|
 |
Having picked my way, line by line, through a few thousand lines of our source
code for various reasons, I thought that it was time to look for a free
debugger.
CLion is free for student and open-source use.
https://www.jetbrains.com/clion/
- BE
Post a reply to this message
|
|
|
|
Once again, I can't more highly recommend the Graphics Gems series, or how
important and helpful it would be for people interested in helping further
develop POV-Ray to begin implementing some of these algorithms.
Doing so in c++ would allow immediate test implementation, however implementing
them in SDL would allow others to play, learn, and become confident in
implementing algorithms of their own choosing.
http://cas.xav.free.fr/Graphics%20Gems%204%20-%20Paul%20S.%20Heckbert.pdf
I'll likely now be poring over the line/ray-cylinder intersection articles.
(Since it is directly related to sor spline bounding, but also there seems to be
issue with correctly rendering cylinders with an essentially unbounded endpoint)
- BE
(I was also excited to see a full polygonolization algorithm for implicit
surfaces just before the line-cylinder intersection articles!)
Post a reply to this message
|
|
|
|
"Bald Eagle" <cre### [at] netscape net> wrote:
> I'll likely now be poring over the line/ray-cylinder intersection articles.
>
> (Since it is directly related to sor spline bounding, but also there seems to be
> issue with correctly rendering cylinders with an essentially unbounded endpoint)
Desmos 3d Graph
All vector math, no algebraic solver needed!
https://www.desmos.com/3d/c795aacwsb
(Just click on the orange surface icon on (at present) line 14 to toggle the
cylinder and see the vectors)
"A benefit of this geometric approach to the ray-cylinder intersection is that
it deals directly in object-space coordinates. This avoids the overhead of
transforming either the rays and/or the bounded cylinder into a canonical form
for testing."
Will work on adding the end cap intersections and surface normals.
Enjoy,
- BE
Post a reply to this message
Attachments:
Download 'screenshot 2025-10-17 140817.png' (64 KB)
Preview of image 'screenshot 2025-10-17 140817.png'
|
|
|
|
"Bald Eagle" <cre### [at] netscapenet> wrote:
> Will work on adding the end cap intersections and surface normals.
Update:
Surface normals got implemented first.
axis-aligned development cylinder worked fine, but when tilted, the normals were
wrong, prompting my vector cross-product question in another thread.
That was the result of using R instead of r, and so once everything gets to the
point of "working", I'll likely do more testing, and then refactor to use more
descriptive identifiers.
1. I found an equation for a general cylinder, however the radius wasn't correct
when I rendered it as an isosurface. (*)
I need to track down the source of that error and document it, because that
would be useful.
As far as I am aware, we only have bounding boxes and spheres for isosurface
shapes. Odd that we don't have bounding cylinders.
I also verified that we cannot transform that bounding box using something like
the point_at or reorient_trans macros.
2. The vector-based method that I'm using will allow cylinder end caps in any
orientation. This would likely allow a great saving in computation to make such
shapes, since we wouldn't have to do CSG and add a whole other layer of bounding
and intersection.
A proposed syntax for such a shape might be something like
cylinder {Start, End, Radius, optional StartNormal, optional EndNormal}
That way if no normals are supplied, then it's just the usual right-angle
cylinder.
There may better ways that make implementation easier.
- BE
After I get that all sorted out, then I can hopefully start on properly bounding
the Catmull-Rom spline to fix the blank-out bug.
(*)
searching again for such an equation (surprisingly hard to find) yields:
\((v_{y}(z-z_{0})-v_{z}(y-y_{0}))^{2}+(v_{z}(x-x_{0})-v_{x}(z-z_{0}))^{2}+(v_{x}(y-y_{0})-v_{y}(x-x_{0}))^{2}=(v_{x}^{2
}+v_{y}^{2}+v_{z}^{2})r^{2}\)
Which looks like it may have the corrective factor included as a coefficient of
r^2
also:
https://math.stackexchange.com/questions/1184038/what-is-the-equation-of-a-general-circle-in-3-d-space
Post a reply to this message
|
|
