MVSmith nous apporta ses lumieres ainsi en ce 2004/01/21 20:50... :

>Hi there...I really love this app. I've been using computers for 35 years,
>and it's been
>a long time since I've wanted to go over every speck of the docs, and get
>into all the
>dark corners.
>
>I have a couple of newbie questions, though, that I haven't had much luck
>finding answers
>for:
>
>  
>
Possible using "photons" and scathering media. Set the media to zero 
absorbtion and set some scathering. The render will take a long time as 
both photons and media will add a lot of computation. (look at 
"optics.pov" in scenes/advanced)

>1) Pointers (links, etc) to info and approaches for using POV-Ray as a
>virtual optics
>    bench. Not asking the app to perfectly model optics--I just want to do
>simple stuff. Example:
>    I'm converting a parabolic satellite dish into a solar collector. I want
>to place a secondary
>    reflector just short of the focal point, and redirect the resulting beam
>down through the
>    axis of rotation. So this is all front-surface reflection, and I'd like
>to look at the result of
>    SORs for the secondary reflector before building something. I think I
>can create an array
>    of multiple-parallel-beam light sources (array of laser beams), so I can
>look at the path of
>    each beam through the system.
>
>    So in this context (other than requesting general info on optics
>projects in POV-Ray), is
>    there a way to "blow smoke" in the air to see each beam without
>attenuating it?
>  
>
Look for "trace.pov" in scenes/language. It use the "math.inc" library. 
It contain a function that return the intersection point of an object 
and a vector.

>2) This is the more serious question. I'd like to model an object (for
>example, the union of
>    two extruded splines that are centered and parallel to the z-axis), take
>cross-sections of
>    the object at intervals, and read each cross-section out into a file so
>the dataset can be converted
>    into commands for a numerical-control milling machine. This is more for
>artistic applications,
>    so extreme precision isn't needed.
>
>    So far, I've discovered only two possible methods, both of which are a
>little kludgy:
>
>    -- Derive the union of the shape and an intersecting plane, and using a
>parallel light source
>    project the slice onto a "screen" of sorts that is coincident with the
>POV of the camera. Save
>    the bitmap, and then use a conversion utility to read out the outline.
>Very kludgy, but creative, no?
>
>    -- Use POV-Ray's _inside_ function to determine if a given point is
>inside or outside the
>    object. Use an interative routine that "scans" along an array of XY
>points on each z-plane
>    cross-section, and then by successive approximation generate a series of
>points that are
>    just inside (or outside) the surface up to some limit of accuracy.
>
>    I can see that I can generate a mesh, save the surface to a file, and
>take cross-sections of that
>    dataset, but I was wondering if there was a more generalized method one
>could use with any
>    given union of objects.
>
>Thanks for at least reading this far! 8)
>
>--Chris
>
>
>
>  
>
Alain

Post a reply to this message