[PD] Gem: Using 2 glsl_program objects ?

[marius schebella]

actually, I did a quick test now, and it does not work.
it seems gemframebuffer and glsl program don't play together.
a chain like

[gemhead]
  |
[gemframebuffer]
  |
[glsl stuff]
  |
[some texture]
  |
[square]

does not produce any texture in gemframebuffer (only black or whatever 
bg color you use).

:(

marius.


vade wrote:
> I agree, this should be possible in theory. I suspect one should be able 
> to build an abstraction that would dynamically load a shader program, 
> render to a quad, send that quad to a texture and output the texture.
> 
> This is basically all jit.gl.slab does for stream processing.
> 
> I still have heard nothing on the multitexture issue for GLSL however, 
> so there may be some gotchas.
> 
> On Nov 25, 2007, at 3:40 PM, marius schebella wrote:
> 
>> Hi alexandre,
>> why do you think it is not possible? or what exacactly do you want to do?
>> If you try to have several glsl programs like a filter bank, first do
>> glsl-blur, then glsl-bloom, then glsl-apply-texture-to-model... then I
>> think this only works with gemframebuffer plus playing around with
>> "texunit 0", "texunit 1", and so on... and maybe also the right
>> in/outlets of texture to reference gpu textures.
>> unfortunately GEM on os x 10.5 is still broken, so I can only do very
>> rudimentary stuff (no developing), but before that I was trying to do
>> exactly this and I think it should be possible to do.
>> marius.
>>
>> Alexandre Quessy wrote:
>>> Hi,
>>> It seems like we cannot use two GLSL shader programs at a time. Hence,
>>> I need to merge both fragment and vertex programs to use them
>>> together.... Is there a way to use more than 2 shaders ? In Jitter,
>>> there is jit.gl.slab that allows that.
>>> http://www.cycling74.com/documentation/jit.gl.slab
>>>
>>> If not possible in Gem, that would be an awesome add-on to it.
>>>
>>
>>
>> _______________________________________________
>> PD-list at iem.at mailing list
>> UNSUBSCRIBE and account-management -> 
>> http://lists.puredata.info/listinfo/pd-list
> 
>