VMD-L Mailing List

From: John Stone (johns_at_ks.uiuc.edu)
Date: Tue May 24 2005 - 09:06:33 CDT

Hernan,
  The "chess board" indicates that your video card ran out of
texture memory for the potential map that you're trying to display.
VMD (currently) displays that pattern when the video board fails to
allocate the necessary texture memory for the 3-D dataset. I plan to
add a piece of code to decimate the volumetric dataset down to a the
highest resolution the video card can display, but that code wasn't written
for 1.8.3, so VMD 1.8.3 gives up and displays the chess board pattern
you're seeing. The maximum 3-D texture size your video board can display
(claimed by the driver anyway) is printed when VMD starts up. Due to the
way OpenGL works, VMD has to allocate 3-D texture maps with dimensions that
are an even power of two. If you're loading a potential map that's say
65x65x65, that means VMD will have to request a texture map of 128x128x128
from OpenGL. If 128x128x128 is bigger than the max 3-D texture map size
your video card supports, then the operation fails and VMD displays the
chess board pattern. Things should improve dramatically in the near future,
as there's a "non-power-of-two texture map" OpenGL extension available
on a number of video boards that helps significantly with this case, and
OpenGL 2.0 eliminates the power of two limitation entirely. In the mean time,
I intend to implement automatic dataset decimation to help low-end laptop
graphics chipsets get by. Hopefully this will be in VMD 1.8.4.
Until then, you'll need to limit the potential map resolution to something
that's within the capability of your graphics board.

  John Stone
  vmd_at_ks.uiuc.edu

On Tue, May 24, 2005 at 06:44:48PM +1000, Hernán Alonso wrote:
> Hello,
> I've been trying to use VMD to represent the electrostatic potential
> of a molecule as a color pattern over a surface representation.
> Although I have no problems in looking at isosurfaces of the
> electrostatic grid, when I choose the Surf representation and the
> Volume coloring method the surface that appears is completely red (or
> dark red), and I cannot see the patterns. I've tried changing the
> range of the scale (under trajectory menu) as well as the color
> scale, but nothing seems to work.
> The VolumeSlice drawing method gives a very weird 'chess board'
> pattern (of red and black squares), so I don't really thing the
> program is reading the numbers properly...
> I've used pymol to represent the same thing and I had no problems.
> I'm using an ibook G4, with 3D support...
> I really don't know what I may be doing wrong,
> Any suggestions would be much appreciated.
> Best Regards,
>
> Hernan
>
> ========================================
> Hernán Alonso
> PhD student
> Computational Proteomics Group
> The John Curtin School of Medical Research
> Building 54
> The Australian National University
> Canberra ACT 0200 Australia
>
> T: +61 2 6125 8302
> F: +61 2 6125 0415
> M: 0431 254 405
> W: http://jcsmr.anu.edu.au
>
> CRICOS Provider #00120C
> ========================================
> 

-- 
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