Ha, they said I wasn't crazy but I didn't believe them.
OK, so which color scheme is "correct"? I would recommend bagging Autocolor and instead use Colormap (and I mean color'map' this time) and Color. Then run your data into both and feed Colormap's output to Color's 2nd input. You can output Colormap output to ColorBar as well.
Now, you will get an error if you feed a generic group to Colormap as previously stated, so keep your ChangeGroupType ("series") in there. But you have the advantage now that the Colormap editor will reveal the min, max, and histogram (optionally) of the data distribution so you can get a better handle on what's going in. For example, you may have an outlier that is throwing the data distribution off and yielding different color schemes.
Also, connect Print('rd') to your group and examine what is really going on in there (in the Message Window).
A simple 'series' of 100-200 morphologically similar fields should not be causing trouble. A 'bounds box' that you've created in some other way may very well be a problem. However, note that DX will automagically create and append a bounding box component to many fields, which you can typically ignore (if you need to see it, use ShowBox). On the other hand, depending on the strictness of the test going on in ColorBar, it may simply be that a 'generic' group is the problem. As I said before, this object type permits any random collection of junk to be bound together but that doesn't permit it to be used ad lib in all modules. A 'series' is a stricter definition that has some rules. I'm not completely positive if ChangeGroupType does any checking to see if your group actually meets these rules, or simply changes the group metadata for better or worse. Maybe Donna knows the answer to that one?
Chris Pelkie Vice President/Scientific Visualization Producer Conceptual Reality Presentations, Inc. 30 West Meadow Drive Ithaca, NY 14850 [EMAIL PROTECTED]
