Who Am I?

Toney, Alabama, United States
Software Engineer, Systems Analyst, XML/X3D/VRML97 Designer, Consultant, Musician, Composer, Writer

Friday, June 19, 2009

3D Movies, Machinima and Camera Nodes for X3D

Delivering 3D in movies is delivering fish in fishnets instead of newspaper. It captures the 3D but it doesn't render it as well as the 3D engine does.

The challenge is adapting the vocabulary of cinema production to the implementations of standard nodes in a common vocabulary.

For musicians, digital sound systems metaphorically resemble physical systems in the GUI. Until now, 3D has been conceived by math geeks, not cinematographers.

Jump cuts still work if used judiciously. If you insist on capture over scripts, it means you get longer sequences. The writer AND the camera director have to think at the scene level.

FWIW, X3D has proposals for new camera nodes that do just what I described. That is a good step forward for X3D as a standard enabling reapplication of the assets into multiple rendered forms. Yes, one can write camera followers into protos but they are painful and why not let the engine do the heavy lifting when a feature is widely used. Some of the best protos I have been given are that sort. I think adding cinematography vocabulary to X3D is a very good idea now that I've spent time making some machinima from my own 3D worlds. As I said, 2D video renders are not as rich but given YouTube, Hulu, etc., eminently practical as a way to get your work in front of more eyeballs.

Fun times. New art forms. Right now cinematographers are trying to make movies with 3D whereas the 3D artists are trying to conceive in real-time 3D.... which is what the new hypermedia is.

Of course the real challenge is realized when you give the audience the cameras which in real 3D systems is unavoidable because of avatars. There are unlimited numebers of scenes that can be shot in the same set simultaneously. The problem of the jump cut is identity reassignment of the user camera. It feels strange because it breaks virtual identity.

A 3D world is a multi-arc scene simultaneously. Adapting to that means deep endowment of character (repetoire, norms, cultural symbols, and history of the instance) and writers have not yet grasped the depth of that or we wouldn't still be talking about movies.

1 comment:

Jordi R Cardona said...

>"X3D has proposals for new camera nodes"

I would suggest just separate the concept of avatar from viewpoint, so collisions are detected between avatar and scenery and not between viewpoint and scenery. This may allow 3D games, avatars in virtual worlds, and of course, machinima.