plib-users

[Plib-users] sgFrustum setting to the same as opengl

[Oliver B. <sh...@ih...>]

I have a function for setting up a view frustum  I want to know if
its getting the right values to be setup the same as the opengl one
... 
I am getting some wierd results in my landscape renderer -
If I roll around etc everything dissapears  but I cant see whats
going wrong ...

here is the setup code ...

void View::Set(float angle, float near, float far) {
    glMatrixMode(GL_PROJECTION);
    glLoadIdentity();				// Reset The Projection Matrix

    gluPerspective(angle, 640.0/480.0, near, far);	// Calculate The Aspect Ratio Of The Window
    glMatrixMode(GL_MODELVIEW);

    frustum.setFOV(angle, angle) ;
    frustum.setNearFar(near, far) ;

    GetGLError("View::Set()");
}

the code which acts wierd looks like : 
	 center=bounds.getCenter();
            sgFullXformPnt3(tform, center, land->camera->rotation);
	 tform[0]-=land->camera->position[0];
	 tform[1]-=land->camera->position[1];
	 tform[2]-=land->camera->position[2];
	
	 temp->setCenter(tform);
	 temp->setRadius(bounds.getRadius());
	
             if(!land->view->frustum.contains(temp)) return;

and the code for setting up the bouds sphere is  :

 for(a=0; a<4; a++)
   bounds.extend(&land->Vertices[Vertex[a]*3]);

Thanks a heap
Oliver Batchelor 

Re: [Plib-users] sgFrustum setting to the same as opengl

[Steve B. <sjb...@ai...>]

Oliver Batchelor wrote:
> 
> I have a function for setting up a view frustum  I want to know if
> its getting the right values to be setup the same as the opengl one
> ...
> I am getting some wierd results in my landscape renderer -
> If I roll around etc everything dissapears  but I cant see whats
> going wrong ...
> 
> here is the setup code ...
> 
> void View::Set(float angle, float near, float far) {
>     glMatrixMode(GL_PROJECTION);
>     glLoadIdentity();                           // Reset The Projection Matrix
> 
>     gluPerspective(angle, 640.0/480.0, near, far);      // Calculate The Aspect Ratio Of The Window
>     glMatrixMode(GL_MODELVIEW);
> 
>     frustum.setFOV(angle, angle) ;
                     ^^^^^^^^^^^^
               That's going to set the HFOV and VFOV to the
               same values - but you (seem to) have a 640x480
               window.  Try using  (angle,-1) - and we'll compute
               the second angle for you assuming a 3:4 aspect
               ratio.

Dunno if that's enough to account for your wierdness.

You could also use a sgFrustum::getMat4 to grab the matrix out
of the sgFrustum - then you can use glLoadMatrix to put it
onto the projection stack - no messing with gluPerspective - and
it guarantees a good match.

-- 
Steve Baker                  http://web2.airmail.net/sjbaker1
sjb...@ai... (home)  http://www.woodsoup.org/~sbaker
sj...@ht...      (work)

Re: [Plib-users] sgFrustum setting to the same as opengl

[Oliver B. <sh...@ih...>]

view.cpp:42: no matching function for call to `sgFrustum::getMat4 ()'
all the other sg functions work fine so  ?

On Wed, 10 May 2000, you wrote:
> Oliver Batchelor wrote:
> > 
> > I have a function for setting up a view frustum  I want to know if
> > its getting the right values to be setup the same as the opengl one
> > ...
> > I am getting some wierd results in my landscape renderer -
> > If I roll around etc everything dissapears  but I cant see whats
> > going wrong ...
> > 
> > here is the setup code ...
> > 
> > void View::Set(float angle, float near, float far) {
> >     glMatrixMode(GL_PROJECTION);
> >     glLoadIdentity();                           // Reset The Projection Matrix
> > 
> >     gluPerspective(angle, 640.0/480.0, near, far);      // Calculate The Aspect Ratio Of The Window
> >     glMatrixMode(GL_MODELVIEW);
> > 
> >     frustum.setFOV(angle, angle) ;
>                      ^^^^^^^^^^^^
>                That's going to set the HFOV and VFOV to the
>                same values - but you (seem to) have a 640x480
>                window.  Try using  (angle,-1) - and we'll compute
>                the second angle for you assuming a 3:4 aspect
>                ratio.
> 
> Dunno if that's enough to account for your wierdness.
> 
> You could also use a sgFrustum::getMat4 to grab the matrix out
> of the sgFrustum - then you can use glLoadMatrix to put it
> onto the projection stack - no messing with gluPerspective - and
> it guarantees a good match.
> 
> -- 
> Steve Baker                  http://web2.airmail.net/sjbaker1
> sjb...@ai... (home)  http://www.woodsoup.org/~sbaker
> sj...@ht...      (work)
> 
> _______________________________________________
> plib-users mailing list
> pli...@li...
> http://lists.sourceforge.net/mailman/listinfo/plib-users

Re: [Plib-users] sgFrustum setting to the same as opengl

[Steve B. <sjb...@ai...>]

Oliver Batchelor wrote:
 
> view.cpp:42: no matching function for call to `sgFrustum::getMat4 ()'
> all the other sg functions work fine so  ?

Well, sg.h, line 871 inside the sgFrustum definition:

  void getMat4 ( sgMat4 dst ) { sgCopyMat4 ( dst, mat ) ; }                 

...um - I wonder if that was added sometime recently?  What
PLIB version do you have?

-- 
Steve Baker                  http://web2.airmail.net/sjbaker1
sjb...@ai... (home)  http://www.woodsoup.org/~sbaker
sj...@ht...      (work)

Re: [Plib-users] sgFrustum setting to the same as opengl

[Oliver B. <sh...@ih...>]

oops sorry
it is in there I looked in the header file ... so why it does that I
dont know ..

Re: [Plib-users] sgFrustum setting to the same as opengl

[Oliver B. <sh...@ih...>]

A stray old version of plib was sitting somewhere where it shouldnt
have been - I still get the wierdness even after that frustum code
so it cant be that ...  

I transform the "center" of each quad
then I put it into a new sphere with the required radius
then i test if its in the frustum

but when I turn around to face the origin they all dissapear
but come back if I roll upside down ...

I am transforming with the same matrix as given to opengl 
then subtracting the camera position as given to gltranslate

I cant see what im doing wrong :(

Oliver


On Wed, 10 May 2000, you wrote:
> oops sorry
> it is in there I looked in the header file ... so why it does that I
> dont know ..
> 
> _______________________________________________
> plib-users mailing list
> pli...@li...
> http://lists.sourceforge.net/mailman/listinfo/plib-users