CompuServe Thread

#Metaballs and 3DStudio

36 messages in this thread
#72360From: krazy imagesDec 18, 1993 12:40 PM
Hello ladies and gentlemen. It's me, again with some exciting news. Could this be possible? I have arrived to Spain yesterday (this is cold). I will be here for a couple of days and then return to the sunny Miami (of course, these are not the exciting news). The thing is that I have hear that a group of spanish programmers are building (and are going to release soon) a module for 3D Studio that will give our beloved animation program _metaball_ capabilities. Sounds strange to me, but it have been released in TV (Canal+). They show the thing running in a Silicon Graphics, but they talk CLEARLY about it running in a PC with 3ds. To believe or not to believe? 🙂 They said that it will be a major breakthrought for the PC animation enviroment. I think that these guys are going seriously (probably the Yosters have news about this). ana /krazy images
#72369From: Danny MercurioDec 18, 1993 2:11 PM
Strangely enough, I'm also working on a "blob" or "metaball" generator for 3D Studio. Do you have any more info about these people? Do they have any samples available? Regards, Danny…
#72370From: Paul LindDec 18, 1993 2:18 PM
Danny, You're actually working on a functional metaball generator that 3DS'ers can use to generate organic sustances? It conforms to actraction levels and size like the rest of the metaball generators? Paul
#72379From: Danny MercurioDec 18, 1993 3:48 PM
Yup, I'm working on it. It will conform not only to attraction levels but repulsion levels as well. That means you can have blobs create "dents" or "dimples" in other blobs based on proximity and size. In other words, blobs can flow together or push apart based on their individual properties. And, as an added bonus, you won't need to morph blobs. They essentially morph themselves as they move about from frame to frame. Currently the program is in VERY rudimentary condition. I have to clean it up, add polygon output (convert blobs to triangles), and then port to IPAS. But first, I have to finish another program I'm working on. It should allow 3D bezier patch generation within 3DS. This tool would be great for creating flowing objects, textiles, and even attaching dynamic "skin" to objects. Talk to you later, Danny…
#72383From: Paul LindDec 18, 1993 4:14 PM
Danny, Sounds like you've got alot of work ahead of you. I've played with some of the major packages that include/support metaballs, and the coding alone seems like a job for a team of software engineers. Interesting about the morphing. So you can simply go from one form to another without worrying about manipulating vertices? Do you forsee the program being robust enough for character generation and manipulation? Paul
#72388From: Danny MercurioDec 18, 1993 4:41 PM
Hey Paul, Were you able to locate some people for those animation positions you advertised a while ago? I hope things work out well for you on that project. > Sounds like you've got alot of work ahead of you. Too much work, actually. This is the third 3D Studio program I've built in the last month or so. It's OK though, I like to keep busy. <g> > So you can simply go from one form to another without worrying about manipulating vertices? No vertex manipulation is necessary if all you're doing is moving the blobs relative to each other. The blobs are simply recalculated to reflect the changes in position and they adapt accordingly. You can even change the sizes of the blobs from frame to frame and have that change reflected as well. Actually, morphing dynamic blobs in 3D Studio would be next to impossible since you are contrained to an equal number of vertices for each morph target. Dynamic blobs have a tendancy to change the number of vertices used in thier contruction as they change shape. I don't see why blobs couldn't be used for character generation. It would take a little getting used to, though. The way blobby objects are defined is quite different from the way 3DSers are used to modelling. Anyway, the program is still in its early stage. I'll let you know more about its capabilities as soon as I figure them out. <grin> Much of that is dependent on what is available in IPAS 3. Take care, Danny…
#72393From: Paul LindDec 18, 1993 5:26 PM
Danny, As a matter of fact, two of the positions still haven't been filled, because we landed another large project. We're trying to make sure the structure and personnel is perfect for long term growth. I'm not directly involved in the process. We have a new full time business manager who is overseeing the whole thing. I've been busy evaluating new platforms, and video equipment. Paul
#72397From: MARTIN G FOSTERDec 18, 1993 6:48 PM
Danny, >> But first, I have to finish another program I'm working on. It should allow 3D bezier patch generation within 3DS. << Wow! Sounds like you are working on some fantastic things. Is the 3d bezier patch generation program equivalent to a NURBS modeler, like Alias et al? I'd love something like that.
