CompuServe Thread

#glassfra

27 messages in this thread
#109325From: DANIEL PASHOLKJun 23, 1994 3:59 PM
I have been trying to model a faceted clear glass object and haven't been able to get the facet reflecting effect of individual facets or model realistically looking through glass onto interior facets. My question therfore is: Will future versions of 3DS have in the materials editor a provision to state the refractive index of a material? It seems to me that this feature would make MUCH easier the design of glass objects and translusent materials. Couple this with ray tracing and it seems to me my problems would be solved. …………….917134832500.1200 N81N GENOA………………914089431231.
#109371From: david W. mennenohJun 23, 1994 10:22 PM
Daniel, You are talking exclusively about ray tracing, not the ability to couple the renderer with ray tracing. Refraction itself implies ray tracing. Ther has been a lot of talk recently about ray tracing engines for 3DS, though I myself am not too excited about them. IMHO, the ray tracers that will be available are not going to be able to touch the rendering quality of the 3DS renderer. They just can't. Take a look at POV. While it does Ray trace it pales in comparison as far as color and AA go. Besides, they'll be SLOOOOWWWW…. Not what production animation is all about…. Dave Mennenoh In one of those moods….
#109494From: M. G. BATCHELORJun 24, 1994 3:43 PM
PMJI, but: << Daniel, You are talking exclusively about ray tracing, not the ability to couple the renderer with ray tracing. >> Actually, the best renderers available selectively combine scanline & ray-trace techniques. << IMHO, the ray tracers that will be available are not going to be able to touch the rendering quality of the 3DS renderer. They just can't. >> What do you mean by this ? …i.e., the RT's that are *_ going to be available_*, or RT renderers in general ? If the latter, I disagree. While I've made supportive statements re scanline renderers here before, personally, I'd prefer a selective RT utilizing a great scanline package as well – it's there if/when you need it. BILL
#109967From: david W. mennenohJun 27, 1994 10:31 PM
Bill, Thanks for the reply and info. regarding different renderers. I agree with you on the ability to selectively raytrace. This would be a great benefit actually and I would love to see it implemented within 3DS. Enough of that. I hope I thanked you for your very gracious upload of the .JPG's of water caustics for me. They were great – just what I needed. I used Apro to combine them into a .FLC instead of using an .IFL – no different I guess, but it seemed easy enough. Talk to ya later. Dave Mennenoh
#110228From: M. G. BATCHELORJun 29, 1994 2:49 AM
You could use the water maps ? Great. You're welcome. BILL
#109404From: James Coulter[Mindscape]Jun 24, 1994 1:36 AM
Yet another plug for the RXP!! (Renderer eXternal Process: Plug-in shader for the Mat. Editor.) — James — MAP —
#109450From: Jonas Ruikis [ADESK]Jun 24, 1994 11:24 AM
Daniel, << Will future versions of 3DS have in the materials editor a provision to state the refractive index of a material? >> Autodesk does not discuss future product features or shipdates. jonas[adesk]
#109459From: ALAN IGLESIASJun 24, 1994 11:56 AM
Daniel, you could try the Place Hilight command with dedicated lights, along with playing with the size and shape of the specular hilight in your "glass" material. With a little fiddling, you should be able to get an acceptable "facet hilight" effect. Hope this kelps, -Alan Iglesias
#109466From: Mike White SunDog Pro.Jun 24, 1994 1:03 PM
Hi I just had to put my two cents in. What about ray traced object exclude just like with shadows. If you wanted an object closer to the camera veiw ray traced you just choose which object you want ray traced. Then with shadows have an option to render the during or after sceen rendering. If you choose to render the shadows after you could render them in an alpha channel and compose the main rendering with the alpha shadows. This would be cool. Also is camera map projection. To simulate different kind of lense and optical effects. Mike White [SDP]
#109516From: ALAN IGLESIASJun 24, 1994 5:33 PM
I agree that those are cool ideas, Mike. Good wish list stuff. Incidently, I've been looking at your routines and they seem to be pretty cool too! Probably be using some of them in production soon. -Alan
#109568From: Mike White SunDog Pro.Jun 25, 1994 3:03 AM
Thans it will be nice to see them in production, to let you know our routines are based on Ray traces texh.. Which are a new kind of SXP. Mike White [SDP]
#109584From: Jim LammersJun 25, 1994 11:28 AM
>>Will future versions of 3DS have in >>the materials editor a provision to >>state the refractive index of a material? >>It seems to me that this feature would >>make MUCH easier the design of glass >>objects and translusent materials. >>Couple this with ray tracing and it >>seems to me my problems would be solved. For animations, usually the frame time for raytracing is too high. For single frame work it can be pretty useful, though. You can raytrace an object right now by translating it to PoV or Vivid (two widely available raytracers) format with 3DS2POV. You have to rebuild some of your textures, though. Render time can be pretty long, depending on the scene. Something with lots of reflection and refraction in it could take maybe 8 hours on a 486/66. They do look pretty incredible, though. I think there may be a workaround that would be pretty effective in 3DS, though. I've seen mention of creating a flic from the refracting objects point of view, and then mapping that to it mirrored (or something like that, not sure). What I would try is creating a cubic environment map and reversing the side assignments. Then use that as a reflection map for the semi-transparent material assigned to the refractive object. The new Topas uses an interesting trick. They create a cubic environment map around the refracting object, and when the eye ray hits the refracting object, they bend it with the specified IOR into the environment map and use that color. It's not raytracing, but in an animation, it would still look pretty great. Jim
