Don't we just love it when adding 88 warnings can speed up rendering by 6? (Message 1 to 3 of 3)
From: Jerome M BERGER
Subject: Don't we just love it when adding 88 warnings can speed up rendering by 6?
Date: 30 Nov 1999 22:34:58
Message: <3844973A.817CAA02@enst.fr>
That's what just happened to me: POV complained that I added
unnecessary bounding boxes to some unions but my rendering time went
from 1511s to 252s...
Jerome
--
*******************************
* they'll tell you what can't * mailto:ber### [at] inamecom
* be done and why... * http://www.enst.fr/~jberger
* Then do it. *
*******************************
From: Thorsten Froehlich
Subject: Re: Don't we just love it when adding 88 warnings can speed up rendering by 6?
Date: 30 Nov 1999 23:07:22
Message: <38449efa@news.povray.org>
In article <3844973A.817CAA02@enst.fr> , "Jerome M. BERGER"
<jbe### [at] enstfr> wrote:
> That's what just happened to me: POV complained that I added> unnecessary bounding boxes to some unions but my rendering time went> from 1511s to 252s...
There is a reason why these are warnings. User bounding for other objects -
most primitives like spheres - is not as efficient as the automatic bounding
and would cause a slow down. For unions (and a few other ones) this of
course doesn't always apply.
Thorsten
From: Jerome M BERGER
Subject: Re: Don't we just love it when adding 88 warnings can speed uprendering by 6?
Date: 30 Nov 1999 23:33:29
Message: <3844A4F1.E574287F@enst.fr>
Thorsten Froehlich wrote:
> There is a reason why these are warnings. User bounding for other objects -> most primitives like spheres - is not as efficient as the automatic bounding> and would cause a slow down. For unions (and a few other ones) this of> course doesn't always apply.> > Thorsten
Not quite: the automatic bounding is supposed to be more efficient for
unions too, hence the warning (you'll notice that if you try with a
difference or intersection, you won't get the warning...). But I always
get my most significant bounding speedups with unions...
Jerome
--
*******************************
* they'll tell you what can't * mailto:ber### [at] inamecom
* be done and why... * http://www.enst.fr/~jberger
* Then do it. *
*******************************