Re: pfLightSource and performance ?!

New Message	Reply	Date view	Thread view	Subject view	Author view

Bernard Leclerc (bleclerc++at++cae.ca)
Wed, 7 May 1997 12:32:41 -0400

Next message: msitech: "I can't hear you?"
Previous message: Bernard Leclerc: "Re: Collision detection"
Maybe in reply to: Niels Hilbrink: "Collision detection"

> > Niels Hilbrink wrote:
> >
> > > I've got a question on the subject of pfLightSources.
> > >
> > > I noticed a performance drop of nearly 50% in using either a local or
> > > global lightsource. Is this normal ??
> >
> > No. This is not normal.
> >
> > What exactly do you mean by 50% drop? In My case, I noticed a small increase
> > in processing time when using infinite and local light sources - something
> > in the order of a milli-second.
> >
> > What kind of hardware are you using?
>
> Onyx (6*R10k) + REII. The delay (in the Draw stage) is around 50 milliseconds
> compared to a similar scene (ie same flt database) in perfly.

This is a huge delay...

As Remi Arnaud mentioned, using projective texture is really costly. However I'm
not convinced that you're using this feature of the pfLightSource class. Here
are a more questions to help us identify what's going on.

What is the pfFrameRate() and pfPhase()? I assume 30 or 60Hz and PFPHASE_LOCK.

What is the pfMultiprocess() mode? PFMP_APP_CULL_DRAW?

What are the timings observed in perfly for the APP, CULL and DRAW stages?

What are the timings observed in your program?

And finally, how is the pfLightSource created? Is it something like this?

        pfLightSource* ls = new pfLightSource;
        ls->setPos(0, 0, -1, 0);
                or
        ls->setPos(0, 0, 100, 1);
        ls->setSpotDir(0, 0, -1);

Or is it more complicated?

Let us know, we'll see what we can do...

Bernard Leclerc, Technical Leader, 3-D Graphics Applications
CAE Electronics Ltd. (http://www.cae.ca)
8585 Cote De Liesse, Saint-Laurent, Quebec, Canada, H4L-4X4
tel: +1 514 341 2000 extension 2275, fax: +1 514 340 5496

=======================================================================
List Archives, FAQ, FTP: http://www.sgi.com/Technology/Performer/
            Submissions: info-performer++at++sgi.com
        Admin. requests: info-performer-request++at++sgi.com

Next message: msitech: "I can't hear you?"
Previous message: Bernard Leclerc: "Re: Collision detection"
Maybe in reply to: Niels Hilbrink: "Collision detection"

New Message	Reply	Date view	Thread view	Subject view	Author view

This archive was generated by hypermail 2.0b2 on Mon Aug 10 1998 - 17:55:12 PDT

This message has been cleansed for anti-spam protection. Replace '++at++' in any mail addresses with the '@' symbol.