Session 4: Keynote Speaker: John Honeycutt, Chief Media Technology Officer, Discovery Network

John Hollycutt, EVP and Head of Int’l Bus Ops, Discovery Networks Int’l (he guides the operational structure in 180+ Int’l markets, and his responsibilities include corporate-level research functions)

The audience for the Sony/Discovery/Imax 3D channel is innovators, social influencers, and idealists.  The three companies have been working on the channel plans for about a year.  3D and HD content can coexist, but dual production costs more.  They are working on new camera technology that will help eliminate this problem (details below).

12% of all people have issues with ‘binocular vision,’ making viewing of 3D images extremely difficult, if not impossible for them.  Not all production works in 3D for all people.  When it doesn’t work, the viewer gets Visually Induced Motion Sickness.  (ex. World’s Deadliest Catch might not be good in 3D.)

Ramping up for 3D will require new crew; stereographer, convergence technician or camera assist, and data handlers (metadata is unbelievably valuable  (also, how do STBs interact with this info?)).  Finding experienced personnel for 3D TV production is a challenge.

3D production issues include beam splitter glare, left/right exposure differences, abd stereo window violations (especially with conversion of IMAX to TV).  Real time monitoring of outputs will be helpful, such as allowing the camera person to see what is being shot.  There will be a stereographer on set for now, although we see the role transitioning to the cinematographer over time.

The metadata strategy needs work ASAP; not just capturing the metadata but making sure that the metadata survives through the postproduction and transmission process.

Eye strain in the edit booth is a major issue!!! Shift lengths may need to be shorter.

Conversion: the race to real time is fine, but we really need flexibility to fix things. Scene depth-matching is key.

Cameras; we must get to small, light form-factor 3D camera.  He showed:

@ The Panasonic AG-3DA1, small, light (6lbs), fixed interaxial but you can control convergence.  (World’s Deadliest Catch – the cameramen refused to use any shoulder mounted cameras because it makes them blind on one side.  A rogue wave could blindside and kill them!)

@ Element Technica Neutron Rig, built for SI-2K minicam, small beam splitter that can be shoulder carried.

NEW CAMERA @@@@ Sony concept prototype designed with Discovery, 3 CMOS chips per eye, full 1920, interchangeable lenses, adjustable interaxial distance 1.5 to 3.5, convergence control via horizontal lens shift, full metadata support, to be field tested in July.

The drivers for HD were sports and natural history,  We are seeing the same pattern with 3D.  The level of consumer demand surprised us.

Q & A

2D-to-3D conversion of library?  It depends on how well it works.  We have compelling content.  I hope to see improved conversion tools at this show.  Today’s conversion options are either $50k/minute of content or you buy at $50k box.  I want a tool that works.  We only have one chance to impress the consumer.  We want to do it right.  Our job is to make people go ‘wow’ and then go out and buy a 3D TV set.

A Discovery Network global tech spec guide for 3D is being written.  It may be release in a month or so.

Session 3: Production and Projection Techniques for Immersive Media

Dr. Foessel, Siegried – Chair, JPEG 2000 committee

What is immersive technology – it is certain behavior of computer and communication that approaches reality.

PRIME (PRoduction and projection technologies for Immersive mEdia) is a consortium of 8 partners.  They are conducting R&D for workflows and key components of multidimensional immersive media.  They analyze markets; do quality evaluation and acceptance testing; work on standardization; and address public relations.

Research areas include 3D cinema, 3D TVs, and panoramic 2d/3d projection.

PRIME has developed new mini-cameras for 3D/multiview capture.  Also they have developed a stereoscopic analyzer (STAN) to assist, control, and correct stereo quality on the set.  STAN does real time analysis of the data from the stereo pair cameras and captures technical metadata about the capture conditions (ex disparity range and convergence plane).  The system has a viewfinder / controller screen with an intuitive GUI.

Their technology for 2D and 3D panoramic production includes a new rig; an omnidirectional camera system that holds up to 12 cameras pointing upward to a 360 degree reflective cone structure.   It has been field-tested in soccer games, a 10 minute documentary, and a 5 minute fiction film.  Two months ago they built a 180 degree theatre with 7 projectors.  They are working towards having the 3D panorama projection system up and running by the end of 2010.  This introduces 3D capture issues, like no common center point, that they are working to resolve.  Also, they are working on higher frame rate and synchronization issues.

PRIME offers lectures on aspects of 3D to students and workers covering concepts like disparity, misalignment, 3D plots, distance and duration of scenes, jumps at cuts, etc.

Regarding the art of panoramic 3D cinema, multiple points of interest with no/some/full narrative connection will impact how to shoot and center the 3D.  Is the screen a geometric continuation of reality and the real space, or a distinct space?

He showed a clip of 3D film – Perigrine