[Uni-verse] Audio demo situation?

Emil Brink emil at obsession.se
Fri Nov 17 10:04:06 CET 2006


Samuel Siltanen wrote:
> Hi,
> 
> We have been putting together and testing the software needed for the 
> demo, and although we have had some nice results with the acoustical 
> simulation part, there are still some issue to be solved.
> 
> 1a) Quelsolaar does not work. We have tested the code currently in CVS 
> (latest changes 13 days ago) and it keeps crashing at least on Linux. We 
> have not got it compiled for Windows. According to a debugger the crashing 
> has probably something to do with textures. So where are the changes 
> mentioned by Emil? Could someone compile a working Quelsolaar for Windows?

http://projects.blender.org/viewcvs/viewcvs.cgi/quelsolaar/?cvsroot=verse&sortby=date#dirlist
shows three files (p_sds_obj_tess.c, p_sds_obj_edge_normal.c and 
e_storage_head.c) changed 18 hours ago, which a bit less than 13 days.

Do you not see them?

I just checked, and my tree has only two local changes that are not checked
in (the render-to-texture disabling, and a debugging printf()). So, you
really should see the code ...

I've built it on Windows with success, and Slas said it worked better for
him, too.

Write back if you still don't see it, I guess I can ship you a Win32 binary
then.

> 1b) An alternative solution is to get the OSG renderer working properly 
> with UVAS. This requires defining the "listener" object properly in the 
> renderer. The updated version was promised, but not seen so far.

Seems Marcus has revolved this?

> 2) We do not have a decent model for the demo yet. The "testscene" is the 
> most promising candidate, since the more complex models are too complex 
> for the real-time acoustical simulation, even when they go through the 
> reduction. The changes needed for the test scene are simple. Add some 
> pieces of furniture (i.e. copy them from the more complex model) and 
> lights. In addition, scale the model down to the proper scale. This might 
> not be significant visually, but acoustically it is.

Here, we have to wait further for Slas to convert the latest from -+ I
guess.

Regards,

/Emil



More information about the Uni-verse mailing list