EnigmA Amiga Run 1997 July/EnigmA AMIGA RUN 20 (1997)(G.R. Edizioni)(IT)[!][issue 1997-07 & 08][EAR-CD IV].iso/lightwave/cgplwlist/cgplw06501-06750.txt/[e-mail]_Sat May 27 10:06:01 PDT 1995.msg

Text File | 1995-06-03 | 2KB | 40 lines

Article: 6586 of comp.graphics.packages.lightwave
Xref: netcom.com comp.graphics.packages.lightwave:6586
Path: netcom.com!ix.netcom.com!howland.reston.ans.net!news.sprintlink.net!beta.inc.net!steve
From: [e-mail] (John Foust)
Newsgroups: comp.graphics.packages.lightwave
Subject: Performance buffer: nifty new idea?
Date: 21 May 1995 20:25:36 GMT
Organization: Syndesis Corporation
Lines: 25
Message-ID: <3po7k0$[e-mail]>
NNTP-Posting-Host: 204.95.160.216
X-Newsreader: WinVN 0.92.6+

I was thinking the other night...  some programs save the
intermediate results of the Z-buffer to disk as a "depth
map" that you might use for smart 3D painting or SIRDs.
Each pixel brightness represents the depth of the object
at that point: dark for far away, say, and white for near.

But what about a "performance map" or a "profiler map."
In this bitmap, each pixel would represent how much
processor time was devoted to that pixel.  Like the depth
map, it could be saved to disk along with the RGB image.

In this way, you'd see "hot spots" in your scene that
would represent the most processor-intensive parts of
the rendering...  and in this way, you might be able to
detect areas of the model that are perhaps overly
detailed or overly processor-consumptive.

For example, an overabundance of detailed polygons in
one area of the image would show up as a white spot.

This technique would provide a true measure of all the
many factors and optimization strategies "under the hood"
of the 3D program.

- John Foust