"Bald Eagle" <cre### [at] netscapenet> wrote:
>
> What if you shielded the screen with a black square or
> mesh, and used no_object so that it wasn't visible?

That idea intrigued me, so I just spent several interesting hours(!) testing it
in all sorts of ways. I thought the answer would be a simple and obvious YES--
but it turned out to be a deceptively complex question. ;-)

The short answer is... YES and NO. My test scene is different from m@b's of
course, but I'll use an analogy to his oscilloscope. Yes, the no_image black
square does block radiosity light that's emanating from the screen's waveform--
but it does not eliminate the animation flicker caused by the waveform's
movement...even if the screen 'object' (assuming it's a separate object) has the
no_radiosity flag. That's rather strange, IMO. It's as if the black object is
blocking the light in one paradigm, but allowing the light to pass in another.

To put it another way: If the entire oscilloscope used the no_radiosity flag,
and was totally enclosed in a big black box that's made no_image-- the moving
waveform would still cause flicker throughout the scene.

There seems to be a simple 'rule of thumb' here: If the camera can 'see' any
movement in the scene at all, there is going to be flicker. (Saving and
reloading rad data from frame to frame can certainly minimize that, like m@b
finally used. Assuming that there are no actual moving OBJECTS in the scene.)

But this no-radiosity behavior seems odd. My current grasp of the docs'
description is that it should not only allow radiosity rays to pass through an
object as if it was invisible, but also to stop an object from *producing*
radiosity light. I'm obviously wrong about that last bit. The other surprise is
that a no_rad object still 'collects' rad-light patches from other objects (even
though it's supposed to be 'invisible' to them.) That may or may not be logical
and consistent, I don't know.

Post a reply to this message