Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Photo realism in z-buffer shading???

0 views
Skip to first unread message

Ken Geis

unread,
Mar 8, 1994, 1:42:28 PM3/8/94
to

Hi. I am trying to write a renderer which has the quality of
RenderMan or AutoDesk 3D-Studio. I'd like to know how they do it. Ray
tracing is not an option, unles someone knows some incredibly fast RT
algorithm.
Please, don't flame me; I know I can't do this in a week and
their renderers are very special... I mean they cost several THOUSAND
dollars. If anyone could help me I'd appreciate it.

Thanks in advance,

Ken Geis

Phil Carlisle

unread,
Mar 10, 1994, 12:26:08 PM3/10/94
to

Here's an excellent book you can read to learn about raytracing, and the
tradeoff's between speed and quality, you mention renderman and 3DS, but
what do you think these programs do but raytracing?? optimisation of the
algorithms, and trading off quality for speed are the only options, you
still have to do SOME form of raytracing at the end of the day...

Anyway, here's a book reference, but dont be put off if you dont like
C, as its a really well readable book, and comes with a ready compiled
raytracer, with full source code, very similar to povray et al...

---------------

Photorealism and Ray Tracing in C - Watkins, Coy and Finlay
ISBN: 1-55851-247-0 published by M&T books...

Description:

I've been looking for a good book that deals with 3D math for a
long time, given that math isn't my strong point, I wanted a book
with clear descriptions, diagrams and useful examples, anyway, I
accidentally ordered this book, (I dont program in C), but I've
got to say, its been better for 3D math, than ANY 3D math book
Ive seen to date! It also covers stuff like colour reduction,
image enhancements and effects, anti-aliasing, and other useful
topics....

Overall, a must have book for any graphics programmer.... pc
-------------

/----------------------------------------------------------------------\
| Phil Carlisle - p...@espr.demon.co.uk Hull,England...... |
| please dont send mail to ph...@freenet.fsu.edu as the mailers screwed|
\----------------------------------------------------------------------/

Donald T. Major

unread,
Mar 14, 1994, 10:29:58 AM3/14/94
to

In article <763320...@espr.demon.co.uk>, p...@espr.demon.co.uk (Phil Carlisle) writes:
|> In article <2liguk$o...@shorter.EECS.Berkeley.EDU> kg...@shorter.EECS.Berkeley.EDU writes:
|>
[clipped for brevity]

|>
|> Here's an excellent book you can read to learn about raytracing, and the
|> tradeoff's between speed and quality, you mention renderman and 3DS, but
|> what do you think these programs do but raytracing?? optimisation of the

[clipped]

Whoops! That's precisely the point, Renderman does NOT depend on ray-
tracing; apparently it depends more heavily on a form of polygonal
rendering, with clever application, to reproduce the effects as if the
models were ray-traced, in far less time (believe me, it's a LOT less
time :-)). Check out the past few days worth of discussion in the REYES
Algorithm thread, in this group--it discusses the operation of Renderman,
and points to some papers which discuss how Renderman works.

By the way, even what I said above isn't strictly true; "Renderman" is
a scene specification language, describing all objects, lighting, etc.,
related to a scene; the specification doesn't apparently forbid using
actual ray-tracing to implement the rendering engine which inputs RIB
(Renderman Interface Bytestream) files, but the implementations from
Pixar (such as MacRenderman) don't, relying on the REYES (Renders
Everything You Ever Saw) algorithm instead.

--
Donald Major SAS Institute Inc. "Cicely, let's fling something!"
sas...@unx.sas.com SAS Campus Drive - Northern Exposure
(919) 677-8000 Cary, NC 27513-2414

0 new messages