sutherla@qtp.ufl.edu (scott sutherland) (04/04/89)
There was some discussion about 8 months ago on comp.graphics on something referred to as the RenderMan interface developed by Pixar. It was described as an interface between solid modeling programs and rendering programs. Some details were given as to what is covered were disclosed. I essentially blew it off at the time. Now, in the graphics supplement of the April issue of BYTE, following the poor representation of the Amiga (including non-Amiga (e.g., IBM 4-color) pics credited to the Amiga and the worst ray-tracing pic I've seen to date on the cover), is a GOOD article detailing the RenderMan interface. At first I was skeptical and felt, since it was covered in BYTE, that it would be another IBM biased article. I was wrong. After reading this article, my toes started to tingle and I got this excited feeling inside. I thought, "this is exactly what the Amiga 3-d graphics modelling and animation community needs"! We have an IFF standard, and an SMUS standard. What the 3-D community has not had is a consistent interface for all the diverse 3-D modelling and rendering packages available today. Each object format is different, since most modellers are associated with a specific renderer, and none of them are directly compatible. One valiant attempt to cure this plague is Syndesis' Interchange. While this program is a must for anyone using more than one modeling package, it is NOT a standard, it is simply a vehicle for converting one format to another. This essentially accomplishes the same goal as the Renderman interface; to allow the user to integrate the "best" modeler for his/her purpose with the "best" renderer for the project at hand. However, Interchange is NOT perfect and a new module must be written for each new product that comes out. What Pixar is trying to do with Renderman (as is described in the BYTE article) is to set a standard which will allow objects to be given attributes which surpass even the most advanced hardware available today. This way, as new technology is developed, it will be able to use the same standard. Thus, an object can be given all the attributes desired and ported to any renderer.A given piece of software, such as Sculpt-Animate, would only use those attributes which it is capable of rendering. For instance, a sphere is defined to be quartz and have a "roughened" surface. SA would ignore the refraction information and the texture mapping, whereas Turbo Silver would be able to use both of these. Videoscape 3D would interpret the quartz as a transparent "color" but ignore the index of refraction and texture data. The major software developers (Byte-by-Byte, AEGIS, and others) got together to agree on an ANIM standard (although the players are as different as snowflakes). So why can't Byte-by-Byte (SA-3/4D), AEGIS (V3D/Modeller), Impulse (TurboSilver), Hash Enterprises (Animator Series), the Forms in Flight II people (name?), Octree (Caligari) and all the others get together and accept the RenderMan interface as THE IDEAL 3D graphics interface for the Amiga, the premiere modeling personal computer? This would, in my humble opinion, be a major step for the Amiga to gain more respect from the general computer community. And we would be in some impressive company. According to the comp.graphics article, companies already endorsing the RenderMan interface include: Apollo, Ardent, DEC, MIPS Computer Systems, NeXT, Prime, Stellar Computer, Sun Microsystems, Sympolics Graphics Div., Alias Research, Autodesk, Digital Arts, Intelligent Light, Synthesis Software Solutions, Industrial Light & Magic, Pacific Data Images, R/Greenberg Associates, Thke Analytic Sciences Corporatin, and Walt Disney Corporation. Although I am only familiar with a few of these names, the ones I do know about are BIG (e.g., IL&M, Disney, Apollo, SUN). I would like to know the opinions of other, more knowledgeable persons than I (Chuck M., Leo S., and others). I wouldn't mind EMAIL responses, but I feel that this is a topic important enough for all to follow. One last word. If this standard is implemented for the Amiga, transition to the new Commodore and Impulse graphics enhancement cards would be almost painless! Thanks, Scott
bradch@microsoft.UUCP (Bradford Christian ms1) (04/06/89)
I have a copy of The RenderMan (tm) Interface specification right here in front of me. I think everyone writing graphics programs should have a copy so I urge you all to send $15 to: Pixar 3240 Kerner Blvd. San Rafael, CA 94901 Attn: RenderMan (415) 258-8100 Fax: (415) 459-4297 The specification is a 100 pages of mostly C function specifications with some general graphics info mixed in. It is a fairly technical document, but anyone writing a ray-tracer shouldn't have any problem reading it. BradCh STD. DISCLAIMER: I have nothing to do with Pixar; this message has nothing to do with Microsoft; I'm sorry AmigaBASIC sucks... *** / Back by * | * / popular demand */|\* / ***\/
