From: John Stone (
Date: Wed Oct 01 2014 - 07:43:22 CDT

  If I follow your description, it sounds like you're making movies
where you have the multiple views shown simultaneously, emulating having
multiple viewports, as opposed to showing the views as part of some

At present, VMD cannot "clone" or link a molecule in the way you
describe without an associated memory use penalty, but adding such a
feature might be interesting for a future version.

Given the limitations of the `current VMD implementation, you would have to
load the trajectory N times where N is the number of simultaneously
shown independent views. If you plan to shown a larger number of total
views V, with N shown at any one time, I would expect that you would still
only need to load N trajectory copies to achieve this, and that to get to
V views you would use script commands to menipulate the view matrices for
the N simultenous views over the course of the movie.

You can effectively program the view matrices for different molecules
independently by toggling all of them "fixed" and then setting the
global view transformation matrix repeatedly, while "unfixing" each
of the molecules one at a time. While "fixed", the changes to the
global transformation matrices are not applied to a given molecule's
transformation matrix. While "unfixed", the changes are applied.

Let me know if this makes sense and if you want further information or
example code.

  John Stone

On Wed, Oct 01, 2014 at 01:25:04PM +0100, Michael Doig wrote:
> Hello,
> I quite often make movies containing multiple viewpoints, from the same
> trajectory file. To do this I load in the same trajectory however many
> times I need to, loading into a new molecule each time, and then move
> the different "molecules" (trajectories), create the representations
> required, and then generate a single movie containing all the different
> required viewpoints. (i.e. a top down view, a side on view, one showing
> just one molecule type etc....) One issue with this is the very large
> amount of memory sometimes required , if I have to load the same large
> trajectory into memory a number of times (sometimes between 4 - 8 times).
> I am wondering if there is any better way of doing this, such as loading
> the trajectory in once, and then creating a new "molecule" by cloning
> (or linking) to the first molecule. This would allow me to have multiple
> viewpoints, without having the same trajectory file loaded into memory
> multiple times.
> I use LAMMPS trajectory files, if that makes any difference.
> Hope this is clear.
> Thanks for your help,
> Michael.
> --
> The University of Edinburgh is a charitable body, registered in
> Scotland, with registration number SC005336.

NIH Center for Macromolecular Modeling and Bioinformatics
Beckman Institute for Advanced Science and Technology
University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801           Phone: 217-244-3349