Hi there,

Recently I started working on reproducing the Dokken logo with POV-Ray. It's a
little difficult because the corners are rounded and the top is beveled
(rounded over). So far I've figured out how to round off and bevel all the
straight edges but I'm having trouble beveling the curved parts of the "D,"
"k," and "n." Does anyone have any suggestions regarding what type of objects
to use and how to make those beveled curves? So far everything is CSG, which is
what I'd like to use.

The attached picture shows my progress so far.

Regards,
Dave Blandston

Post a reply to this message

Attachments:
Download 'dokken.jpg' (134 KB)

Preview of image 'dokken.jpg'

Dave Blandston wrote:
> Recently I started working on reproducing the Dokken logo with POV-Ray. It's a
> little difficult because the corners are rounded and the top is beveled
> (rounded over). So far I've figured out how to round off and bevel all the
> straight edges but I'm having trouble beveling the curved parts of the "D,"
> "k," and "n." Does anyone have any suggestions regarding what type of objects
> to use and how to make those beveled curves? So far everything is CSG, which is
> what I'd like to use.

  A straight edge can be rounded by using two boxes and a cylinder.
Likewise a curved edge can be rounded by using two cylinders and a
torus, using the same principle.

  (Of course if the curve is not close to circular, then it becomes
slightly more difficult. You would need an elliptical torus which is not
impossible, but laborious. Note that an unevenly scaled torus is not a
real elliptical torus, as the minor radius does not remain constant.)

Post a reply to this message

Warp <war### [at] tagpovrayorg> wrote:
>   A straight edge can be rounded by using two boxes and a cylinder.
> Likewise a curved edge can be rounded by using two cylinders and a
> torus, using the same principle.
>
>   (Of course if the curve is not close to circular, then it becomes
> slightly more difficult. You would need an elliptical torus which is not
> impossible, but laborious. Note that an unevenly scaled torus is not a
> real elliptical torus, as the minor radius does not remain constant.)

That's very helpful. The "k" and the "n" both use circular curves. The curve of
the "D" is not circular so I'll have to experiment with that one. Thanks!

Regards,
Dave Blandston

Post a reply to this message

Dave Blandston wrote:
> The curve of
> the "D" is not circular so I'll have to experiment with that one. Thanks!

  If it's close enough to circular, then you can simply scale the
rounded cylinder unevenly to match the shape, and it will be close enough.

Post a reply to this message

Warp <war### [at] tagpovrayorg> wrote:
>   If it's close enough to circular, then you can simply scale the
> rounded cylinder unevenly to match the shape, and it will be close enough.

I'll give it a try. I think it'll be close enough to be ok.

Post a reply to this message

Dave Blandston nous illumina en ce 2009-05-27 08:05 -->
> Warp <war### [at] tagpovrayorg> wrote:
>>   If it's close enough to circular, then you can simply scale the
>> rounded cylinder unevenly to match the shape, and it will be close enough.
> 
> I'll give it a try. I think it'll be close enough to be ok.
> 
If you can't come close enough with a scalled torus, then, you'll probably need 
a sphere sweep.

Post a reply to this message

Alain <ele### [at] netscapenet> wrote:
> If you can't come close enough with a scaled torus, then, you'll probably need a
sphere sweep.

A sphere sweep combined with a prism may be a possible solution, because that
would also work for the inside of the curves, but I think that would be
extremely difficult to get just right. Plus I'd have to figure out how to
remove the square notch that the sphere sweep would fit into. This might turn
out to be a really difficult problem to solve with CSG...

Post a reply to this message

Dave Blandston wrote:
> Alain <ele### [at] netscapenet> wrote:
>> If you can't come close enough with a scaled torus, then, you'll probably need a
sphere sweep.
> 
> A sphere sweep combined with a prism may be a possible solution, because that
> would also work for the inside of the curves, but I think that would be
> extremely difficult to get just right. Plus I'd have to figure out how to
> remove the square notch that the sphere sweep would fit into. This might turn
> out to be a really difficult problem to solve with CSG...

  If the curve is very elliptical (although in this case it didn't look
like it), the perfect elliptical torus can be achieved with the poly
primitive (or, alternatively, an isosurface, which basically does the
same thing). You'll need a bit advanced high-school math to come up with
the formula, though.

  An unevenly-scaled regular torus suffices for most cases just fine,
though.

Post a reply to this message

Dave Blandston wrote:
> Alain <ele### [at] netscapenet> wrote:
>> If you can't come close enough with a scaled torus, then, you'll proba
bly need a sphere sweep.
> 
> A sphere sweep combined with a prism may be a possible solution, becaus
e that
> would also work for the inside of the curves, but I think that would be

> extremely difficult to get just right. Plus I'd have to figure out how 
to
> remove the square notch that the sphere sweep would fit into. This migh
t turn
> out to be a really difficult problem to solve with CSG...
> 
	The square notch is actually very easy to do once you have the 
prism for the main shape and the sphere sweep: make the prism 
shorter by the radius of the sweep and add another prism using the 
same spline as the sweep (i.e think "union" instead of "difference").

		Jerome
-- 
mailto:jeb### [at] freefr
http://jeberger.free.fr
Jabber: jeb### [at] jabberfr

Post a reply to this message

Attachments:
Download 'us-ascii' (1 KB)

"Dave Blandston" <nomail@nomail> wrote:
> Hi there,
>
> Recently I started working on reproducing the Dokken logo with POV-Ray. It's a
> little difficult because the corners are rounded and the top is beveled
> (rounded over). So far I've figured out how to round off and bevel all the
> straight edges but I'm having trouble beveling the curved parts of the "D,"
> "k," and "n." Does anyone have any suggestions regarding what type of objects
> to use and how to make those beveled curves? So far everything is CSG, which is
> what I'd like to use.
>
> The attached picture shows my progress so far.
>
> Regards,
> Dave Blandston

This is what I've been working on recently. Apply a bezier curve to that edge
and have some control of the curve AND the face.

Idea for the 3rd version just hit me. The first 2 connected paths from perimeter
up and down the backbone...

so now just need a scaled perimeter, range a function zero to end of that scaled
path (make quad edges) and fill face with 1st method ... all of this then
writtten to a meshfile

thanks for the idea, it will be at the site when done

Post a reply to this message