Re: Here we go again

[P3D Gregory J. Wageman <gjw@xxxxxxxxxx>]

I wrote:

>> It strikes me that a whole new crop of special effects software would
>> have to be developed for 3D films.  In a conventional flat film, the
>> effect only has to simulate the appearance of a 3D object effectively.
>> 
>> For a stereo film, the effect would have to reproduce not only shading
>> and color but actual depth.  In some cases this would simply mean
>> casting rays from two different viewpoints to obtain your stereo pair,
>> but with other types of effects, entirely new algorithms would be
>> required.

Eric replies:

>This to my knowledge turns out to be untrue. The current crop of 
>mainstay 3D modeling software used in television and film (Wavefront, 
>SoftImage, etc) does not merely "shade and color," but models true 
>special relationships in the proper perspective, and thus is more than 
>capable of rendering "true" stereo. The modeler or artist simply 
>"cha-chas" the camera POV to whatever the desired stereo base is, and 
>rerendering the scene(s) for the other "eyeball."

And if you read my second paragraph, you will see that I said exactly
that (casting rays from two different viewpoints) would work for some
cases (specifically, ray-casting- or ray-tracing-based software).

The algorithms to which I was referring are those that generate things
like artificial flames, particle systems effects and other
algorithmically-generated effects that are NOT (currently) 3D modelled.

And of course, all of the "forced perspective" tricks which are employed
to composite models in-camera into scenes with live action would be
completely useless in 3D because the depth cues would give them away.

Is there such a thing as blue-screen effects that work in stereo?
At the very least it seems that it would require twice the work,
with separate masks being required for each view.

The more I think about it, the more I see that many (most?) current films
simply could not have been made in 3D without drastic changes.  One
only has to flip through a recent issue of "Cinefex" to see how often
such tricks that don't work in 3D are used.  This suggests to me that,
until the F/X industry perceives a need to evolve new approaches for stereo,
any new 3D films will be confined to subjects that can be done without
special effects (e.g. nature programs, documentaries, skin flicks), or
done entirely with 3D modelled effects only.

        -Greg


------------------------------