ap@tukki.jyu.fi (Patric Aalto) (01/08/91)
Hi!
I just finished a small demo I have been working on a couple of months.
It performs real-time ray-tracing and shading (a sort of mixture of them
both) on an IBM PC-compatible computer with VGA graphics. I find it pretty
impressive (although the environment is pretty simple: It has a planet,
a moon and a sun). It runs at 70 frames/second on my 80386/20 Mhz machine.
So, is anybody interested to see this? If so, where should I upload/send
this demo? It is an .EXE program of about 15KB, programmed in assembler
and Turbo Pascal. (Sorry, sources not available, at least not yet :-).
Requirements to run this demo are:
- VGA-card that is register-compatible with IBM model. (Demo uses the
non-BIOS-supported 320x400x256 mode).
- A faster-than-4.77Mhz CPU (though this is not absolutely necessary,
you probably don't want to view this on a standard PC. :-)
This demo also shows how fast it runs on your computer (frames/second),
so you can use it for testing this also.
Patrick Aaltoong@d.cs.okstate.edu (ONG ENG TENG) (01/09/91)
From article <61606@masscomp.ccur.com>, by mark@calvin..westford.ccur.com (Mark Thompson): > In article <1991Jan8.072628.13689@tukki.jyu.fi> ap@tukki.jyu.fi (Patric Aalto) writes: >>I just finished a small demo I have been working on a couple of months. >>It performs real-time ray-tracing and shading (a sort of mixture of them >>both) on an IBM PC-compatible computer with VGA graphics. I find it pretty >>impressive (although the environment is pretty simple: It has a planet, >>a moon and a sun). It runs at 70 frames/second on my 80386/20 Mhz machine. > > Yeah right....and pigs can fly. It takes an AT&T pixel machine (which > cranks out more MFLOPS than a roomful of 80386/20 Mhz machines) about > 3 seconds to ray-trace a simple image. Sounds to me like you are doing > some SERIOUS cheating. Accusations aside, does anyone have a relatively simply program to do ray-tracing on PC. It might require math coprocessor or massive amout of EMS or XMS memory or RAM drive, and it might take 2 weeks to do the job, but just so I can see what myself what's in there. (please, no books, unless it's got a ray-trace program in it). Thanks in advance. E. Teng Ong (ong@d.cs.okstate.edu)
dbuck@ccs.carleton.ca (Dave Buck) (01/09/91)
In article <1991Jan8.190941.16830@d.cs.okstate.edu> ong@d.cs.okstate.edu (ONG ENG TENG) writes: > >Accusations aside, does anyone have a relatively simply program to >do ray-tracing on PC. It might require math coprocessor or massive amout of >EMS or XMS memory or RAM drive, and it might take 2 weeks to do the job, >but just so I can see what myself what's in there. (please, no books, unless >it's got a ray-trace program in it). > >Thanks in advance. > >E. Teng Ong (ong@d.cs.okstate.edu) My raytracer (DKBTrace) runs on a PC. You can get it by anonymous ftp from alfred.ccs.carleton.ca (134.117.1.1) or from the "You Can Call Me Ray" BBS in Chicago at (708) 358-5611. David Buck dbuck@ccs.carleton.ca -- _____________________________________________________________________ | David Buck | My employer is not responsible for| | dbuck@ccs.carleton.ca | my opinions. I'm not even sure | | | I am. |
randys@cpqhou.uucp (Randy Spurlock) (01/09/91)
in article <1991Jan8.190941.16830@d.cs.okstate.edu>, ong@d.cs.okstate.edu (ONG ENG TENG) says: > > From article <61606@masscomp.ccur.com>, by mark@calvin..westford.ccur.com (Mark Thompson): >> In article <1991Jan8.072628.13689@tukki.jyu.fi> ap@tukki.jyu.fi (Patric Aalto) writes: >>> >>> ***** Claims and accusations deleted ***** >>> > Accusations aside, does anyone have a relatively simply program to > do ray-tracing on PC. It might require math coprocessor or massive amout of > EMS or XMS memory or RAM drive, and it might take 2 weeks to do the job, > but just so I can see what myself what's in there. (please, no books, unless > it's got a ray-trace program in it). > > Thanks in advance. > > E. Teng Ong (ong@d.cs.okstate.edu) I have a couple of public domain ray tracers [QRT (Quick Ray Tracer) and DBW render and another one that I can't remember right now] I have the source code and the executable ported to run under MS-DOS. They won't ray trace in real-time but you can look at the source code to see how it all gets done, aside from generating some