comp.lang.idl-pvwave archive
Messages from Usenet group comp.lang.idl-pvwave, compiled by Paulo Penteado

Home » Public Forums » archive » iTools hardware rendering
Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend 
Switch to threaded view of this topic Create a new topic Submit Reply
iTools hardware rendering [message #58914] Mon, 25 February 2008 15:43 Go to next message
sjwelch is currently offline  sjwelch
Messages: 5
Registered: October 2007
Junior Member
Greetings all,

I'm having a string of problems relating to hardware rendering in IDL
itools:

Before I start, system specs are as follows:
Hardware: Dual-core P4 @ 3Ghz, 2GB, 256MB ATI Radeon X600
Software: IDL 6.3 on Windows XP Pro, ATI Catalyst Driver 08.2 (and an
earlier version..)

I noticed that my CPU usage was maxed out (at least in one core) when
using interactive itools, this wasn't a problem as the software
rendering was still fast enough to use interactively. This was for an
iplot 3D scatterplot of around 50,000 points (lidar point clouds if
anyone's curious :). Now I needed to visualise around 1-2 million
points. Software rendering was not going to cut it.

Just for comparison a colleague, using an indentical machine, working
in R (which uses OpenGL for its 3D stuff) under Win XP can smoothly
pan and rotate a visualisation of around 2 million points.

I read: http://www.dfanning.com/ographics_tips/slowrendering.html and
made corresponding changes in the IDL preferences window. This didn't
change anthing.

I then upgraded my video drivers to the latest version. No help there,
same speed and CPU usage as in software. In fact that made things
worse. Now I repeatedly get the following error when I mouse over the
plot:

Program caused arithmetic error: Floating illegal operand

Everthing works fine, but slow when I force software mode.

So, conclusions I'm drawing from all this:
IDL is managing to pass the rendering to OpenGL, but thats as far as
its going: whatever software OpenGL renderer ATI has supplied is being
used. The OpenGL software renderer I've now upgraded to is somehow
incompatible with IDL and spits erroneous data back to it, leading to
the error message.

Given that I'm stuck with this hardware and this operating system, any
ideas what I should do next?
(except for smart ones like, "just use R then" ;)

Thanks!
-Sam
Re: iTools hardware rendering [message #59143 is a reply to message #58914] Tue, 11 March 2008 10:08 Go to previous messageGo to next message
Steve Houston is currently offline  Steve Houston
Messages: 3
Registered: November 2007
Junior Member
sjwelch wrote:
> Greetings all,
>
> I'm having a string of problems relating to hardware rendering in IDL
> itools:
>
> Before I start, system specs are as follows:
> Hardware: Dual-core P4 @ 3Ghz, 2GB, 256MB ATI Radeon X600
> Software: IDL 6.3 on Windows XP Pro, ATI Catalyst Driver 08.2 (and an
> earlier version..)
>

Do you have access to IDL 6.4 or later?

Significant improvements were made to rendering performance in 6.4 by
using OpenGL vertex buffer objects whenever possible.

In 6.3 one of your CPU cores is maxed out sending points one by one to
OpenGL. So 50,000 points means 50,000 calls to glVertex.

In 6.4 only one call is required to render all 50,000 points, moving the
workload from the CPU to the GPU.

Regards,
Steve.
Re: iTools hardware rendering [message #59282 is a reply to message #59143] Mon, 17 March 2008 16:29 Go to previous message
sjwelch is currently offline  sjwelch
Messages: 5
Registered: October 2007
Junior Member
On Mar 12, 4:08 am, Steve Houston <shoustonrem...@ittvis.com> wrote:
> sjwelch wrote:
>> Greetings all,
>
>> I'm having a string of problems relating to hardware rendering in IDL
>> itools:
>
>> Before I start, system specs are as follows:
>> Hardware: Dual-core P4 @ 3Ghz, 2GB, 256MB ATI Radeon X600
>> Software: IDL 6.3 on Windows XP Pro, ATI Catalyst Driver 08.2 (and an
>> earlier version..)
>
> Do you have access to IDL 6.4 or later?

I've moved up to IDL6.4, still no improvement. I've also updated
graphics drivers again. I've now gotten rid of the "arithmetic error:
Floating illegal operand " message.

Cheers,
Sam
  Switch to threaded view of this topic Create a new topic Submit Reply
Previous Topic: index arrays of structures
Next Topic: Re: Embedded formatting: !M single-symbol tag misbehaving?

-=] Back to Top [=-
[ Syndicate this forum (XML) ] [ RSS ] [ PDF ]

Current Time: Wed Oct 08 19:43:00 PDT 2025

Total time taken to generate the page: 0.00550 seconds