I created a structure from bicubic patches. I noticed that if I use a shiny
finish the reflected image is not smooth over the edges of the bicubic
patches. With a normal (red) pigment the object seems smooth, so there is no
reason why the reflection of a shiny surface should be broken.

Note: I created the object manually, but I understand the nature of bicubic
patches and I am 99.9999% certain that the individual patches line up
smoothly.

An example picture is available in the newsgroup povray.binaries.images and
is titled:
"Possible bug in shiny bicubic_patch (82kb)"

The border - indicated by a red ellipse - shows exactly the problem.
Another example is marked by the blue ellipse, but on this scale it is more
difficult to see.

Note: The red object has some marks, but these are originated by a
conversion to jpg format. They are not visible in the original BMP (300kb).

Post a reply to this message

Are you sure they line up smoothly? It looks like the edges meet, but the
slopes are different. Just looking at it I can see a sharp corner at the
meeting place in the red one.

Both the edges have to meet *and* the slopes at the edges have to be
identical for the reflection to be smooth.

 - Slime
[ http://www.slimeland.com/ ]

Post a reply to this message

jansb000 <jan### [at] planetnl> wrote:
> Note: I created the object manually, but I understand the nature of bicubic
> patches and I am 99.9999% certain that the individual patches line up
> smoothly.

  It's not enough for the patches to line up, their slopes should also be
the same at the common edge.
  Achieving this is rather simple (even though perhaps not immediately
obvious). Suppose we name the points in the two patches like this:

 1  5  9 13 17 21 25
 2  6 10 14 18 22 26
 3  7 11 15 19 23 27
 4  8 12 16 20 24 28
===========
Patch 1  ===========
         Patch 2

  The points 13-16 are the ones which form the common edge of the two
patches (ie. they have to be the same points in both).
  In order to get the same slope at both edges, all the triplets 9-13-17,
10-14-18, 11-15-19 and 12-16-20 should lay in the same lines (that is,
for example the points 9, 13 and 17 should lay on the same line, ie. they
must be co-linear).
  The points 1-8 and 21-28 don't affect this edge and can be anything (of
course if we want to attach more patches to these two, they have to follow
the same rule with them).

-- 
#macro M(A,N,D,L)plane{-z,-9pigment{mandel L*9translate N color_map{[0rgb x]
[1rgb 9]}scale<D,D*3D>*1e3}rotate y*A*8}#end M(-3<1.206434.28623>70,7)M(
-1<.7438.1795>1,20)M(1<.77595.13699>30,20)M(3<.75923.07145>80,99)// - Warp -

Post a reply to this message

> Both the edges have to meet *and* the slopes at the edges have to be
> identical for the reflection to be smooth.
>
>  - Slime

>   It's not enough for the patches to line up, their slopes should also be
> the same at the common edge.
>
> - Warp

Yep, this is how I modelled this scene. The edges line-up AND the slopes are
identical.
I have uploaded the scene file: "Possible bug in shiny bicubic-patch (11kb)"
in povray.binaries.scene-files  and I uploaded some pictures to show how the
controlpoints are defined in "Re: Possible bug in shiny bicubic_patch
(82kb)" in povray.binaries.images.

Post a reply to this message