cute pictures. If these programs are not available from somewhere on the net I could post them, or mail them to someone on the net so that other people can get them...sorry but there is no FTP access here! =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= - Randy Spurlock - | Compaq Computer Corporation --------------------------------------------------------------------------- These opinions are mine...all mine... | He fired his hyper-jets and... just ask anyone who's heard them! | blasted into the 5th dimension! --------------------------------------| UUCP: ...!uunet!cpqhou!randys | Space Man Spiff =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
fwb@hpfcdj.HP.COM (Frank Bennett) (01/11/91)
small Flame on!! It seems like this is one of the more popular topics for students interested in making some sort of contribution to the field of Computer Graphics. I think better areas for contribution might be in a modeler or in 3D animation. I don't know of any PD software for either. (In "C" please) You may in well achieve "near real time" performance if you are only procedurally tracing spheres, but in order to get in the ball game review some of the following work: o rayshade, weedeater.math.yale.edu, Craig Kolb Excellent "C" code (if only everybody coded this good) Primitives include: sphere, triangle, triangle with normals, poly, plane cylinder, cone, heightfield, box, superq Features: depth of field jitter or adaptive supersampling for anti-aliasing 3 light source types surface description texture & bump maps fog mist Algorithms: triangle intersection Bounding Volume Hierarchy o vort, munnari.oz.au Has some interesting procedurally defines surface primitive. o Dante, Eric Haines, 3DEYE & Hewlett Packard Build a model & view it in "real time" using polygons or NURBS, then go get a raytraced or Radiosity rendition of the scene. The only tracer I know which will trace Bspline patches. o Wavefront Build a model, animate it ( that is get the key frames right ), use splines to get the inbetweening right, then ask for raytraced frames of interest. Frank Bennett, Graphics Technology Division, Hewlett Packard my views only, of course.
ap@tukki.jyu.fi (Patric Aalto) (01/11/91)
In article <61606@masscomp.ccur.com> mark@calvin.westford.ccur.com (Mark Thompson) writes: >In article <1991Jan8.072628.13689@tukki.jyu.fi> ap@tukki.jyu.fi (Patric Aalto) writes: >>I just finished a small demo I have been working on a couple of months. >>It performs real-time ray-tracing and shading (a sort of mixture of them >>both) on an IBM PC-compatible computer with VGA graphics. I find it pretty >>impressive (although the environment is pretty simple: It has a planet, >>a moon and a sun). It runs at 70 frames/second on my 80386/20 Mhz machine. > >Yeah right....and pigs can fly. :-) >It takes an AT&T pixel machine (which >cranks out more MFLOPS than a roomful of 80386/20 Mhz machines) about >3 seconds to ray-trace a simple image. Yes, I know.. And it takes many hours to raytrace a simple image using QRT on a 80386/20 Mhz machine without math processor. Using the conventional approach, that is (normalized real (double) -type vectors etc). >Sounds to me like you are doing >some SERIOUS cheating. Sure. Or well.. Depends on what you mean with cheating? My demo does not use the conventional approach to this problem, so feel free to say that it DOES NOT do raytracing. The results LOOK LIKE raytracing though, so what's the difference?-) I'm not cheating anything else than the computer to do something it should not be able to do. Try out this demo and then say what you think. Patrick Aalto ap@tukki.jyu.fi tkp72@jytko.jyu.fi
ap@tukki.jyu.fi (Patric Aalto) (01/11/91)
Hi again! I've been told that the demo I mentioned earlier can be found at chyde.uwasa.fi (sorry, don't have the IP number here now) in the directory PC/demo. I am trying to FTP it to other places as well. Thanks to all of you who sent e-mail and showed interest in this demo. I tried to e-mail it to everyone who asked for it, some of the mails kept bouncing, though. So, if you asked me to e-mail it, and didn't receive anything, try to e-mail me again, with as many reply address options as possible. Patrick Aalto ap@tukki.jyu.fi tkp72@jytko.jyu.fi Ps. Anyone yet able to share first impressions of it on the net?
