Re: SelectRange performance

Hi Doug,

I'm not surprised at the time to reorganize topology. Do
you have texture mapping disabled? That really slows down
images, and especially if topology needs to be reorganized.

However, you should be able to avoid the SelectRange
mappings by using clipping. See visad/examples/Test35.java
for an example of clipping to the unit (x, y, z) cube by
six calls to DisplayRendererJ3D.setClip().

Good luck,
Bill
----------------------------------------------------------
Bill Hibbard, SSEC, 1225 W. Dayton St., Madison, WI  53706
hibbard@xxxxxxxxxxxxxxxxx  608-263-4427  fax: 608-263-6738
http://www.ssec.wisc.edu/~billh/vis.html


On Thu, 28 Feb 2002, Doug Lindholm wrote:

> Hello,
> 
> My performance is an order of magnitude better (from 10s of seconds to a
> few seconds for a colortable change) now that I am using setRange on my
> ScalarMaps (in effect disabling autoscaling).
> 
> I'm still getting a big performance hit (changing colortable is an order
> of magnitude slower - 10s of seconds as opposed to a few seconds) when I
> use ScalarMaps to SelectRange to keep data from hanging beyond my axis
> ranges. This only happens with image data that extends beyond my axis
> range and thus needs to be cropped. If my axis range is larger that my
> image domain, I get no performance hit. If the data are contoured, no
> hit. I can imagine that crunching the triangles (or whatever) to render
> only the image within the axes would be a hit, but I wouldn't expect an
> order of magnitude, especially just changing the colortable. 
> 
> Does that kind of hit sound reasonable or do you have any ideas to
> optimize this one. Doing a setRange on the SelectRange map (in addition
> to setting the range of its Control) doesn't help.
> 
> Thanks,
> Doug
> 
> -- 
> *----------------------------------------------------------------------*
> | Doug Lindholm, Software Engineer          |  E-mail: lind@xxxxxxxx   |
> | Research Applications Program             |   Phone: 303-497-8374    |
> | National Center for Atmospheric Research  |                          |
> | P.O. Box 3000                             |     There's no place     |
> | Boulder, Colorado 80307-3000              |        like $HOME        |
> *----------------------------------------------------------------------*
>