Making movies in the future

Continuing the theme from yesterday…

At some point technology will be able to accurately digitize an actor’s performance and then play it back, either with faithful reproduction of appearance or with any modification desired. Right now this process is rather expensive, using wonderful technology developed by Paul Debevec and others. Yet Moore’s Law suggests that eventually the process will become inexpensive — less expensive than worrying about lighting and camera placement on set.

In other words, George Clooney will be able to just show up in his street clothes, do the scene against a rough projected digital backdrop (to provide him with context and eye-line cues), and know that niceties such as lighting, make-up, costumes, camera placement and final set dressing will all be done in a computer during post-production. Not only that, but Paul Giamatti will be able to play that same scene, mimicking Clooney’s performance. If he does a good enough job in his acting, you might not be able to tell which was the real Mr. Clooney.

And sooner or later none of this will require millions of dollars of equipment. Anyone with enough money to afford a laptop computer will have all the equipment they need to create what is today considered a high quality Hollywood film. Of course they will still need the talent to provide script, acting, editing and post-production decisions about lighting and camera work, but it no longer becomes a question of money, only of talent.

It might still cost you a lot of money to license George Clooney’s likeness, but if enough people license that likeness, the cost per unit license might become surprisingly small (much like the cost of licensing the sampled sound of a very good Steinway Grand). At that point we’ll start seeing the economics of movies really start to change. In particular, big studios will no longer be able to dictate what movies can get made, any more than they can now dictate what books get written.

One thought on “Making movies in the future”

  1. Audience literacy. Artefacts of these processes will be cool, as the likeness will include temporal sequencing and loops that will take much longer to solve than the static implementation – Max Headroom says hi.
    And a question you brought up at the siggraph panel… at what point will our grey matter implode?

Leave a Reply

Your email address will not be published. Required fields are marked *