[FLASH-USERS] 3D visualization, also Visit issues

[D.F. Duffin]

Hello everyone,

So, I'm getting headaches going through many different visualization
programs, most needing modifications to handle FLASH (I use FLASH2.5).
 If anyone has any program recommendations, I'm all ears.  For now,
VisIt is my 3D preference (has anyone ever tried to do stuff 3D with
IDL?).

VisIt has by far the best 3D surfaces, and has very good features,
especially for animations.  However, being that it is by far and away
the only choice for 3D visuals, I have some serious concerns (I've
included images of the problems below).  Perhaps someone here has found
some solutions (I've also tried the visit mailing list):

1) minor, but most quantities have to be scaled for features to work
(e.g. distances if distances are small, some surfaces turn black.  If
your vector field or distances are to small, vector and streamline will
not work.  This is because VisIt is single precision, still.  The fix
is to scale things.  But then, what units will graphs show?

2) 3D surface over multiple refinement levels will cut out grid like
features in the surface (ghost cells?).  A fix that sometimes works
(hardly ever for me) is that you isolate the refinement levels that are
important for that surface.  However, if your surface is on two
different levels, you're screwed.  You would think there is a way to
fix this...anyone? 

3)  3D streamlines:  If I plot up a magnetic field, my streamlines will
be drawn for each refinement level.  The result is a set of field lines
per starting point with varying degrees of accuracy.  Magnetic field
lines (in my problems) are always living on several if not more
refinement levels per visual, so avoiding multiple levels is
impossible.  Even if I wanted to, the streamline plot will not allow me
(I receive an error, improper pipeline use or something like that). 

As I would like to do 3D surfaces with 3D streamlines and do a movie
over time (and space) in a sim that has adaptive mesh refinement (so
it's changing from plot file to plot file), these errors kill the
pleasure and ease of using VisIt and limit its use (for me) to a data
prospecting tool.  

One possibility is the use of Ifrit (it currently reads HART AMR data),
though it requires a bit of coding to read FLASH data (I'm still not
understanding how to do this, though the fix seems straightforward with
HART already set up).  The maintainer of Ifrit, Nick Gnedin, thinks
this would be straightforward to setup however, so if anyone is
interested+capable...

Thanks to anyone who has read this, and thank you for any feedback!
Dennis