#72404From: John EllisDec 18, 1993 8:16 PM
Martin – Wouldn't we all! <G> -JE
#72486From: Danny MercurioDec 19, 1993 3:03 PM
> Is the 3d bezier patch generation program equivalent to a NURBS modeler, like Alias et al? Well, yes and no. <g> Why yes: The core of the adaptive bezier patch code was co-developed by me and a friend who works at Alias here in Toronto. Actually, he's "Team Leader – Rendering" for Alias. (I LOVE hanging out there! <G>) Why no: In order for the patch modelling to work like the Alias NURBS, the 3D Editor would have to be re-written. I'm in no position, (nor do I consider myself knowledgeable enough), to do that. Right now, I'm waiting for the release of the IPAS 3 kit to find out how I'll implement the interactive portion of the program within 3D Studio. Talk to you soon and Merry Christmas, Danny…
#72551From: MARTIN G FOSTERDec 19, 1993 10:33 PM
Danny, well whatever you are doing it sounds terribly exciting and you get my vote for "man (or woman) of the week" for having a crack at improving the modeling capabilities of us 3ds types.
#72761From: Danny MercurioDec 20, 1993 9:27 PM
> you get my vote for "man (or woman) of the week" "man of the week" please. <g> Danny…
#72448From: krazy imagesDec 19, 1993 6:47 AM
They have their program running in the Silicon Graphics and they say that it will be a hit for the PC market. They were showing an impressive demo on the SGI (it was under the desk, you only can see the monitor). They basically explain how metaballs work and talk about 3ds, no more. I will call AutoDesk Spain and check if they know anything. If they don't, I will start to dig in Canal+, animation BBSs and the usual magicians circles 🙂 ana /krazy images
#72489From: Danny MercurioDec 19, 1993 3:19 PM
Thanks Ana, I'd appreciate whatever info you can get a hold of. Merry Christmas, Danny…
#72390From: John EllisDec 18, 1993 4:48 PM
Ana – I wondered where you'd disappeared to. <G> Jesus's brother wouldn't happen to be involved in this, would he? (tease, tease). Nobody has solved the problem of Metaballs and texture maps or have they? Will you be able to convert your Metaball objects to texture mapped splines, polygons? Can you find out more about these things? Inquiring minds are thirsty to know! -JE 🙂
#72449From: krazy imagesDec 19, 1993 6:47 AM
Hey you tell me you were in a trip and to wait for more news about you! How is everything there? I hope that all is going well in the US. I am digging now looking for this people. It has take us by surprise. We know that here are a lot of hackers with 3ds and the IPAS kit doing a lot of things, but this people with a Silicon Graphics??? Wosh! It is pretty impressive. I don't know if you know about what is the level of computing in Spain. We have pretty good graphic programmers here (a huge market of AutoCAD -lisp, etc…) and Jesus' brothers and friends, of course. A gang of hackers!!! <g> Do you know about Ricardo Montesa and his company? The people at Techex tell me their story. They only have money for an AT-Vista many years ago, so they buy it (with some video hardware). They said that they don't have the money to buy Topas, so they decided to program their own software based on the Vista. The results were impressive. They recieve offers from Techex to market the product, which were far better than Topas, but they rejected it. Today, they continue with AT-Vistas and their own software, which is beyond 3D Studio. In fact, they build software each time they want a new effect. An entire program each time. They have one of the biggest companies of animation in Europe and they don't use even one SGI. They have won a lot of prices in Imagina, etc… Ah, pionners! adventurers! …well, I prefer 3dsr3 🙂 Regarding this guys and the Metaballs, don't know any details, sorry. The demo was impressive, the usual metaball thing. Something like the effects of the metal blobs in Terminator (don't know if ILM use metaballs in Terminator, btw). They explain how metaballs work, but anything more, just said that it will be available as an "external module" for 3ds and the PC platform. I am digging now in different sources, so stay tunned for more info about it. Take care, ana /krazy images
#72475From: John EllisDec 19, 1993 1:33 PM
Ana >> Hey you tell me you were in a trip << In a trip? Only when I'm working "in" 3DS <G>. This time I was "on" a trip up North, but it was very short, not like your jaunt across the big pond. Everything is fine, and it sounds like you're having quite a time of it too. It would be fun to have metaballs technology for 3DS for liquid forms, but its application without some ability to texture map would be limited IMO. I'm not sure that its going to revolutionize PC animation but it would certainly be a step in the right direction. Anyway look forward to more news as you track these guys down and learn more about what they are up to. Interesting story about Ricardo Montesa, (great name) and the development of his company. Curious to know how far along they are with their IPAS, and what other adventures you're "in". <G> -JE
