VMD-L Mailing List
From: barrett (barrett_at_biop.ox.ac.uk)
Date: Fri Feb 06 2004 - 04:57:01 CST
- Next message: John Stone: "Re: Densities for protein and ligand"
- Previous message: Oliver Beckstein: "Re: random graphics outages on startup"
- In reply to: John Stone: "Re: movie making in batch mode"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
John
Late last year you gave some helpful advice on movie making.
I have conclude that I rally need do need to run in batch mode. You
kindly offered to give some hints on improving performance of the
tracers, and asked for details of the environment.
VMD will be executed on a cluster of dual processor PCs running redhat
7.2. I don't think they have graphics cards.
What else would you like to know?
I suspect you will advise me to make changes to ray tracer settings. It
would be very helpful if you also told me how to do that (e.g. what
files I need to edit)
Thanks again
Paul
-----Original Message-----
From: owner-vmd-l_at_ks.uiuc.edu [mailto:owner-vmd-l_at_ks.uiuc.edu] On Behalf
Of John Stone
Sent: 17 December 2003 02:47
To: barrett
Cc: vmd-l_at_ks.uiuc.edu
Subject: Re: vmd-l: movie making in batch mode
Dear Paul,
At present VMD doesn't have any support for rendering to off-screen
OpenGL buffers. The "snapshot" render command is only available when
using OpenGL, and since VMD doesn't have code for off-screen buffers,
this means that in order to use it, you MUST be running VMD with the
OpenGL window displayed. To be clear, even this is not strictly
sufficient
to guarantee that everything will work fine: the VMD OpenGL window must
normally be the "top" window displayed, i.e. not obscured by any other
windows, as the glReadPixels() routine VMD uses to grab the OpenGL
buffers can behave strangely on some implementations if the window
is obscured by other windows. (even though its double-buffered)
Similarly, you wouldn't want a screen saver to kick on while VMD
is in the middle of making a movie as this has also been known to
cause problems on some platforms (corrupt movie images..).
Down the road I may have time to add code to VMD for rendering to
off-screen OpenGL buffers, but this is tricky and requires a connection
to the local windowing system. If I were to implement such a feature,
it'd probably get a new display device name such as "pbuffer" which
is the term OpenGL uses for these off-screen pixel buffers.
For now, you'll either need to use the ray tracers (you can tune
the rendering speed a number of ways by running on SMP machines,
reducing the number of antialiasing samples taken per pixel, etc),
or you'll have to run VMD as usual rather than in text-only batch mode.
If you must truly run in batch mode, then the ray tracers are probably
the best option, and I can make a few suggestions for improving your
performance if you tell me what kind of machine(s) you're using, and
what operating system you're on.
Thanks,
John Stone
vmd_at_ks.uiuc.edu
On Tue, Dec 16, 2003 at 04:07:33PM -0000, barrett wrote:
> Hi
>
> I want to make movies in VMD in "batch mode" i.e. using vmd -e [file].
>
> I can do this successfully when I run VMD normally within X windows on
> Linux.
>
> However if I run "vmd -dispdev none ..." (which is important for what
Im
> trying to achieve) there is no screen for the snapshot renderer to get
> its image from and the movie doesn't work.
>
> I can successfully get around this by picking a ray tracing renderer,
> e.g. tachyon but it is much much slower.
>
> So the question is... can anyone advise me how to make movies quickly
in
> batch mode when -dispdev is set to none? E.g. is there a way of
> rendering a snapshot of the non-existant window?(!)
>
> Many thanks
>
> Paul
>
> Dr Paul Barrett
> Laboratory of Molecular Biophysics
> Department of Biochemistry
> University of Oxford
> Oxford
> OX1 3QU
>
> Tel 01865 275380
> Fax 01865 275182
>
> barrett_at_biop.ox.ac.uk
> hellopaul1_at_hotmail.com
>
>
-- 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 ------------------------------------------------------------------------ -------- Received: by biop.ox.ac.uk (MX V4.2 AXP) with SITE; Wed, 17 Dec 2003 04:27:37 +0000 Received: from tx1.oucs.ox.ac.uk by fennel.biop.ox.ac.uk (MX V4.2 AXP) with SMTP; Wed, 17 Dec 2003 04:27:36 +0000 Received: from scan1.oucs.ox.ac.uk ([129.67.1.166] helo=localhost) by tx1.oucs.ox.ac.uk with esmtp (Exim 4.20) id 1AWTHc-0000zY-Ga for barrett_at_biop.ox.ac.uk; Wed, 17 Dec 2003 04:27:36 +0000 Received: from rx1.oucs.ox.ac.uk ([129.67.1.165]) by localhost (scan1.oucs.ox.ac.uk [129.67.1.166]) (amavisd-new, port 25) with ESMTP id 03586-07 for <barrett_at_biop.ox.ac.uk>; Wed, 17 Dec 2003 04:27:35 +0000 (GMT) Received: from rama.ks.uiuc.edu ([130.126.120.150]) by rx1.oucs.ox.ac.uk with esmtp (Exim 4.20) id 1AWTHb-0000yw-5f; Wed, 17 Dec 2003 04:27:35 +0000 Received: from rama.ks.uiuc.edu (rama.ks.uiuc.edu [130.126.120.150]) by rama.ks.uiuc.edu (8.12.10/8.12.10) with ESMTP id hBH2l8tX021899 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NOT) for <vmd-l-outgoing_at_rama.ks.uiuc.edu>; Tue, 16 Dec 2003 20:47:09 -0600 (CST) Received: (from majordom_at_localhost) by rama.ks.uiuc.edu (8.12.10/8.12.9/Submit) id hBH2l85i021898 for vmd-l-outgoing; Tue, 16 Dec 2003 20:47:08 -0600 (CST) Received: from moline.ks.uiuc.edu (moline.ks.uiuc.edu [130.126.120.60]) by rama.ks.uiuc.edu (8.12.10/8.12.10) with ESMTP id hBH2l4tW021893; Tue, 16 Dec 2003 20:47:04 -0600 (CST) Received: (from johns_at_localhost) by moline.ks.uiuc.edu (8.11.6+Sun/8.9.1) id hBH2l4I02668; Tue, 16 Dec 2003 20:47:04 -0600 (CST)
- Next message: John Stone: "Re: Densities for protein and ligand"
- Previous message: Oliver Beckstein: "Re: random graphics outages on startup"
- In reply to: John Stone: "Re: movie making in batch mode"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]