Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!uupsi!njin!limonce From: limonce@pilot.njin.net (Tom Limoncelli +1 201 408 5389) Newsgroups: comp.sys.amiga.programmer Subject: Re: Fast 3d Graphics Keywords: fast 3d optimizing assembly cycles Message-ID: Date: 6 Apr 91 00:32:12 GMT References: <9529@star.cs.vu.nl> Organization: Drew University - Madison NJ Lines: 27 To: mmblom@cs.vu.nl In article <9529@star.cs.vu.nl> mmblom@cs.vu.nl (Marc Blom) writes: > 'm wondering how fast 3D graphics can really go. I've written some rotation > and perspective routines (in assembly) and profiling shows that 60 % of the > time is spend in my routine RotatePoint. This routine rotates 1 point around > all three axes in only about 900 cycles minimum and about 1100 cycles max. Yow! You're using the wrong algorithm! I recommend "Computer Graphics, principles and practice" by Foley, vanDam, Feiner & Hughes. Chap 5 has some better ways of doing this kind of stuff. It really makes you appreciate those linear algebra classes that were so tempting to sleep through. They bring a big rotate/scale/move routine down to one precomputation then a matrix-multiple for every point in the display. By doing one precomputation you save a LOT of work for all those other points. Also, comp.graphics has a Monthly Posting Of Frequently Asked Questions (and their Answers). It includes *many* good tips. Some of them are actual simple formulas, some are snotty "here's a list of books... go to the library!" but, hey, the net isn't a replacement for good research. Tom -- One thousand, one hundred, seventy five people died of AIDS last week. Did someone mention a war in Iraq?