Re: Re: Re: Hacking actors

[Doyle Saylor <djsaylor@xxxxxxxxxxxxxx>]

Greetings Economists,

E Nilson writes back to me:

Doyle wrote:
What it exploits is an appearance of a fear of job loss by >
actors to computer images, but is that what is going on?

Eric,
Yes. It is already happening. But "stars" are not those being affected
right now. Rather, those not being hired right now are certain types of
"extras," in particular those in groups in the deep background of a scene.
It is true, however, that some scenes with lots of extras wouldn't be made
if not for the ability to replace real extras with computer animated people
because of the high cost of using real people.

Doyle
That has been true throughout the history of making movies that people
have sought to use some cost cutting device to limit how many people they
hire for scenes (which is capitalism).  I suppose the idea of replacing
people with puppets has been around for thousands of years.  The
computing problem of producing images (conversational agents) that are
realistic in real time (however many frames per second the medium supports)
is probably insurmountable.  What one sees in recent animation in movies is
very costly animation productions crowding out small scale non animated work
that might have give more work for more actors were those ventures financed.
"The Titanic" and other technically costly works soak up the budgets for
many smaller real world movies.  The computing time costs are feasible in
relation to big budgets but does not explain well what is lost by not
shooting a dozen or more movies on small budgets.  Your claim about extras
not being hired is very hard to substantiate;

http://www.sag.org/pressreleases/pr-la001220.html

is an example of where stats are being used to track minority employment in
the screen actors guild (SAG), but there is no sign of statistics about
animation taking over jobs in the sense of the NYT's article.

The NYT's article indicated that in animation, photo realism is more
difficult but the scare talk about that being an avenue to create movies
with no actors is unrealistic about what computers can do.  The most
difficult computational problem in reproducing a human face is hair.  The
NYTs article expresses the opinion of major capitalist figures in the
industry about what is possible to do with computer animation.  Spielberg
said it was 'non-issue' in replacing actors. And Hanks wasn't worried about
them computing an animation that render him obsolete, Tom Hanks was worried
about control of his image in contexts he wasn't party to.  Remember Hanks'
voice was a toy character in "Toy Story".

I want to repeat what I wrote about in my previous posting, the purpose of
seeing a persons face in movies is to understand how they feel, and also to
know better through watching how the actor uses their body to show where
their attention is focused.  Parallel to that observation is the work in
games and fantasies on the web or cd in which an "avatar" like the current
movie character "Laura Croft, Tomb Raider" provide aspects of body centered
information which works well for games but isn't photo realistic.

The reason these elements matter is that we need to know how to share
attention in order to socially work together.  A movie does not share
attention it is a monologue in pictures, but it surpasses writing in
conveying important aspects of communicating humans knowledge need.

Eric
But, still, come real people have not been hired in some movied because
computer-generated people have been used instead. As 3-D modelling of people
gets better and better over the years I would bet that more actors will be
replaced by computer generated images. Of course, for many decades
film-makers have used various methods to use a few real people to appear to
be a large crowd. But the new powers of 3-D modelling open up new avenues
that have not existed before. ERic

Doyle
Cartoons have been around for decades with no real human being in the
image. The relative form of realism in an avatar (Laura Croft) is not so
much about photo realism, but social realism (a person "really" feeling
connected to the avatar)?   Cartoons didn't look real in the last century,
and now they do, but the realism of the image is not what drives the need
for the image. We aren't able to understand why a movie works if we focus
upon the appearance of a realistic animated face (Engels would say that is
empiricism).  Ultimately the deeper issue is how we share communication to
create a social system.  Accurate modeling of human faces does not replace
the need for human beings to be a part of social network.  Ultimately what
drives people to see movies is to understand in some way more directly the
social fabric.  In essence we see a movie because that is why a face matters
in communicating.  And we would expect that movies that could be used in a
collaborative way instead of as a dialogue, would be going further down the
road of social realism.

Stereoscopic vision is a minor part of seeing the world.  To understand that
point see the literature on the magnocellular pathway in the optic nerve as
the conduit of stereo vision and what it does in primate vision.  "Vision
Science", Stephen Palmer, MIT Press, 2000, page 149, especially the
difference between that pathway and the parvo-nuclear channel.

thanks,
Doyle
http://citeseer.nj.nec.com/context/11616/0

To better understand what goals 3d modeling accomplish,
examine the internet research into "caves";


http://graphics.stanford.edu/projects/imtv/

Stanford Immersive Television project

http://graphics.stanford.edu/projects/shading/

rendering project at Stanford see comments on problems with rendering real
time shading

http://graphics.stanford.edu/projects/iwork/

interactive workspaces where shared attention is critical aspect of the
graphic

http://www.cs.caltech.edu/~heirich/anl_report.html

Cal tech on rendering methods including Neural Networks (Monte Carlo methods
is a reference to Neural Networks) to supplement computation.