[Chimera-users] odd glare when volume rendering

[Tom Goddard]

Hi Jeff,

  Aha!  Now I see Chimera is working as expected and I can reproduce the 
glare on my computer.  If you have depth cueing set to "No color", which 
is the default, *and* you use a light background color, then Chimera 
interprets "no color" as white.  With a dark background color, I use 
black, "no color" for depth cueing is interpreted as black.  So all I 
had to do was set by background color to a light shade to see the glare.

  So this is bad behavior.  The white depth cue color does not work with 
volume rendering -- it is way too powerful -- all the far planes in the 
data become slightly white, and this ends up making a very white brick.

  Oddly, I have already fixed the problem, accidently!  My current 
development version of Chimera (not on the web) has a new blending mode 
for volumes.  I made the change to address a serious problem you raised 
earlier that the brightness in volume rendering changes a lot when you 
change step size.  The new mode fixes that and will be the default in 
the next snapshot.  Coincidentally the new blending mode also makes 
white fog work as expected without an overly powerful effect.

  I have also added better handling of large EM tomography maps -- file 
loads report progress and can be canceled, default initial display can 
show a single z plane, the z plane slider is built into volume viewer, 
hdf5 format map files can be written which contain subsampled versions 
of the map allowing rapid display of the data at any scale.  I'm testing 
all this on an 2048^3 map.  A snapshot should be out in a month or so, 
though if you really want bleeding edge stuff I could probably give you 
an earlier version for mac os.

    Tom

William Jeffrey Triffo wrote:
> hello,
>
> I am getting an odd glare when I use solid rendering in VolumeViewer - 
> it is orientation dependent, which makes me think it has to do with 
> lighting? I am using build 2416; is it possible to forward a screenshot 
> to the list?
>
> thanks,
>
>