We have been using VoxBlast for Windows to reconstruct cytoskeletal
organization in frog oocytes from confocal Z-series, running on a 90 MHz
pentium with 64 MB RAM. While there are a few bugs remaining in the version
we use, it has proven quite useful for 3-D reconstruction of confocal
images. It does take some practice to set up the color lookup tables and
opacity function, but the newer versions will have a maximum brightness mode
that should be very useful for casual exploration of a volume.
 
Speed is very dependent upon data set size vs. RAM: at "highest quality,"
reconstuctions take about 1 sec per MB of data (including interpolated
sections), until physical RAM becomes limiting. Once disk swapping is
necessary, it slows by a factor of 5-10x. On our system (with 64 MB RAM), I
routinely reconstruct data sets of 100-175 MB (30-50 confocal sections
@768x512 with 3 interpolated sections between each) in about 10 minutes. Two
stereo pairs rendered with Vox Blast are shown in Gard et al., Current
Topics in Dev Biol 31, 383 (1995). While rendering is not real time, it
could probably be sped up considerably by increasing RAM to 128 MB and using
a faster CPU (RAM is more important than CPU speed).
 
When we purchased our system, VoxBlast cost $2000 (I think it is now $2500)
and the computer (with 64 MB RAM and 20" moniter) was about $8000. We have
been quite happy with it.
 
Dave Gard
 
>We're getting lots of info on different vendors but little on price.  How
>much does ImageSpace cost?
>
>Anyone care to comment on the utility (pros, cons) of the VoxBlast system
>from VayTek (roughly $us10K; software only) that runs on PCs, MACs, and
>SGIs?
>
>Rob Palmer
>Ctr Env Biotech
>UT/ORNL
>
>
David L. Gard
Assoc. Professor
Dept. of Biology
University of Utah
Salt Lake City, UT 84112
T: 801-581-7365
F: 801-581-4668
E: [log in to unmask]