I'm trying to write a program that converts Moray's povray output, which
uses the right hand coordinate system, to povray's native left handed
system. As you can see from the images bellow I'm not quite 1/2 way there
yet. The method I use is simply to comment out the sky and up key words in
Moray's output thereby defaulting to povray's lefthanded system. Then for
all vectors I switch the z and y positions. For example

translate <10,-50,100>

becomes

translate <10,100,-50>

This seems to work fine for translate and scale but rotation is a different
matter all together. I found a way to get rotation to work some of the time
by reversing the sign of the magnitude and rotating around the Z axis before
the y axis, so for example

rotate <-28.678608, -34.926117, 42.857418>

becomes

rotate 28.678608*x rotate 34.926117*z rotate -42.857418*y

As you can see from the image below this works for the rotated box in the
background but not for the superellipsoid in the front. Also I get all kinds
of nightmarish things happening if I try to work with csg and rotation. Can
anyone help here cause I feel like I'm close to a solution.

--
----------------------------------------------------------------------------
-----------
Home Page: http://www.geocities.com/~thomaslake/
----------------------------------------------------------------------------
-----------

Post a reply to this message

Attachments:
Download 'PovRay Before.jpg' (26 KB) Download 'Moray Original.jpg' (27 KB) Download 'PovRay After.jpg' (26 KB)

Preview of image 'PovRay Before.jpg'

Preview of image 'Moray Original.jpg'

Preview of image 'PovRay After.jpg'

There is an old program out there somewhere that is called Lefter that does
the conversion. You have to make sure to change the *y and *z stuff though.
"Thomas Lake" <tla### [at] REMOVE-THISshawca> wrote in message
news:3ccbe337@news.povray.org...
> I'm trying to write a program that converts Moray's povray output, which
> uses the right hand coordinate system, to povray's native left handed
> system. As you can see from the images bellow I'm not quite 1/2 way there
> yet. The method I use is simply to comment out the sky and up key words in
> Moray's output thereby defaulting to povray's lefthanded system. Then for
> all vectors I switch the z and y positions. For example
>
> translate <10,-50,100>
>
> becomes
>
> translate <10,100,-50>
>
> This seems to work fine for translate and scale but rotation is a
different
> matter all together. I found a way to get rotation to work some of the
time
> by reversing the sign of the magnitude and rotating around the Z axis
before
> the y axis, so for example
>
> rotate <-28.678608, -34.926117, 42.857418>
>
> becomes
>
> rotate 28.678608*x rotate 34.926117*z rotate -42.857418*y
>
> As you can see from the image below this works for the rotated box in the
> background but not for the superellipsoid in the front. Also I get all
kinds
> of nightmarish things happening if I try to work with csg and rotation.
Can
> anyone help here cause I feel like I'm close to a solution.
>
> --
> --------------------------------------------------------------------------
--
> -----------
> Home Page: http://www.geocities.com/~thomaslake/
> --------------------------------------------------------------------------
--
> -----------
>
>
>

Post a reply to this message

I have done that in the past with a program I wrote to convert AutoCAD
models to POV.  I used much the same method, but it got a little more
complex with rotations around the y and z axes (they have to be reversed),
and I thing I remember that lathes and prisms are very messy since they can
only be defined in 2-dimensions.  I can email code snippts if you're really
interested.  Hope this helps your cause...


"Thomas Lake" <tla### [at] REMOVE-THISshawca> wrote in message
news:3ccbe337@news.povray.org...
> I'm trying to write a program that converts Moray's povray output, which
> uses the right hand coordinate system, to povray's native left handed
> system. As you can see from the images bellow I'm not quite 1/2 way there
> yet. The method I use is simply to comment out the sky and up key words in
> Moray's output thereby defaulting to povray's lefthanded system. Then for
> all vectors I switch the z and y positions. For example
>
> translate <10,-50,100>
>
> becomes
>
> translate <10,100,-50>
>
> This seems to work fine for translate and scale but rotation is a
different
> matter all together. I found a way to get rotation to work some of the
time
> by reversing the sign of the magnitude and rotating around the Z axis
before
> the y axis, so for example
>
> rotate <-28.678608, -34.926117, 42.857418>
>
> becomes
>
> rotate 28.678608*x rotate 34.926117*z rotate -42.857418*y
>
> As you can see from the image below this works for the rotated box in the
> background but not for the superellipsoid in the front. Also I get all
kinds
> of nightmarish things happening if I try to work with csg and rotation.
Can
> anyone help here cause I feel like I'm close to a solution.
>
> --
> --------------------------------------------------------------------------
--
> -----------
> Home Page: http://www.geocities.com/~thomaslake/
> --------------------------------------------------------------------------
--
> -----------
>
>
>

Post a reply to this message

Hmm... See my message/work-around in moray.win.
I think my work-around there might also help with lathes and prisms? I don't
think I tried, but it should work the same as for any CSG object...
Anyway, the conversion is only intended for a group of objects? Not a whole
scene? because that would not make really sense as the end result would be
the same, whether traced in Moray or POV-Ray...

Thomas de Groot


"cadman" <cad### [at] graffitinet> schreef in bericht
news:3ccd5bc7$2@news.povray.org...
> I have done that in the past with a program I wrote to convert AutoCAD
> models to POV.  I used much the same method, but it got a little more
> complex with rotations around the y and z axes (they have to be reversed),
> and I thing I remember that lathes and prisms are very messy since they
can
> only be defined in 2-dimensions.  I can email code snippts if you're
really
> interested.  Hope this helps your cause...
>
>

Post a reply to this message