rsilvers@hawk.ulowell.edu (Robert Silvers) (04/06/89)
In article <488@orange6.qtp.ufl.edu> sutherla@qtp.ufl.edu (scott sutherland) writes: > > After reading this article, my toes started to tingle and I got this >excited feeling inside. I thought, "this is exactly what the Amiga 3-d >graphics modelling and animation community needs"! We have an IFF standard, I suggested to the president of Byte by Byte that Sculpt-4D should generate RenderMan output files. (I know it could never trace RenderMan input files!). He said that he will not implement it unless he can get paid for it, and that there is not enough intrest in it to increase the sales of Sculpt. There may be a Mac version thought. I personally think it would be great. I could edit objects with Sculpt, upload them to our Pixar, and than trace them very fast. We need a standard like this and must insist on it. We have too many different formats already and interchange is just a temporary fix. Please call up Byte by Byte, and the other companies and ask for RenderMan. --Rob. Robert Silvers. I'm the NRA. Box #1003 University of Lowell. Lowell Ma, 01854 (508) 452-5000 x2233
ewhac@well.UUCP (Leo L. Schwab) (04/07/89)
In article <12642@swan.ulowell.edu> rsilvers@hawk.ulowell.edu (Robert Silvers) writes: > I suggested to the president of Byte by Byte that Sculpt-4D should >generate RenderMan output files. [ ... ] There is a problem with this: There is no Renderman *file* specification. Unless I'm very much mistaken (and I probably am), the Renderman spec simply says that, if you're writing a renderer, thou shalt name thy functions this and they shall accept these parameters and they shall do these things. This implies that, currently, the only way to express a Renderman scene portably is by C source code. I'll have to read my copy of the spec again to be sure. (It'd be kinda neat to have renderman.library in the LIBS: directory for 1.5.) _-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_-_ Leo L. Schwab -- The Guy in The Cape INET: well!ewhac@ucbvax.Berkeley.EDU \_ -_ Recumbent Bikes: UUCP: pacbell > !{well,unicom}!ewhac O----^o The Only Way To Fly. hplabs / (pronounced "AE-wack") "Work FOR? I don't work FOR anybody! I'm just having fun." -- The Doctor
shf@well.UUCP (Stuart H. Ferguson) (04/08/89)
It's always fun to be excited about things, but I don't think RenderMan will unite the Amiga 3D commumity real soon. If I understand it correctly, it has no associated RenderMan file format. Instead it is a specification of how a rendering engine will talk to the outside world -- specified in the document as function calls though I suppose they could be anything. The interface is real low level, so rendering your scene will require programming in some language or other. In fact, part of the document is a spec for the RenderMan "shader" language, a 'C'-like language with privatives for computing lighting and reflection models and a whole lot more. (The things that Pixar does with custom shaders boggles the mind.) So at the very least, a RenderMan rendering system will need a shader compiler. RenderMan might go a long way towards unifying alot of the varied jargon that comes with each renderer, though. Each 3D program seems to have its own idea about what a color is, what index of refraction is and so on. If developers followed the RenderMan conventions for these terms, then users might start to see some correlation between the parameters in the requesters and some sort of "reality" (the Renderality). The other aspect of RenderMan that's rather off-putting for a viable Amiga standard is that it's so *COMPLEX*. The minimum renderer is far beyond what we have on the Amiga today. Just coming up to that minimum level will be hard enough, let alone trying to be RenderMan compatible. These were my impressions after browsing the RenderMan doc and having some extended discussions with Those Who Know More Than I About RenderMan, so I may have it all backwards. RenderMan is a registered trademark of Pixar. -- Stuart Ferguson (shf@well.UUCP) Action by HAVOC
karsh@trifolium.SGI.COM (Bruce Karsh) (05/11/89)
In article <11262@well.UUCP> ewhac@well.UUCP (Leo 'Bols Ewhac' Schwab) writes: > > There is a problem with this: There is no Renderman *file* >specification. I have a real old version (V3.0 May 88) of the RenderMan Interface spec. It says in the introduction: A byte stream representation of the RenderMan interface will also be defined in a separate document. This byte stream is meant to serve as an external file format and network transport protocol for passing information between modeling systems and rendering systems. Does anybody know if this document is available? -- Bruce Karsh karsh@sgi.com In fact men will fight for a superstition quite as quickly as for a living truth - often more so, since a superstition is so intangible you cannot get at it to refute it, but truth is a point of view, and so is changeable.