#72490From: Danny MercurioDec 19, 1993 3:19 PM
> It would be fun to have metaballs technology for 3DS for liquid forms, but its application without some ability to texture map would be limited IMO. What do you mean by "ability to texture map"? If the blobs exist as objects in 3D Studio, it should be as easy to apply materials to them as to any other object. Danny…
#72494From: John EllisDec 19, 1993 3:40 PM
Danny – I wish this were true, however from my readings on meta-balls this has not been achieved although some hy-brid systems have been developed. The problem has to do with co-ordinates to map to. If the "blobs" as you put it are translated into polygons then of course you could, but this somewhat defeats the purpose. The other approach I've seen applied is Bezier Splines which can be texture mapped but there are problems with this too because the maps can't be applied across control points. DGS had the capability to create 3d Splines which are really neat but in the end after the modeling was complete the model had to be converted into polygons. But a significant advantage is that you can get fairly fluid forms by modeling this way. You can change the color of splines but only 1 color per model and if you map them, they will only map between control points. To create the Dinosaur in Jurassic park they used a hy-brid system which incorporated thousands of meta-balls to get the bumpy-texture and color. And even then you need a fairly sophisticated modeler to have the kind of control to produce the quality of models produced by Alias for instance. Pretty intensive computational stuff! Not real suited to the PC or even a cheapy SGI which is kind of a misnomer.<G> -JE
#72501From: Paul LindDec 19, 1993 4:08 PM
John, The dino's in Jurassic park were virtually entirely polygon based. CGW once assumed that metaballs were integral to the process, but they printed a retraction the following month. I've actually seen the TRex model inside of SoftImage where the animation was applied and the models were totally polygon based. The creation process included cyberware scans of clay models, and then patch addition and refinement inside of Alias. The textures on the dino's were extremely complex Renderman shaders accompanied by a custom routine that allowed them to paint in 3D space. ILM actually made a statement that the meta technology was not developed enough to base that amount of filmwork on it. Just thought you'd like to know. Paul
#72520From: John TissavaryDec 19, 1993 6:54 PM
Very interesting info, Paul. I knew they digitized clay models, but thought they used metaballs for some of the refinements. Guess I should keep an eye out for retractions <g>. John Tissavary (La Luna cie)
#72524From: Paul LindDec 19, 1993 7:34 PM
John, Actually they didn't use the technology at all according to some of the participants. The retraction in CGW, July 1993 page 11: >Corrections >Due to a miscommunication with Industrial Light & Magic, the article "The >Magic of Metaballs" (May issue) incorrectly speculated that the dinosaurs in >the movie Jurassic Park were created using metaballs technology. According >to ILM, the production house that created the dinosaurs, metaballs >technology was not used. While ILM could not elaborate on how the dinosaurs >were created, a spokesperson did say that SoftImage software was used for >animation, Alias software was used for 3D modeling, and Parallax software >was used for painting effects. Since that point ILM has elaborated on the creation process and several trade magazines devoted to film production have presented in depth interviews of Dennis Murren and his crew, outlining the creation process. We've managed to gather close to a hundred pages of articles devoted to the work in an attempt to further our knowledge of how such a talented group went about setting new standards for character animation. Paul
#72536From: John EllisDec 19, 1993 9:02 PM
