I've got all my arrays and macros and splines "working" - I get a render that
approximates what I'm going for.

It also leads me to wonder about some of the finer points of creating these
complex shapes.

As you can see from the attached png, I use a loop to draw cylinders starting at
the "top" of the propeller hub, and work my way around halfway - with the other
endpoint returned by the spline value corresponding to the loop value.  The
weird thing is - the spline appears to give values for the sphere sweep that
don't match up with the cylinder endpoints. (Orange sphere sweep vs multicolored
cylinder fill)

Things get really weird when I replace the cylinders with triangles.  The
vertices are the hub edge, the spline-loop value, and the spline-loop+step
value.  As you can see, there's a big blank space around the hub as though
there's no longer a semicircular path around the hub edge.

I'm confused.

If anyone can explain this strange behaviour, I'd like some insight.
If anyone can suggest some better, faster, or more interesting ways to fill in
the space between the hub and the sphere sweep, I'd love to hear some
suggestions.

I'm specifically looking for a way to get a slight convex curve along the top of
the propeller face.

OR - a way to get a crackle / Voronoi pattern like a dragonfly wing.

Post a reply to this message

Attachments:
Download 'propeller2_cylinders.png' (93 KB)

Preview of image 'propeller2_cylinders.png'

Triangle Fill:

Post a reply to this message

Attachments:
Download 'propeller2_triangles.png' (58 KB)

Preview of image 'propeller2_triangles.png'

On 01/08/2014 21:22, Bald Eagle wrote:
> OR - a way to get a crackle / Voronoi pattern like a dragonfly wing.

I have something like that on a fork of povray:

http://wiki.povray.org/content/User:Le_Forgeron#masonry

http://wiki.povray.org/content/User:Le_Forgeron#voronoi


-- 
IQ of crossposters with FU: 100 / (number of groups)
IQ of crossposters without FU: 100 / (1 + number of groups)
IQ of multiposters: 100 / ( (number of groups) * (number of groups))

Post a reply to this message

Attachments:
Download 'leforgeronmasonry.png' (87 KB)

Preview of image 'leforgeronmasonry.png'

On 01/08/2014 21:26, Bald Eagle wrote:
> Triangle Fill:
> 
Are you using the same type for the sphere_sweep and your spline ?

Is the radius along the sphere_sweep constant ?

Any actual code to share ?

-- 
IQ of crossposters with FU: 100 / (number of groups)
IQ of crossposters without FU: 100 / (1 + number of groups)
IQ of multiposters: 100 / ( (number of groups) * (number of groups))

Post a reply to this message

I fixed the problem - my spline was out-of-whack, which is why it wasn't lining
up with the sphere sweep, and producing all sorts of weird results.

I thought that using triangles would be faster than cylinders - but it's slower?

The other thing that I noticed is that using triangles with a small loop step
made the SKY render slower.  I'm not sure why that would be.

Anyway, after much plotting and puzzling, and backtracing, I fixed it up and I
think it looks pretty good.

Shown with the propeller are the coordinates of the spline and the scaled trig
functions I use to curve around the hub.

Post a reply to this message

Attachments:
Download 'propeller2_dragonfly.png' (133 KB)

Preview of image 'propeller2_dragonfly.png'

> I fixed the problem - my spline was out-of-whack, which is why it wasn't lining
> up with the sphere sweep, and producing all sorts of weird results.
>
> I thought that using triangles would be faster than cylinders - but it's slower?
>
> The other thing that I noticed is that using triangles with a small loop step
> made the SKY render slower.  I'm not sure why that would be.
>
> Anyway, after much plotting and puzzling, and backtracing, I fixed it up and I
> think it looks pretty good.
>
> Shown with the propeller are the coordinates of the spline and the scaled trig
> functions I use to curve around the hub.
>

Looking at the cylinders fill, you see that the corresponding triangles 
are long and narrow. Those are known to be slower. If your triangles 
where oriented perpendiculat to the lenght of the blade, they would 
render faster, but would make the face of the blade to flat. To 
compensate, you'd need to have some other splines running the length of 
the blade according to the desired curvature and use smooth triangles. 
Have two sets of such splines, one for the front and one for the back of 
the propeller so the blades have the desired thickness.



Alain

Post a reply to this message

Alain <kua### [at] videotronca> wrote:
>To compensate, you'd need to have some other splines running the length of
> the blade according to the desired curvature and use smooth triangles.
> Have two sets of such splines, one for the front and one for the back of
> the propeller so the blades have the desired thickness.

> Alain

I follow you.  Luckily, I'm halfway there already - I thought about doing a
feather-type fill, and so I varied my starting point from hub-toward tip-and
back.

I just haven't tested that method with the triangles yet.  I'll have to read up
some on triangles vs smooth_triangles, and add that little bit of curve to the
central rib.

Post a reply to this message

Attachments:
Download 'propeller2_fixed_vein.png' (136 KB)

Preview of image 'propeller2_fixed_vein.png'