Re: The reflector problem

[T3D John Bercovitz <bercov@xxxxxxxxxxx>]

John Vala writes:

> About Lambertian Surfaces:  By far the best reflector for the type of viewer
> we usually discuss here would be one which has a spherical or (more feasibly)
> a cylindrical shape.  In this case the "point source" would be positioned at
> the center of curvature.  All light rays would travel the same distance
> before hitting the surface and would therefore be the same intensity upon
> incidence (remember the surface of a parabola just keeps getting farther and
> farther away in distance from the source as you go from the Vertex to the
> Latus Rectum!).  

And beyond in my diagram.  Hey - you're not talkin' dirty, are you?

> Since the surface is coated with a Lambertian material, it
> really won't matter how close or far away the reflecting surface is to the
> film gate, it will all look the same intensity.

Right.  Well I think this is a conclusive argument.  Since we have a
cylinder, we need a line source.  I wonder what we can get in the way
of a tubular incandescent.  How do we handle the problem of light from
the source itself?  If we block, it, it becomes a black line; if we
leave it, it becomes a bright line.  OK, so we go to an integrating 
sphere.  888-)  I've built some too.  Not cheap compared to curved
white paper.

If you had a hidden light source, ala red button, what then the shape 
of the reflector?

John


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