Re: [Jmol-developers] Initial .cube file isosurface support

[<rka...@ri...>]

On Jul 17, 2005, at 12:11 PM, Miguel wrote:
>
> The problem is ... I think that you want to be able to color the=20
> positive
> and negative orbitals differently.

Indeed, and you can only associate a single identifier with the surface.

>
> I thought that this was the best scheme to be able to give control =
over
> color and display attributes of each independently.

It does allow you to do that, but with the penalty of having to read=20
the same cube file twice. I don't know whether those are being cached=20
by the browser if one does this in the applet, but otherwise it could=20
get time consuming if you don't have too fast a connection to the=20
internet...

On an aside: if you can keep the original cube file information in=20
memory, it would be possible to modify the isosurface value for an=20
existing surface, and possibly create new isosurfaces from the original=20=

cube information, without having to reload it.

>
>>> Q: Are you saying that a particular surface should be tied to a
>>> specific
>>> 'model' ?
>>
>> Yes indeed.
>
> This would solve your 'animation' problem too ... I'll think about=20
> this.

You'd still have to read multiple models: one for each step in the=20
slideshow (as opposed to 'animation' :-)), and be able to associate one=20=

or more isosurfaces with that particular model.
Since I don't think that the cube file format allows for the creation=20
of multiple frames (like for instance gaussian output does), one would=20=

have to build up the animation information inside Jmol. This would=20
require one to be able to

1) load a single model (done by loading the cube file, or a single=20
model input file, which zaps the previous models)
2) read one or more cube files to generate the isosurfaces for that=20
model
3) read the next model file without blowing away the old model (in the=20=

previous frame)
4) read one or more cube files to go with this new model\
5) repeat steps 3 and 4 for every model in the sequence in your=20
slideshow

So far only step one is working, since a surface is not tied to a=20
model. If the tying of an isosurface is done by associating it with the=20=

currently viewed model, steps 2 and 4 are solved. To address step 3 one=20=

may need an 'append' feature, which will read a new model and make that=20=

the currently active one (so any follow up operations automatically=20
work on that one).

Ren=E9