Tarik

I'd like very much to create Escher-like effects by having certain shapes which should normally be obscured by objects which are nearer to the camera, rendered to the screen anyway.

I've got 2 equally sized cubes standing behind each other. The camera is placed in such a way that it sees only the one in the front.

Now I want the possibility of drawing the cube in the back to the screen also, as if it were just a little cube standing in front of a bigger cube.

The fun of this effect would be that by just rotating the camera we'd get an impression of a little and a big cube at one point, and the impression of two equally sized cubes at another.

I sure hope this kind of stuff can be done in Max, and that someone knows how to do it!

I've got 2 equally sized cubes standing behind each other. The camera is placed in such a way that it sees only the one in the front.
  Now I want the possibility of drawing the cube in the back to the screen also, as if it were just a little cube standing in front of a bigger cube.
  The fun of this effect would be that by just rotating the camera we'd get an impression of a little and a big cube at one point, and the impression of two equally sized cubes at another.

creating-escher-in-opengl

You can get 360' panoramic using the lens_angle attribute of 

jit.gl.render. I believe there is also an ortho attribute which 

removes perspective distortion completely. But I'm not sure if this 

You might be able to delve deeper and muck around directly with the 

OpenGL projection matrix though jit.gl.sketch or using jit.gl.lua 

etc. It would be interesting to know what is possible...

On Jan 30, 2008, at 11:32 AM, Tarik wrote:

> I'd like very much to create Escher-like effects by having certain 

> shapes which should normally be obscured by objects which are 

> nearer to the camera, rendered to the screen anyway.

> I've got 2 equally sized cubes standing behind each other. The 

> camera is placed in such a way that it sees only the one in the front.

> Now I want the possibility of drawing the cube in the back to the 

> screen also, as if it were just a little cube standing in front of 

> The fun of this effect would be that by just rotating the camera 

> we'd get an impression of a little and a big cube at one point, and 

> the impression of two equally sized cubes at another.

> I sure hope this kind of stuff can be done in Max, and that someone 

You can get 360' panoramic using the lens_angle attribute of  
jit.gl.render. I believe there is also an ortho attribute which  
removes perspective distortion completely. But I'm not sure if this  
is what you're after.

You might be able to delve deeper and muck around directly with the  
OpenGL projection matrix though jit.gl.sketch or using jit.gl.lua  
etc.  It would be interesting to know what is possible...

>
> Hi,
>
> I'd like very much to create Escher-like effects by having certain  
> shapes which should normally be obscured by objects which are  
> nearer to the camera, rendered to the screen anyway.
>
>   Concrete example:
>
> I've got 2 equally sized cubes standing behind each other. The  
> camera is placed in such a way that it sees only the one in the front.
>   Now I want the possibility of drawing the cube in the back to the  
> screen also, as if it were just a little cube standing in front of  
> a bigger cube.
>   The fun of this effect would be that by just rotating the camera  
> we'd get an impression of a little and a big cube at one point, and  
> the impression of two equally sized cubes at another.
>
> I sure hope this kind of stuff can be done in Max, and that someone  
> knows how to do it!
>
> Thanks - Tarik

But I do think that indeed I'll have to delve deeper.

I have an idea of what to do, but don't know how to do it. I'll try and explain it in these 2 steps:

1. draw a certain 3d image, all normal-like

2. draw a second 3d image on top of the first one, of which none of the pixels will be obscured.

I have discovered that a very basic version of this can be done by turning off the depth_enable attribute of the gl objects that I want to be drawn. But I don't want to completely disregard all of the depth information: I just want the depth information generated in step 1. to be disregarded by the rendering process used in step 2.

Well, if not getting closer to the anwer, I'm at least getting closer to the real question that I want to ask!

1. draw a certain 3d image, all normal-like
2. draw a second 3d image on top of the first one, of which none of the pixels will be obscured.

Quote: Tarik wrote on Thu, 31 January 2008 03:53

----------------------------------------------------

But I don't want to completely disregard all of the depth information: I just want the depth information generated in step 1. to be disregarded by the rendering process used in step 2.

Quote: Tarik wrote on Thu, 31 January 2008 03:53
----------------------------------------------------
But I don't want to completely disregard all of the depth information: I just want the depth information generated in step 1. to be disregarded by the rendering process used in step 2.
----------------------------------------------------

On Jan 31, 2008, at 12:29 PM, Robert Ramirez wrote:

> Quote: Tarik wrote on Thu, 31 January 2008 03:53

> ----------------------------------------------------

> But I don't want to completely disregard all of the depth 

> information: I just want the depth information generated in step 1. 

> to be disregarded by the rendering process used in step 2.

I believe that Tarik wishes the following:

1. render some geometry which respects depth buffering for itself 

(i.e. back faces won't render over front faces when rendering 

2. render some other geometry which also respects depth buffering for 

itself yet not the previously rendered geometry.

This is accomplished by enabling depth buffering, but clearing the 

depth buffer between rendering each geometry. You can either send the 

message depth_clear to jit.gl.render, or set an individual jit.gl.* 

object's @depth_clear attribute (which will clear the depth buffer 

before rendering, most useful in conjunction with the @layer 

attribute), or you can use the "glclear depth" or "gldepthclear 

" messages to jit.gl.sketch, or jit.gl.lua OpenGL solutions, 

>
> Quote: Tarik wrote on Thu, 31 January 2008 03:53
> ----------------------------------------------------
> But I don't want to completely disregard all of the depth  
> information: I just want the depth information generated in step 1.  
> to be disregarded by the rendering process used in step 2.
> ----------------------------------------------------
>
> i don't understand the difference.

1. render some geometry which respects depth buffering for itself  
(i.e. back faces won't render over front faces when rendering  
something like a sphere or whatever)

2. render some other geometry which also respects depth buffering for  
itself yet not the previously rendered geometry.

This is accomplished by enabling depth buffering, but clearing the  
depth buffer between rendering each geometry. You can either send the  
message depth_clear to jit.gl.render, or set an individual jit.gl.*  
object's @depth_clear attribute (which will clear the depth buffer  
before rendering, most useful in conjunction with the @layer  
attribute), or you can use the "glclear depth" or "gldepthclear  
" messages to jit.gl.sketch, or jit.gl.lua OpenGL solutions,  
or...

Not only have I gotten a dozen of perfect solutions for my problem, also now I can finally concentrate on doing very nifty openGL stuff thanks to my introduction to jit.gl.lua! This makes me very happy.

Creating escher in openGL