"Trevor G Quayle" <Tin### [at] hotmailcom> wrote:
> "cpmac" <web### [at] cpmaccom> wrote:
> >  #declare panel= prism {
> >     linear_sweep
> >     linear_spline
> >     0, // sweep the following shape from here ...
> >     0.01, // ... up through here
> >     4, // the number of points making up the shape ...
> >     <0,0>, <5,0>, <5,5>, <0,5>
> >      texture {
> >       pigment{
> >       image_map {png "noddy2.png"
> >         }
> >        }
> >
> >       }
> >  rotate<-90,0,0>
> >  translate<0,6,0>
> >  scale 0.5
> >   }
> > The file parses with no error message but no visible result.
> > How is an image file supposed to adapt to it's suport. The image file has a
> > transparent part. Could it be that only the transparent part is on the
> > panel?
> >
> > cpmac
>
> You should prisms are created as curves on the x-z plane and swept through
> the y-axis.  On the other hand, image_maps are mapped on the x-y plane and
> swept through the z-axis.  In order to get what I assume you are expecting,
> you need to rotate the image_map to suit the prism face orientation.  Also
> note that image_maps are mapped from <0,0> to <1,1> and repeated (unless
> the once keyword is used) so if you want the image_map to fit the 5x5 prism
> face exactly, you will need to scale it as well.
> (see 3.5.1.5.1  Specifying an Image Map)
>
> -tgq

Thanks for the advice . I'll look into all that.
I did manage to solve my problem by adapting the "sign" from the cliche
sample in POV. That uses a "superellipse"
From my trials and errors I seems to me that the image map format must be
the same as the set output format. In the standard setting of BMP output
the script wouldn't work with a png, but with +FN in the control box, the
png file was read and the output was a png file.

Post a reply to this message