Path: utzoo!censor!geac!torsqnt!news-server.csri.toronto.edu!cs.utexas.edu!wuarchive!usc!apple!vsi1!zorch!xanthian From: xanthian@zorch.SF-Bay.ORG (Kent Paul Dolan) Newsgroups: comp.sys.amiga Subject: Re: 3D Objects Message-ID: <1991Jan8.045918.5178@zorch.SF-Bay.ORG> Date: 8 Jan 91 04:59:18 GMT References: <6715@crash.cts.com> Organization: SF-Bay Public-Access Unix Lines: 53 seanc@pro-party.cts.com (Sean Cunningham) writes: >In-Reply-To: message from hastoerm@vela.acs.oakland.edu > You're probubly right. If you watch closely, a lot of the segments > showing the ship flying are the same from episode to episode. They > could render a whole collection of views of the ship from different > angles, in front of different colored planets, and then just use those > throughout the series. I don't remember where I learned all this stuff; maybe at SIGGRAPH; anyway, some more interesting information. You don't directly capture onto film the model "in front of diffeent colored planets". You film the model against a deep blue background that happens to look blacker than black when photographed (it reflects no wavelength the film happens to record). This is half your effort. It's also interesting in passing that the model hangs from a thread the same unphotographable color, though it is still, of course, opaque. Then you make a second, separate film of the backdrop, the "colored planet". Next, in the days when I learned this, came a very expensive hand effort called "matting", that blotted out a model shaped hole in each frame of this second film (with paint, I think). This could perhaps be done cheaper today with a computer process, since no raytracing is involved, but in those days, the matter was a highly paid, highly respected part of the film effort, and the success of the model parts of films was attributed in the trade primarily to the skills of the matter, not to the camera operators or model makers. Then the backdrop was exposed onto yet a third film with the matted area of course not exposed. Last, the film of the model was exposed onto the same film in a double exposure, fitting neatly into the unexposed area. This could be repeated to quite a few layers of complexity, reminiscent of animation cels in concept, but completely different in actuality since transparency and opacity change importance in the two methods. (This takes thinking about.) The more layers, the more expense, and the more matting needed. > I don't know how much miniatures (if you can call a 6ft long model a > miniature) like this cost, so CG may or may not have been cheaper to > produce. Well, you look at two curves, model makers salaries going up and cpu cycle costs going down, but you have to leverage your computer model builder's effort a lot before the curves intersect, even today. Kent, the man from xanth.