mp@pixar.UUCP (Mike Paquette) (08/13/87)
Does any of the current crop of Amiga animation and modeling
software support some form of machine-independent model
representation? I'm looking for some form of scene description
language [1] and geometric primitive support, such as patches,
meshes, or polygons.
This would let an animator whip up models and animation sequences
on the Amiga, and then take the resulting model and scene information
and feed it through a very high quality production rendering system
once the animation 'proofed' on the Amiga looks good.
Also, has anyone writing rendering and ray tracing code looked
into antialiasing? "The Dream Comes Alive", 'Perpetrated by
Leo L. Schwab', suffered badly from aliasing artifacts. Granted,
antialiasing can be expensive in terms of computation time,
but it sure makes the end result look better. After sweating
days or weeks over an animation sequence, its not worth saving
a few minutes a frame only to have your masterpiece stuffed
chock full of jaggies and banding.
An antialiasing option (so it could be switched off for quick
proofs) would be a big win.
Mike Paquette
{ucbvax,sun}!pixar!mp
[1] Not necessarily SDL (Registered Trademark of Alias Research, Inc.)
but any sort of description of a scene.
[The opinions expressed here are my own, and do not necessarily
reflect the opinions of my employer, wife, or pets. In fact, this
whole article is an artifact of line noise....]czei@osupyr.UUCP (Michael S Czeiszperger) (08/16/87)
In article <992@pixar.UUCP> mp@pixar.UUCP (Mike Paquette) writes: >Does any of the current crop of Amiga animation and modeling >software support some form of machine-independent model >representation? I'm looking for some form of scene description >language [1] and geometric primitive support, such as patches, >meshes, or polygons. > >This would let an animator whip up models and animation sequences >on the Amiga, and then take the resulting model and scene information >and feed it through a very high quality production rendering system >once the animation 'proofed' on the Amiga looks good. > At Ohio State University, the Computer Graphics Research Group is doing just that. Their 'scn assembler' program using a text dexcription of polygons, objects, eye point, etc, to generate each frame. The program was ported over to the Amiga, and has been used to teach beginning animation to Art students for over a year. It is possible to perfect an animation on the Amiga, and then modem the text description over to CGRG's custom 32 bit buffer for rendering. (The only problem is a beginning animation student having enough clout to use the buffer when it's being fought over by half a dozen grad students) They currently are making a large investment toward developing a standardized scene description language. The project is known by the name ApE, although I can't remember what it stands for. If you'd like more information contact cbosgd!osu-eddie!osu-cgrg!mark. Michael S. Czeiszperger | Disclaimer: "Sorry, I'm all out of pith" Sound Synthesis Studios | Snail: Room 406 Baker Phone: (614) College of the Arts Computer Lab | 1971 Neil Avenue 292- The Ohio State University | Columbus, OH 43210 0895 UUCP : {decvax,ucbvax}!cbosgd!osupyr!czei
kent@xanth.UUCP (Kent Paul Dolan) (08/19/87)
In article <107@osupyr.UUCP> czei@osupyr.UUCP (Michael S Czeiszperger) writes: >In article <992@pixar.UUCP> mp@pixar.UUCP (Mike Paquette) writes: >>Does any of the current crop of Amiga animation and modeling >>software support some form of machine-independent model >>representation? I'm looking for some form of scene description >>language [1] and geometric primitive support, such as patches, >>meshes, or polygons. >> >>This would let an animator whip up models and animation sequences >>on the Amiga, and then take the resulting model and scene information >>and feed it through a very high quality production rendering system >>once the animation 'proofed' on the Amiga looks good. >> > >At Ohio State University, the Computer Graphics Research Group is doing >just that. [...] >They currently are making a large investment toward developing >a standardized scene description language. The project is known >by the name ApE, although I can't remember what it stands for. If >you'd like more information contact cbosgd!osu-eddie!osu-cgrg!mark. > (Why, WHY am I trying to help someone at Pixar?!? Oh, well:) Mike, Besides IGES (Initial Graphical Exchange Standard, or some such), of which you're probably already aware, the following current new book might be of interest to Pixar and to you as a way to get in on the ground floor of some megabucks. It is a (published softcover book form of a) draft of an ISO (the international standards organization) WG2 (working group 2) proposed standard, as best I can figure out. Research Reports ESPRIT Project 322. CAD Interfaces (CAD*I). Volume 1. Specification of a CAD*I Neutral File for Solids, Version 2.1 Edited by E.G. Schlectendahl Copyright 1986, Springer-Verlag, Berlin and Heidelberg ISBN 3-540-16297-X ISBN 0-387-16927-X (Why two ISBNs?) QA 76.27 .S67 1986 (Library callout) From