#109969From: david W. mennenohJun 27, 1994 10:31 PM
Hey Jim, Sorry I was a little on edge last week when I responded to your message about "hair". It jsut seemed you missed my point and that's fine. I have never even tried to do real hair as it would be just too much work. The point I was making is that rapture would most likely work in cases where you didn't need to be up close and personal. Oh well, I'm sure you can relate to being a little edgy while replying to messages at 2am. Dave Mennenoh
#110197From: Jim LammersJun 29, 1994 12:06 AM
>>well, I'm sure you can relate to being a little edgy while Hello, David. Thanks for your nice letter. Sure, I can relate. Don't know if we'll find a middle ground on this "IXP-for-hair-effect" situation. I've done extensive work with all the currently available image processors (both IXPs and batch processed Gallery effects/Kai's Power Tools, etc) and never seen a process that could avoid a "flat" look in an animated 3D scene. If the renderer could send some of the scene info to the IXP, maybe it would be possible. However, as it stands, using alpha to use the effect on a single object would have a 'rubber-stamp' effect in the course of an animation. Stills might look ok, especially with a little hand-retouching in Photostyler. I'll recant immediately if/when you can show me a video with IXP'd in hair that looks 3D… <G> BTW, I do get on the system every other day or so. Also, I read the whole thread before I replied. When I saw the general consensus of the thread leaning toward the conclusion that an IXP could do a passable 3D fur object, I felt I had to put in a word, and I picked your letter to reply to. Didn't mean to offend…. Regards, Jim
#110390From: david W. mennenohJun 29, 1994 7:15 PM
> I'll recant immediately if/when you can show me a video with IXP'd in hair > that looks 3D… <G> Doubt you'll be recanting anytime soon Jim. Obviously you've done a lot more work than I have in this area and hopefully you'll figure out something that works better than the currently available options. Well I'm glad I didn't offend you, you must have a good sense of humour – a good thing in this area. Thanks for the reply. Dave Mennenoh
#110423From: Yost GroupJun 29, 1994 10:11 PM
I saw the Metrolight "JuJu Shampoo" video again today (with a bunch of other folks) and 4 of the people in the room refused to believe that the long, lustrous, silky hair in the video wasn't real. I think it's the best fake I've ever seen. – G
#110465From: MARTIN G FOSTERJun 30, 1994 4:41 AM
Gary, >> room refused to believe that the long, lustrous, silky hair in the video wasn't real. << Slight deviation from the subject, but have you seen the latest stuff from Acclaim? It's a sequence from a game in progress called "Alien Trilogy", I believe. It was shown at Summer CES and I saw a camcorded tape taken off the tradeshow monitor. It just blew me away again! These guys must have the best (proprietary) motion capture facility in the world. It's a cinematic sequence of about 4 space marines getting into a serious "disagreement" with the Alien we've all come to know and fear. Their system seems to capture every nuance of human motion, unlike any other system I've seen. I believe it's a 4 camera infra-red system, but may be wrong (wouldn't be the first time <g>).
#110509From: Yost GroupJun 30, 1994 10:16 AM
I haven't seen Acclaim's newest work yet, but I'm looking forward to seeing it at Siggraph. Those guys are the BEST at biomechanics. – G
#110566From: MARTIN G FOSTERJun 30, 1994 2:07 PM
Gary, >> but I'm looking forward to seeing it at Siggraph. << Well, they've outdone themselves again. Up until now I hadn't thought computer generated actors (synthespians) could convincingly replace real ones (except in distance shots). Now, I think they could. Many people at the show were convinced it was a clip from the actual "Aliens" movie??!
#110652From: Yost GroupJun 30, 1994 10:44 PM
Give it another 3 years before you won't be able to tell if the video that you're watching on the evening NEWS isn't simulated. A very very scary thought. We're going to have to have some very firm laws regulating the use of simulated imagery. Think about it (and shudder). – G
#110670From: M. G. BATCHELORJul 1, 1994 12:50 AM
<< Give it another 3 years before you won't be able to tell if the video that you're watching on the evening NEWS isn't simulated. >> What do you mean 3 years ? I personally know the team that renders Dan Rather every night. (bizzare rumors are better than flaming, right ?) BILL
#110680From: James M. MurphyJul 1, 1994 1:41 AM
G >>>Give it another 3 years before you won't be able to tell if the video that you're watching on the evening NEWS isn't simulated.<<< Oh my GAWD!!! You mean….Bill and Hillary are REAL!!!! ARRRRRRGGGGGG! Murph'
#110727From: Yost GroupJul 1, 1994 10:24 AM
We'll never be able to legislate media illusion. Either our perceptions will alter to keep up with the media, or our reality will alter to adapt to the perceptions. – J
#110771From: MARTIN G FOSTERJul 1, 1994 1:34 PM
Gary, >> Give it another 3 years before you won't be able to tell if the video that you're watching on the evening NEWS isn't simulated. << I think that's why the Acclaim stuff makes me somewhat uneasy. I have uncomfortable feelings about how this could be misused, but I'm still excited about the possibilities.
#110479From: Accu-WeatherJun 30, 1994 7:54 AM
>…"JuJu Shampoo"< What is this? I suspect it's a demo clip & not on TV or anywhere someone could easily see it. Yes, no, maybe so?
#110510From: Yost GroupJun 30, 1994 10:17 AM
It's a piece done by Metrolight Studios in LA, and was shown in the Film & Video Show (Electronic Theater) at Siggraph/Anaheim '93. I think it's on Siggraph's VHS copy of the ET, and that tape is available from ACM Siggraph (I don't have the number, but I'm sure someone on the forum does). – G
#110694From: Accu-WeatherJul 1, 1994 6:32 AM
Thanks. I'm gonna make sure I go to this year's ET; probably be combs going thru the hair this year <g>.