home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Light ROM 4
/
Light ROM 4 - Disc 1.iso
/
text
/
net_news
/
1994
/
2.doc
/
000170_shf@netcom.com_Wed Dec 7 12:35:02 PST 1994.msg
< prev
next >
Wrap
Text File
|
1995-03-23
|
2KB
|
38 lines
Article: 423 of comp.graphics.packages.lightwave
Xref: netcom.com comp.graphics.packages.lightwave:423
Newsgroups: comp.graphics.packages.lightwave
Path: netcom.com!shf
From: shf@netcom.com (Stuart Ferguson)
Subject: Re: Pentium Bug WAS Re: LightWave 4.0 Delays
Message-ID: <shfD0GC3G.Mw@netcom.com>
Organization: The Blue Planet
References: <1fc65d78.1d8eecab@amuc.mtroyal.ab.ca> <3bcuqr$pap@xmission.xmission.com> <3bhb6f$d6s@nnrp.ucs.ubc.ca> <shfD0938M.4vu@netcom.com> <3bt0b7$5ae@news.cs.tulane.edu> <Norman-061294183226@red_knight.msfc.nasa.gov>
Date: Wed, 7 Dec 1994 17:47:40 GMT
Lines: 24
+-- Norman@eisner.decus.org (Richard) writes:
| Of course they didn't mention how many billions of FP divisions that
| lightwave makes during a week long render session. So you might see it
| before your drive melts down ;-) I don't know.
The number I heard was something like one in a billion calculations,
which sounds like a lot until you realize that to render a single
frame of high-quality video, LightWave computes about one MILLION
samples (752 x 480 pixels x ~3 samples per pixel). There's probably
between 10 and 100 FP calculations per sample depending on scene
complexity, so you only need to render 10 to 100 frames before you've
reached the point of a billion calculations. That's one day's work
for many people.
The other thing making it worse is that any errors will be converted
to a visual form and flashed at high speed with nearly identical
images computed with slightly different FP calculations. Even small
errors are made highly visible under these circumstances. I also
agree that the earlier any error occurs in the rendering pipeline,
the more severe the result.
--
Stuart Ferguson (shf@netcom.com)
"How do you compute that? Where on the
graph do `must' and `cannot' meet?"