Dave Akers (dla)
Thu, 21 Oct 1999 15:11:17 -0700
Sorry for the delay.. We've been on a roadtrip for the past week or so..
The crash you're experiencing is actually related to lookup table
support for O2. On the O2, post interpolation lookup tables are not
available, and pre-interpolation lookups are very slow. So, in this case
Volumizer does its own pre-interpolation lookup in software (i.e., it
internally allocates an intermediate buffer to store the results of
lookup-table interpolation.) One of the assumptions it makes is that the
external format of the lookup table is the same as the external format of
the data... and hence the crash occurs when this assumption fails to be
true.
So, in summary, in this case you're actually getting a 4X bloat of your
volume data in main memory.. We could probably find ways in future
releases of the software to reduce this to 3X, at least. (after all, it
seems a bit silly to require that LUMINANCE data be replicated _twice_ to
LUMINANCE_ALPHA.)
Hope that helps,
-Dave
On Tue, 19 Oct 1999, Mark Davey wrote:
> Hi..
> Can anyone help me with the following.
>
> When running the demo
>
> voglBasic $(Datafile) -sc 1 1 3.2
>
> on an O2 with an 8 bit dataset
> CT.Head.Small.char.Tiff
>
> the best external format returns as LUMINANCE_ALPHA while the internal
> format is LUMINANCE8_ALPHA8_EXT.
> I am not sure why this is optimal but that is not my problem however. I
> tried to change these formats directly after the
> voAppearanceActions::getBestParameters command in voglInitAppearance.cxx
> to
> external format = voExternalFormatTypeScope::LUMINANCE
> internal format = voExternalFormatTypeScope::LUMINACE8_EXT
>
> This causes the demo to crash on a
> voAppearanceActions::textureBind(abrick)
> within voglDraw.cxx.
>
> Why is this. Having identical formats on an Onyx2 does not cause the
> same problem.
>
> I assume that generating luminace_alpha formats from a disk format of
> lumince on an O2 is done for speed reasons, but I am more interested in
> memory usage at the moment so I would like luminance internal and
> external formats.
>
> Can anyone help...
> Mark
>
> --
> From: Mark Davey
> Dept. of Medical Physics, UCL.
>
> TEL: +44 (0)171 915 1673.
> FAX: +44 (0)171 837 9279.
>
> Institute of Laryngology and Otology,
> 330 Gray's Inn Road,
> London.
> WC1X 8GE.
>
>
>
This archive was generated by hypermail 2.0b2 on Thu Oct 21 1999 - 17:01:44 PDT