the jacket back: "ESPRIT Project 322, 'CAD Interfaces', has been established to define the most important interfaces in CAD/CAM systems for data exchange, data base, finite element analysis, experimental analysis, and advanced modelling. The definitions of these interfaces are being elaborated in harmony with international standardization efforts in this field. "The Project is to contribute to the compilation of European Know-how, and to facilitate the application of CIM methods in industry on a broad basis. In this way, the European influence on international standardization bodies is to be strengthened." Replace "European" by "human" here, and my urge to upchuck at the Jingoism would lessen considerably [ When are folks going to understand that we all live on the same little blue rock? ], but still it looks like a possible help for your needs or related to your needs. By the way, this is a definite cure for insomnia, like all standards documents. Kent, the man from xanth. -- Kent Paul Dolan, LCDR, NOAA, Retired; ODU MSCS grad student // Yet UUCP : kent@xanth.UUCP or ...{sun,harvard}!xanth!kent // Another CSNET : kent@odu.csnet ARPA : kent@xanth.cs.odu.edu \\ // Happy USPost: P.O. Box 1559, Norfolk, Virginia 23501-1559 \// Amigan! Voice : (804) 587-7760 -=][> Last one to Ceres is a rotten egg! -=][>
jdm@pnet02.CTS.COM (John Mesiavech) (08/19/87)
Mike;
Taken in combination, there are some programs out there on the Amiga scene
that do what you ask quite admirably.
Examples:
Videoscape-3D (a 3-D animation system used to produce Leo's animation) uses
output files in an ASCII vertice format for it's objects. Videoscape-3D does
no anti-aliasing that I know of, resulting in the jagginess seen in Leo's
otherwise excellent animation. This format could be easily converted to other
formats, due to it's sime ASCII list structure.
For programs that do more sophisticated things, you might want to look into:
a) Sculpt-3D: This is a raytrace program that works with simple polygons, and
produces quite excellent results. It's data files are in it's own binary
format, and it can produce IFF-format raytrace pictures. I am tols that a
list converter is in the works for it that will let the program produce
VideoScape-format files, but that is net.rumor. As it is, the program can
produce high-quality output in a reasonable amount of time (reasonable for a
MICRO...a full-screen raytrace with 64,000 light rays still can take hours,
but there is a "quick" mode available).
b) DigiPaint: This picture can take a Sculpt-3D image (or any other 4096-color
"HAM" image) and produce some sophisticated effects. It has a superior type
of dithering algorithm and excellent auto-shading abilities...One use for it
that is commonly shown is to use the program to colorize a pre-digitized black
and white picture..The results are excellent.
Both programs are readily available, and are quite inexpensive considering
what they can do..DigiPaint lists for $50, and Sculpt-3D for $99.
Just some programs you might want to look into..
John
UUCP: {ihnp4!crash, hplabs!hp-sdd!crash}!gryphon!pnet02!jdm
INET: jdm@pnet02.CTS.COMjohn13@garfield.UUCP (08/20/87)
-- Sculpt 3D offers support for a 6 million colour framebuffer, not sure about the resolution. This could be attached to either the Amiga (is he thinking about the Mimetics thing that's been mentioned?) or something more expensive. I only browsed the manual, but it did mention something about the file format being easy to transfer and display on another system of different resolution/ # of planes. Could someone who has the software shed some enlightment? I would also be interested in hearing details about the Mimetics hardware. John -- "She's sort of a 'pit baby', with interlocking jaws. We feed her on chicken parts." "But baby-fighting has been outlawed, hasn't it?" -- Tracy Ullman describing her infant daughter to David Letterman
bart@amiga.UUCP (Barry A. Whitebook) (08/24/87)
[ eat this line, please! ] this is amiga!bart: in response to mp@pixar.UUCP (Mike Paquette) who writes >Also, has anyone writing rendering and ray tracing code looked >into antialiasing? Granted, antialiasing can be expensive in terms of >computation time, but it sure makes the end result look better. yes, i have. from experience with implementing a version of heirarchical integration at sub-pixel level using adaptive stochastic sampling i found that with from 4 to 256 subpixel samples the average pixel is only sampled about 4.1 times (that is, if one is willing to wait four times as long one can get an image "effectively" anti-aliased in an efficient fashion). this was implemented a year ago shortly after SIGGRAPH 86 but has been in storage for a year pending the availability of a frame buffer (to display my 48 bit deep images i had to settle for 12 bits by ganging THREE amigas together and taking 4 bits of red from one, 4 bits of blue from the second, and 4 bits of green from the third: possible if you work for c/a but not recommended otherwise :-) ). all parties interested in fast algorithms for ray-tracing on the amig please contact amiga!bart via email. bart