erich@eye.com (Eric Haines) (01/11/91)
In article <17400021@hpfcdj.HP.COM> fwb@hpfcdj.HP.COM (Frank Bennett) writes: >o rayshade, weedeater.math.yale.edu, Craig Kolb > Excellent "C" code (if only everybody coded this good) I agree - it's the best public domain ray tracer. CSG coming soon! >o Dante, Eric Haines, 3DEYE & Hewlett Packard > > Build a model & view it in "real time" using polygons or NURBS, >then go get a raytraced or Radiosity rendition of the scene. The only >tracer I know which will trace Bspline patches. Also does trim curves, textures & quadrics. I did much of the ray tracer, but lots of people are responsible for the creation of the whole package. The internal name is "Dante" (because it was hellish to implement), externally it's the "Starbase Radiosity and Ray Tracing" package. This package is a library providing an additional set of calls to the Starbase graphics library, and comes bundled in free with HP high-end graphics machines. The radiosity part of the package is unique to the marketplace (plug, plug) - no one else I know has radiosity available commercially. The only other radiosity package I know of is Greg Ward's RADIANCE software, based on his research. If seriously interested, contact: gjward@lbl.gov By the way, the "real-time raytracer" on the IBM PC is kinda cute: some little planets moving around the screen. No shadows or reflections, and the one planet that's shaded is rendered incrementally (using a 3D modification of Bresenham's circle algorithm for XYZ locations, for example), so this whole demo seems to be "ray tracing" without any rays. Fast, though, and I agree with the basic philosophy: why trace rays in cases where it's faster to use incremental techniques? The demo executable is quite small (~12K). Eric Haines
ap@tukki.jyu.fi (Patric Aalto) (01/14/91)
In article <1579@hagbard.dc.luth.se> Ingvar.Strand@ext.luth.se (Ingvar Strand) writes: >Why not give us the name of this demo also. > >/Ingvar Oh.. How shamefully stupid of me... Yes, the name of the demo is RTSYSTEM (as in Ray-Traced system or Real-Time system), so it should be available as rtsystem.zip in chyde.uwasa.fi, for instance. Patrick Aalto ap@tukki.jyu.fi tkp72@jytko.jyu.fi
markv@taylor.Princeton.EDU (Mark VandeWettering) (01/15/91)
>In <1991Jan8.072628.13689@tukki.jyu.fi> ap@tukki.jyu.fi (Patric Aalto) writes: >>Hi! >>I just finished a small demo I have been working on a couple of months. >>It performs real-time ray-tracing and shading (a sort of mixture of them >>both) on an IBM PC-compatible computer with VGA graphics. I find it pretty >>impressive (although the environment is pretty simple: It has a planet, >>a moon and a sun). It runs at 70 frames/second on my 80386/20 Mhz machine. Please, don't call it raytracing if it isn't raytracing. Caveat: I haven't seen this program, although I have read the discussions of it. When one speaks of raytracing, one generally means a renderer that works by generating view rays, and probing a data base of objects to find the closest one. The main advantage: shadows and reflections can easily (if not cheaply) be done within the same framework. Disadvantage: requires (possibly) accessing the entire database of objects for each pixel. It sounds like this demo does some highly optimized phong shading, which is nice, cute, interesting, but it AIN'T RAYTRACING. We have enough muddled terminology (like distributed raytracing) without calling everything that renders a raytracer. It just ain't so. Mark Mark VandeWettering markv@acm.princeton.edu
zap@lysator.liu.se (Zap Andersson) (01/16/91)
markv@taylor.Princeton.EDU (Mark VandeWettering) writes: >>In <1991Jan8.072628.13689@tukki.jyu.fi> ap@tukki.jyu.fi (Patric Aalto) writes: >>>Hi! >>>I just finished a small demo I have been working on a couple of months. [1 Gigabit bla bla deleted] >Please, don't call it raytracing if it isn't raytracing. I agree. Norty you ;-) But it has nothing waz-zu-ever to do with my good friend, Ray. BTW, I have some "demos" that I've made that are "real time" of TRUE raytracing! I.e. compressed animation files that took hours and hours to render, but replays in realtime... should I call THAT real time raytracing for the PC ;-) ? Well, it LOOKS ;-) (coz it IS) like raytracing... and it shows in REALTIME. So now I've invented Real-Time raytracing Q.E.D.? Nah.... dont think so.... problem is, do YOU? ;-) >It sounds like this demo does some highly optimized phong shading, which is >nice, cute, interesting, but it AIN'T RAYTRACING. Oh Mark, my dear friend, do you have to YELL at the poor guy ;-) >We have enough muddled terminology (like distributed raytracing) without >calling everything that renders a raytracer. It just ain't so. >Mark Yeah, rite. >Mark VandeWettering >markv@acm.princeton.edu P.S. My Animations will appear ftp-able realsoonnow with the help of a friend from finland... I just got tons o' work to doy right now and havn't been able to send him disks. P.P.S.S. My anims won 2:nd price in the magazine 'Cadalyst' competition 'Annual Image Award Whatever' last summer, and they were just bits and pieces of demo renderings.... I'll enter something with a story this year ;-) /Z -- * * * * * * * * * * * * * * * * * * My signature is smaller than * * yours! - zap@lysator.liu.se * * * * * * * * * * * * * * * * * *