"Mike Miller" <mil### [at] gmailcom> wrote:

> Thanks BW.
> I like that idea - using the cameras frustum shape for an inside/outside test.
> Using the cameras' hither and yon planes for space partitioning. BSP?

Nope.  I'm currently too ignorant and unskilled to implement a BSP.  ;)

IIRC, it was just a check to see which side of the 6 planes that bounded the
View Frustum the selected point was on.  Simple.  Fast.

I suppose there might be a way to convert the frustum into a signed distance
function (SDF) and do it that way.   It would be slick, and then you'd have the
distance calculation integrated into each check as well.

- BW

Post a reply to this message

Op 08/03/2023 om 17:37 schreef Mike Miller:
> Thanks Thomas. Very nice - big help.
> Objects not seen by the camera but reflecting in visible surfaces could present
> a problem. I guess you could 'add-back' objects not seen but within a distance
> of the frustum - if that object type was important to the reflections in the
> view.
> Mike.
> 
> 
Indeed. It does not happen often, but then... ;-)

Found the demo. Attached (the macro is written out in the scene again; 
can be replaced by a call to the external macro file of course).

-- 
Thomas

Post a reply to this message

Attachments:
Download 'gt_visibilitytest.pov.txt' (5 KB)