Re: Musing about stereoscopic plotting.

[sfeam <sf...@us...>]

On Wednesday, 12 November 2014 12:24:26 AM Daniel J Sebald wrote:
> With regard to what I wrote in the previous post about better 
> organization of 2D/3D data aiding new feature development, one could 
> think of the 2D plot as just a projection of 3D data (with added axis 
> decoration capabilities not suitable for 3D).
> 
> This brings to mind something I've wondered about from time to time, 
> which is whether stereoscopic plotting (i.e., two plots side by side, 
> not "stereographic" plotting) is something that would be useful to maybe 
> a few fields like protein analysis, etc. 

This was common 20 years ago, but has been largely replaced by the use
of true stereo, e.g. Zalman monitors, for interactive viewing.
Side-by-side is still used for publication (usually relegated to on-line
supplementary material these days).

In any case the plots one would want this for are not the sort of thing
that gnuplot is good at.  We use specialized programs that know about
protein structure, crystallographic data formats, and relevant conventions
for display of chemical features.

Try an image search for "protein active site stereo" in Google.

	Ethan

> It would be more than merely 
> doing a multiplot, for example.  Rotating the mouse would rotate the two 
> images in concert to give the illusion of rotating a 3D object.  I've 
> limited experience with stereoscopic "projection", and maybe it is more 
> difficult than I imagine.  (There probably not only needs to be viewing 
> angles, but depth of view, or maybe that's called parallax.)
> Nonetheless, it would be interesting.  But again, gnuplot internals 
> would need to be revamped for something sophisticated like that.
> 
> Dan