Hi
Lars O. Grobe wrote:
"BTW, as I know that Georg Mischler is on the list: dxf2rad is a great
way to import
data from formZ!!! I wrote dxfr14 and translated it with dxf2rad, it
works perfectly!
So this is an alternative besides the obj-way and the
3ds/mgf/rad-conversion."
What about the 3ds->ConRad->Rad conversion ? 
Ole Lemming
Consultant Engineer & Manager
OpenEntry - www.openentry.dk
Hi!
Ole Lemming wrote:
What about the 3ds->ConRad->Rad conversion ?
The problem with 3ds is that it will triangulate
EVERYTHING - and only the windows-version of formZ
has an 3ds-export.
CU, Lars.
With programs like 3DWin by Thomas Baier (PC-only), file conversion should not really be a problem.
What you say about polygonising everything is true. Only when using file formats that support other kinds of geometry (NURBS, Bezier-patches, CSG-constructs) then you'll be able to have the most economic conversion. There is no "general 3D-file format" at the moment, so unless you can write a plug-in for your software directly, you'll have to live with these limitations...
If we could get the software-firms aware of the importance of Radiance, then they could write export-plugins with a much better support of the radiance-primitives.
Radiance is however capable of rendering files with a lot of polygons. It is not the fastest way, but it'll get you there.
I use the ConRAD-program from Ole Lemming and most models come through. It is not unlikely that these models generate octrees of more then 100MB, which takes time... but the results are OK.
--- stefkeB ---
From: "Lars O. Grobe" <[email protected]>
Reply-To: [email protected]
To: [email protected]
Subject: Re: [Radiance-general] 3DS to RAD
Date: Tue, 24 Jul 2001 17:36:49 +0200Hi!
Ole Lemming wrote:
> What about the 3ds->ConRad->Rad conversion ?
The problem with 3ds is that it will triangulate
EVERYTHING - and only the windows-version of formZ
has an 3ds-export.CU, Lars.
_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp
Hi!
stefke b wrote:
I use the ConRAD-program from Ole Lemming and most models come through. It
is not unlikely that these models generate octrees of more then 100MB, which
takes time... but the results are OK.
Do you use a modified radiance? As far as I know there are limits
concerning the
number of entities.
CU, Lars.
The Desktop Radiance distribution on a PIII-600/256RAM with Win2000.
There are limitations. In the (so far) largest project I used it with, I had to redo the terrain in ArchiCAD, to be able to calculate the octree. This octree was about 350MB. The "error.log" grew one time to 450MB of aiming failures! I guess there were slight rounding errors, causing troubles.
As I am still learning Radiance, the process was a bit cumbersome, but it got through.
--- stefkeB ---
From: "Lars O. Grobe" <[email protected]>
Do you use a modified radiance? As far as I know there are limits
concerning the
number of entities.CU, Lars.
_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp
Lars O. Grobe wrote:
Hi!
stefke b wrote:
> I use the ConRAD-program from Ole Lemming and most models come through. It
> is not unlikely that these models generate octrees of more then 100MB, which
> takes time... but the results are OK.
Do you freeze all your octrees?
Scene data in the tens of MB are common, but a non-frozen
octree beyond 10 MB should be rare.
Do you use a modified radiance? As far as I know there are limits
concerning the number of entities.
There are no inherent limits, if you answer the right questions
when compiling (or at least I'm sure you haven't run into *those*
limits yet).
I finally found the avlmemi error in some old mail:
Your problem most likely has to do with a defective qsort()
routine in Linux. Radiance uses this routine in the context
of an attempt at saving some memory when the ambient data
grows beyond a certain margin. If you have lots of RAM, then
you don't really need this, and you can set the margin to
a very big size with the following compilation option:
-DSORT_THRESH=0x7fffffffffffffff
Have fun!
-schorsch
--
Georg Mischler -- simulations developer -- schorsch at schorsch.com
+schorsch.com+ -- lighting design tools -- http://www.schorsch.com/
Hi!
Georg Mischler wrote:
I finally found the avlmemi error in some old mail:
Thank You!
Your problem most likely has to do with a defective qsort()
routine in Linux.
I remember that older versions used an option -D_QUICKSORT or similar.
The p20 doesn't have it. I also deleted this in older versions, because
it didn't work on my system.
of an attempt at saving some memory when the ambient data
grows beyond a certain margin. If you have lots of RAM, then
you don't really need this, and you can set the margin to
a very big size with the following compilation option:-DSORT_THRESH=0x7fffffffffffffff
I will try this, if I get the error again!
Thank You, CU, Lars.
Hi!
Georg Mischler wrote:
-DSORT_THRESH=0x7fffffffffffffff
Works fine here! Thank you, just got the image.
CU, Lars.