[PD] GEM GEMglClearColor alpha ?

vade doktorp at mac.com
Tue Sep 13 18:39:55 CEST 2005


Hi James, List

Im not at a machine as we speak that has GEM installed, but ill try  
to make the simplest patch to get it working. I appreciate the  
feedback about other openGL calls being 'made behind my back' with  
regard to the other GEM objects.

Im taking an into to openGL course, learning GLUT, so im programming  
with Cocoa/GLUT on OS X, but, then I realized that GEM has wrappers  
for the openGL, so I thought id give some stuff a shot.. interesting  
that you have to use multiple GEMglWrapper objects together.. you  
cannot change certain patch/GEM machine state objects with a  
GEMgWrapper object on its own? l

I did use floats between 0,1, and GLDefine and whatnot, but I put the  
GEMglClearColor between a gemhead and a rectangle. I *was* able to  
change the window background color, but I made no specifications to  
GEMglClear(GL_COLOR_BUFFER_BIT) or anything of that nature.

I will play around a bit when I get home and send an example patch,  
and maybe someone can point me in the right direction.

> ...hmm, now going thru this, you may just want to play with that  
> message to [gemwin] in conjunction with the clearmask message,  
> which I think you'll need to send a [GL_COLOR< - [GLdefine] to...?   
> I'll have to try that out when I have GEM running...

I am not familiar with this message, and I dont see it in the  
documentation that I have.

thanks!


v a d e //

http://homepage.mac.com/doktorp/
doktorp at mac.com


On Sep 13, 2005, at 11:12 AM, james tittle wrote:

> On Sep 13, 2005, at 1:19 AM, vade wrote:
>
>
>> Hello
>>
>
> hiya,
>
>
>> Playing with the gl objects in GEM (0.90) on OS X (Pd 0.38-4  
>> Extended, via Hans) , ive noticed that GEMglClearColor has 5  
>> inputs, one for the gem render chain, 2 - r, 3 - g, 4 - b and  
>> presumably 5 - alpha, yet, changing alpha results in no change in  
>> the window Changing the RGB values does, just like [color] to  
>> gemwin (as it should).
>>
>> Presumably, changing the alpha would act to let the scene draw  
>> 'over' itself, and give an easy way to do trails/blurring. Jitter  
>> supports this feature (not griping.. just saying :), any chance ,  
>> or workaround to get similar functionality, as I love the  
>> aesthetic of it. Id rather not use snap2pix/snap2tex.
>>
>
> ...ok, you have the correct inputs to the object, but there are a  
> coupla things to make sure of:
>
> 1.  the rgba inputs need to be 0-1
> 2.  after [GEMglClearColor] you need a [GEMglClear] that clears the  
> color buffer (GL_COLOR)
> 3.  what is the gemhead ordering?  Are you trying to clear  
> everything via [gemhead 1]?  Or are you inserting this into every  
> drawn object's gemchain?
>
> ...that's the theory, anyway :-)  In practice in GEM, things are  
> not exactly as they may appear...in other words, while I think the  
> glWrapper is one of GEM's most powerful features, there are a few  
> decisions made as default behaviors that may cause unexpected  
> behavior and frustration...fr'instance, glClearColor is called many  
> times behind your back, such as windowInit, window reset, and the  
> color message to [gemwin]...
>
> ...hmm, now going thru this, you may just want to play with that  
> message to [gemwin] in conjunction with the clearmask message,  
> which I think you'll need to send a [GL_COLOR< - [GLdefine] to...?   
> I'll have to try that out when I have GEM running...
>
> ...anyway, as ya can see, there's a bit more to the glWrappers than  
> just trying an object...it's really important to have a copy of the  
> "Red Book" on hand, but in general I've found that it's an  
> extremely quick and powerful way to prototype:  no more compiling ;-)
>
> l8r,
> jamie
>
> _______________________________________________
> PD-list at iem.at mailing list
> UNSUBSCRIBE and account-management -> http://lists.puredata.info/ 
> listinfo/pd-list
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.puredata.info/pipermail/pd-list/attachments/20050913/aaf47c5b/attachment.htm>


More information about the Pd-list mailing list