Paul – Yes I read the CGW article and they were doing more than speculating, and I also read the retraction which didn't really elaborate like you have, and then I saw a SF/X program on TV on the creation of Jurassic Park and they stated meta-balls were used and even showed an operator at the controls manipulating the model which looked like a Bezier Spline model. Then they went on to explain how many meta-balls they needed to create a model of this complexity. I'm not discounting what you've stated just letting you know that I didn't pull this out of a hat. <G>. Curiously enough Alias takes credit for the Dino work and has had several ads to imply this. Softimage has meta-balls technology and a hy-brid of it. So who do you believe? I suspect you have a better inside track on this stuff and am glad for the clarification, at least I won't talk about it without a definitive assurance that what I read and see is what its supposed to be. So do you have knowledge of texture mapping meta-balls, which is really the issue? -JE
#72541From: Paul LindDec 19, 1993 9:35 PM
John, I've sent you an e-mail. I really can't elaborate here. Paul
#72582From: John EllisDec 20, 1993 12:11 AM
Paul – I got to thinking and went back and took a look at the articles and advertisements and I saw the dark hand of the respective marketing departments, which I think pretty well explains the apparent contradictions. <G> Its what they say and what they don't say but imply. Thanks for the e-mail. -JE
#72546From: Danny MercurioDec 19, 1993 10:19 PM
OK. I understand where you're coming from now. I agree that texture mapping of any complexity is a MAJOR undertaking for metaballs if they remain in that state from the modelling to rendering stages. What I'm trying to write is a "metaball" modeller which deals with objects as blobs during the modelling and keyframing stages, but converts them to polygons for the rendering stage. This should allow complex texture mapping while maintaining a good approximation of true metaball model construction. I know it isn't metaballs in the truest sense but it should serve as a good work around for the time being. Merry Christmas and Best Wishes, Danny…
#72581From: John EllisDec 20, 1993 12:11 AM
Danny – Sounds like an ambitious project. But you're approach to the problem sounds workable. Just wonder how you're going to create mapping co-ordinates. Maybe you can take advantage of the Snapshot feature which will enable the user to create a mesh model. If you decide to pursue this approach, may I also suggest that the user be able to specify the face count. Having the ability to create a mesh model would eliminate the need to create your own method for texture mapping. Anyway best of luck, and if you need a Beta tester or just want to bounce some ideas around you know where to find me! And Merry Christmas to you. -JE
#72762From: Danny MercurioDec 20, 1993 9:27 PM
The metaballs will indeed be converted to a mesh model. That will allow for standard texture mapping within 3D Studio. I agree about letting the user determine the face count. Obviously, there would have to be some minimum number based on the number of blobs and their interactions, etc. but the user should be able to set higher face counts to achieve very smooth and fluid models. Thanks for the offer of bouncing ideas off you, I may take you up on that from time to time. Talk to you soon, Danny…
#72800From: John EllisDec 21, 1993 2:16 AM
Danny >> Thanks for the offer of bouncing ideas off you… << Feel free, sounds like you're definitely heading in the right direction. -JE
#72592From: Michele BousquetDec 20, 1993 12:52 AM
Take it easy on the guy. In Spanish, "on" and "in" are the same word. <g> Michele
#72617From: John EllisDec 20, 1993 2:43 AM
This Gal "Ana" is a friend and is accustom to my teasing. -JE
#72724From: krazy imagesDec 20, 1993 5:38 PM
Montesa aren't the developers of any IPAS routines. They produce their entire software for theirselves, but don't work with 3DS. They are probably one of the best animation companies in Europe and in the world. And their software is awesome. I just hope to hear that they have become involved with 3ds, rebuilding some of their programs for use in an engine like 3ds. However, I doubt that they will change their idea about software (exclusive tools for exclusive artists). I even think that 3ds can't provide the structure for their programs (but again, I don't have idea about high-end programming). ana /krazy images
#72741From: John EllisDec 20, 1993 8:08 PM
Ana – That pretty much clears that one up! <G> -JE
#72415From: John TissavaryDec 18, 1993 9:51 PM
Holy mackeral! Keep us informed, pleeeeeze! Will this be a full featured Metaball modeling IPAS? John Tissavary (La Luna cie)
#72450From: krazy imagesDec 19, 1993 6:47 AM
Seems that things will go this way, John. I will keep everybody informed, don't worry. ana /krazy images
#72576From: Yost GroupDec 20, 1993 12:02 AM
Metaball modelling is certainly possible to do in IPAS3. (I've heard the same rumour myself and am looking forward to seeing what they've got!) – G