Interview with Howard Postley, 3ality Digital Systems - Part 1

Howard Postley is Chief Operating Officer/Chief Technical Officer at 3ality Digital Systems. Howard is responsible for research, product development, and overall operational management. He also designed and launched the first all digital 3D online post facility at 3ality Digital. For U2 3D, he was instrumental in the film’s production and post-production. Mr. Postley’s varied background includes writing for industry publications, and he has an installation of the most complex consumer multimedia product developed to date on display at the Library of Congress.

Could you please describe a typical 3-D live multi-camera captation?
For a football game, we'll have between five and 12 3D cameras (except for really big events where we will have more). Each camera is connected to the input of a SIP2100 (actually, we usually use the SIP2900, which handles up to nine cameras in one box). The SIP has a variety of outputs which are connected to the router of an OB truck and the different views are routed appropriately. We generally add two 3D displays in the truck and a stereographer to an HD truck. Depending on the type of show, we may also need to add convergence operators for each camera but, generally, we use our multi-camera management system to allow a single stereographer to control all of the cameras. We also commonly add a third 3D display outside of the truck for people to look at because otherwise they crowd into the truck and get in the way.

Beyond what we add, the operation is pretty standard. All of the HD gear is normal, as is the vision mixer, EVSs CCUs, etc. We generally shoot Sony HDC1500s but can accommodate other types of cameras. We can use either manned cameras or remote heads. The engineer needs a small amount of training for dealing with 3D. Likewise, the Director needs some experience there as well to make the pacing right.

What kinds of 3-D camera rigs are you using?
Our 3flex rigs, of course! We use a combination of beam-splitter and side-by-side rigs, depending on the shots that are needed. All of our rigs employ dynamic compensation to address mechanical imperfections of the cameras and lenses. In addition, the image processors take out whatever misalignment may be left, in addition to color wedges and keystoning.

Ours is the only system that can shoot and broadcast watchable live 3D. We can generally shoot as much faster than anyone else. We just finished shooting an episodic TV show doing up to 50 setups a day. Our systems auto align and output no alignment or color error. We can overlay graphics. We can freely cut between cameras without worrying about transitions because of the multi-camera management system. There are so many things that we do which are really hard. Not only in the rigs, themselves, but also in control and image processing electronics. It goes well beyond subjective claims, as well, as our systems track numerous qualitative metrics that to assure that what comes out is good 3D.

Have you got a preference between the three main 3-D digital cinema technologies available today (RealD, Dolby, LCS glasses)?
Each has its plusses and minuses. Personally, I have a hard time with the Infitec process because the colors are not balanced for me. On the other hand, I prefer the reduced ghosting. Faster projectors will help.

What is your feeling about the 3-D home market?
It is what will make or break 3D. It is where we are primarily focused. My worry is that the studios will push their 3D theatrical content to the home in anaglyph, in the very short term, which could ruin the whole industry. As such, we're going to work very hard to push a better at home experience to the market quickly. We have a variety of technologies that we can bring to bear here.

Which technological approach woud you choose for a common distribution 3-D codec targeted to the home (spatial compression, 2D-plus-Depth, delta information into video codec's user data or transport layer, etc.)?
2D+Z leaves a lot to be desired. Even Declipse isn't really great. The format really needs to be Left + delta Right + Metadata. We have a lot of metadata that we believe needs to be delivered to the receiver in order to maximize the viewing experience. I don't think that the particular format matters that much, actually. I don't think that there will be a real standard soon so the receivers will have to be flexible, which isn't that hard to do. I think 3D codecs/formats will be more like the web than TV, and that is good. People can improve them instead of getting stuck with something that got compromised by a committee of people with differing agendas.

What is your point of view about existing 3-D home displays (3D DLP, micro-polarization, 120Hz LCD, etc.)?
I like 3D DLP because it is inexpensive. I dislike it because it is U.S. only and people don't want rear projection. I like the quality of micro-pol. I dislike the cost of the film and the yield is far too low. 120Hz LCD, unless you're very clever, isn't fast enough for active glasses; you need at least 180 Hz. Laser, probably too expensive and really not that different from DLP. In general, I think we need passive glasses in the home but good actives are OK. It would be a big improvement if they improved the excitement angle of the receivers...

Part 2 here