Hi Alastair, 50-100k triangles comes down to 600-1200k floats that have to be stored in memory. That's heavy stuff! Of course there's quite a bit of redundancy, since many points arepart of several triangles, but I don't think that this is considered for CGO. There is a way though to get the triangulated VdW surface, colored by atoms. It's the default of 'surface'. Added benefit of that is that I think it does have an optimized mesh representation internally. Of course there must be a specific reason why you want to have your own triangles... Maybe POV-Ray is the way to go then. It does allow optimized meshes and can deal with really huge ones. Oh, and it comes natively with easily adjustable transparency ;)
Cheers, Tsjerk On Mon, Apr 19, 2010 at 10:22 PM, Alastair Fyfe <af...@soe.ucsc.edu> wrote: > Hi Tsjerk, > many thanks - I had missed the "ALPHA, 1.0" and was only specifying RGB > values. The surfaces I've been working with have 50-100K triangles, which is > why I was wondering whether there was a more direct binary interface. Still > everything works, and with the addition of that alpha setting I should be > able to control transparency. > > It is puzzling though that there's not a ready way to generate a VDW surface > colored by atom name. It's doable by checking each triangle for its > adjacent atom and coloring accordingly, but this seems such a common task > I'm surprised there's not already code available for it. > Best, > Alastair > > Tsjerk Wassenaar wrote: >> >> Hi Alastair, >> >> The CGO file should come out something like: >> >> cgo_obj = [ >> BEGIN, TRIANGLES, ALPHA, 1.0, >> COLOR, 1.0, 1.0, 1.0, >> NORMAL, -1.70693, -11.93375, 23.78227, >> VERTEX, 39.96822, 39.96822, -36.33474, >> VERTEX, 34.37814, 37.81595, -37.81595, >> VERTEX, 38.78347, 35.25770, -38.78347, >> COLOR, 1.0, 1.0, 1.0, >> NORMAL, -8.49790, -9.67383, -13.55758, >> VERTEX, 36.09036, 36.09036, 32.80942, >> VERTEX, 35.67732, 32.43392, 35.67732, >> VERTEX, 32.32192, 35.55411, 35.55411, >> ... >> ] >> >> and then you can do cmd.load_cgo(...), but that I guess you already >> know. Right after the TRIANGLES directive, you can notice the ALPHA >> directive, followed by the value for the alpha channel. In this >> example it's set to 1.0, so it's completely opaque; lowering it will >> make it transparent. I have a simple C program somewhere to convert a >> set of triangles specified as vertices and indices into a cgo >> representation. If you're interested I can send you that. I've been >> able to work with quite large triangulated surfaces with this (4096 >> triangles, and I think even up to 16384, but that was too slow to work >> with). >> >> Hope it helps, >> >> Tsjerk >> >> >> On Mon, Apr 19, 2010 at 7:09 PM, Alastair Fyfe <af...@soe.ucsc.edu> wrote: >> >>> >>> Hi, >>> I'm using cgos of fairly large triangulated surfaces from other programs >>> and was wondering if anyone had guidance regarding the following: >>> - the cgo encoding of individual triangles is somewhat bulky and makes >>> for huge files which are slow for pymol to load. Is there an >>> alternative, more efficient means of cgo input? >>> - set cgo-transparency() does not seem to behave as expected, the >>> surface remains opaque. Is there a way to specify alpha along with RGB >>> for individual triangles or any way to investigate why cgo-transparency >>> might not be working? >>> - is there a way to have obtain a VDW surface colored by atom name (eg >>> N3,N7 etc., not atom type)? >>> thanks! >>> Alastair >>> >>> >>> ------------------------------------------------------------------------------ >>> Download Intel® Parallel Studio Eval >>> Try the new software tools for yourself. Speed compiling, find bugs >>> proactively, and fine-tune applications for parallel performance. >>> See why Intel Parallel Studio got high marks during beta. >>> http://p.sf.net/sfu/intel-sw-dev >>> _______________________________________________ >>> PyMOL-users mailing list (PyMOL-users@lists.sourceforge.net) >>> Info Page: https://lists.sourceforge.net/lists/listinfo/pymol-users >>> Archives: http://www.mail-archive.com/pymol-users@lists.sourceforge.net >>> >>> >> >> >> >> > > -- Tsjerk A. Wassenaar, Ph.D. post-doctoral researcher Molecular Dynamics Group Groningen Institute for Biomolecular Research and Biotechnology University of Groningen The Netherlands ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ PyMOL-users mailing list (PyMOL-users@lists.sourceforge.net) Info Page: https://lists.sourceforge.net/lists/listinfo/pymol-users Archives: http://www.mail-archive.com/pymol-users@lists.sourceforge.net
