VMD-L Mailing List

From: John Stone (johns_at_ks.uiuc.edu)
Date: Tue Nov 08 2005 - 14:08:01 CST

Bram,
  If you're running with separate projectors for each eye, then I'm
almost certain you're experiencing a race condition. I've seen it before
on setups using different versions of CAVElib from the one that VMD is
built against. If it's reliably reproducing the problem each time, it
could be that the slave renderer is "winning" the race every time and
the init code that sets up the lighting parameters is leaving the slave
renderer uninitialized. I've never figured out what causes this as the
problem doesn't occur with the CAVElib VMD is compiled against here,
but the version that was used here at NCSA is quite out of date now.
If disabling lighting gives you visible geometry on the other wall(s),
then I'm almost certain the slave renderer is working with bad initial
OpenGL state. Which rev of VMD are you running on this CAVE by the way?

Thanks,
  John Stone
  vmd_at_ks.uiuc.edu

On Tue, Nov 08, 2005 at 09:01:25PM +0100, Bram Stolk wrote:
> John Stone wrote:
> >Bram,
> > If you're getting all black geometry, it sounds like there's a race
> >condition in the CAVE display device with your version of CAVElib.
> >Try restarting the program and see if it behaves this way every time or
> >if you get correct lighting some of the time. If you get correct lighting
> >some of the time, then there's indeed a race condition in the VMD per-wall
> >OpenGL initialization with the version of CAVElib you're using. The bad
> >news is I don't have a way to solve the problem for you since we don't have
> >a CAVE running here currently so I have no way to test. If you want to try
> >changing the CaveDisplayDevice.C startup etc to log the sequence of OpenGL
> >commands that are being executed, you'll probably find that some OpenGL
> >state isn't getting set properly in the slave renderer for one of the eyes.
> >Is your CAVE system an active stereo or passive stereo system? From your
> >description it sounds like a passive stereo system, but it would be helpful
> >to know what the display configuration is so I have some idea what the
> >mapping of renderers to projectors/walls is.
> >
>
> John,
>
> It is reliably reproducible, so it does not look like a race.
>
> Also, we have an active stereo CAVE, but we have two projectors per wall,
> and the projectors are shuttered. This means that we dont use quad-buffered
> stereo, but a non-stereo window per eye, per wall.
>
> I'll have a further look into the code, but my current guess is that
> a lightsource is not enabled.
>
> The reason that linesets are drawn OK is probably because lines are
> commonly drawn with lighting disabled (not to confuse with a lightsource
> that is disabled). Interestingly, if you disable opengl lighting, all
> geometry
> is drawn with lighter colours :-) Lighting only darkens stuff, especially
> if no lightsource is present.
>
> More info later,
>
> Bram
> 

-- 
NIH Resource for Macromolecular Modeling and Bioinformatics
Beckman Institute for Advanced Science and Technology
University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801
Email: johns_at_ks.uiuc.edu                 Phone: 217-244-3349
  WWW: http://www.ks.uiuc.edu/~johns/      Fax: 217-244-6078