VMD-L Mailing List

From: Eckhart Guthöhrlein (eckhart.guthoehrlein_at_uni-bielefeld.de)
Date: Thu Aug 26 2004 - 07:14:17 CDT

On Wed, Aug 25, 2004 at 09:29:13AM -0500, John Stone wrote:
>
> Besides the rotation and centering matrices, you'll also need to
> make molscript and VMD use the same projection matrix, which is probably
> the trickiest part. If the projection matrix isn't the same, then you'll
> get different degrees of perspective out of Molscript and VMD. You could
> probably start out by doing a simple orthographic projection, until you
> get them lined up with regard to all of the other matrices, then when that
> works, you could move up to doing the projection matrix. If molscript has
> an orthographic projection mode, I'd suggest trying that quickly to see
> how different the views are when that issue is eliminated. Let me know
> where things stand once you've tried that, and I can help you deal with
> getting the perspective projections to line up.

Ah! If I use orthographic projection mode in VMD, I can simply use the
rotate_matrix in molscript like this:

transform atom *
  by centre position atom *
  by rotation
    0.835655 0.477917 -0.270702
    0.534086 -0.591997 0.603565
                0.128199 -0.648951 -0.749955 ;
                                                                                                                                                                                                        
And the views are perfectly identical! That's already great.

As far as I can see, molscript only has an orthogonal projection mode. I
browsed through the manual, but there is only a parameter to adjust
depth-cueing.

Regards,
Eckhart