock
Senior Analyst and Editor
for Insight Media

Having so far only attended the first day of this three-day event, my observations should be considered preliminary. Notable in the events this year were several panel discussions. One focused on the standards activity now underway around 3D. It was a good summary from experts in the field covering activities at SMPTE, CableLabs/SCTE, HDMI and CEA.

When asked what worries them, Sony’s Peter Lude, representing SMPTE said, "It’s the little things like menu graphics, handling aspect ratios, etc, that will impact the 3D user experience."

CableLabs’ Dave Broberg is worried about the interoperability issues with set top boxes, AV Receivers and 3DTVs. He noted that cable MSOs do not have to include the two mandatory 3D formats required for HDMI 1.4 compliance as the broadcast industry will use frame compatible Over/Under or Side-by-Side formatting. In order to provide a good user experience, display of menus and icons, channel changing and other operations must work properly — and he is worried that they won’t. That is why he is opening up CableLabs to the industry to come and test their products to help ensure good operation and interoperability over many combinations of connection.

Steve Venuti of HDMI echoed these sentiments, as did Mark Stockfisch of Quantum Data, who represented CEA. Stockfish also expressed concern about the way 3D information will be passed back from sink devices to source devices in the HMDI spec. Since this is handled in the vendor-specific data blocks, it could open the door for some problems. He is concerned as well about the range of 3D image quality that will be delivered to consumers, as was his experience when viewing the numerous 3D demos at CES.

A second panel on Wednesday will focus on successful business models around 3D. These panels show the growing sophistication of this event and a desire to educate its attendees, who are mostly academic researchers and 3D specialists, about the bigger dynamics that are shaping this nascent 3D industry. I applaud this move.

I was also impressed with the first session I attended which covered free viewpoint TV. The idea is to use the collective information from a number of cameras to be able to create a "virtual camera position" from nearly any vantage point. This can be done in 2D or in 3D. Most of the research presented used linear arrays of 2D cameras to first create depth maps of the scene. This was then used to create the virtual camera position at intermediate camera points. Algorithms to develop the depth map, to fix occlusions and edges and other anomalies in this process were described by several groups. I can see that given time, these algorithms will become quite good and end users will someday be able to joystick around a sporting event looking at the action from nearly any angle.

I see the same thing happening in 2D to 3D conversion algorithms. These get better and better every year and at some point may actually get the grudging acceptance of this critical crowd (well, maybe not this crowd).

SD&A also affords a chance to see the latest research in integral imaging displays, volumetric displays and other novel approaches to 3D visualization. Some of these concepts are many years away from commercialization, but it shows where we are likely to end up in 10 or 20 years. Good stuff.

Well, that’s it for now — off to learn more at the last day of this fine event.