I've read a lot of posts that say people don't like modellers because
you can't do multiple rotates etc. Just to let you know, this is what I
do: I write my source directly in POV. When there's a complex object I
can't just measure, I use a 2-D metafile drawing program (In my case,
Micrografx Designer) to draw an orthographic plan of the object. When I
like my results, I transfer the measurements back to POV. It's also
great for zooming way in to check tangent lines and such.
This also allows for a great "cheat" of sorts--if you have a real 3-D
object sitting in front of you, scan it or photograph it or whatever,
then bring the picture into your drawing program, draw your curves over
the pic, delete the pic, and copy the curve measurements into POV
source.
The only thing I really can't do well is bezier patches and
height_fields. Height_fields for things like mountains or whatever are
fairly easy, but what if I want a height_field that's a perfect
hemisphere? Anyway, I'll leave you with this idea.
PS. I've always thought a modeller with hard-code POV on one side and
the wireframe preview on the other would be great. Then you could use
the tools, or just change the POV code directly and get instant results.
And support for multiple transforms and CSG. I've got Visual C++ and
Visual Basic, so maybe I could make it myself someday...

Post a reply to this message