CompuServe Thread

refraction fakeouts

24 messages in this thread
#105766From: dave aguilarJun 4, 1994 3:24 PM
I've seen some pretty elaborate refraction schemes including some mentioned here. But I have a question on either a capability I'm not aware of already in the package or a feature which might make this process easier: Is there a way to have a fixed map in space? Example… right now I can create an object (like a magnifying glass) and map what the magnifying glass is hovering over at a larger scale than what it actually IS hovering over to give the effect that magnification is taking place. The glass is curvilinear using spherical so it distorts, reflects perfectly. But when the magnifying glass hovers to another area, the map of course travels with it. It's been a bear making an animated tecture map to anticipate the new hover regeion and give the proper effect. So here's the question restated again with this example… is there a way to freeze the "map" so that as the object moves "within the map boundaries" the object acts as a sort of "mask", revealing only that portion of the map where the object exists? The map area is larger than the object obviously, and creates an area where the object can show it based on its shape and location within the map field? Everything I know about mapping locaks the mapping to the object's motion. -dave
#105780From: ARA GREGORIANJun 4, 1994 5:44 PM
dave: the two methods i outlined in this lib under "refractions" work very well. place your camera exactly in the center of the magnifying glass and assign the same motion of the glass to the camera and then hide the glass and render the background from this cameras point of view(use a very wide angle lens). then use this animation as a tex. map on the glass and render the new view that shows the glass. it works very well. also the projector light effect works good too. these two methods should get you the effect you want. as far as freezing a tex.map, i dont think there is a way to do it. reflection maps would be the closest to that but the effect is not very good for refraction. i am hoping that the next ver. of 3ds will offer raytracing or at least selective raytracing for refractions(like shadows in r3). we will see what happens. take care ara
#105906From: dave aguilarJun 5, 1994 5:13 PM
/* the two methods i outlined in this lib under "refractions" work very well. place your camera exactly in the center of the magnifying glass and assign the same motion of the glass to the camera and then hide the glass and render the background from this cameras point of view(use a very wide angle lens). then use this animation as a tex. map on the glass and render the new view that shows the glass. it works very well. */ I gave this technique a shot… it does work well… only problem is that it takes twice the rendering time (map for glass, map for overall) to get. /* as far as freezing a tex.map, i dont think there is a way to do it. reflection maps would be the closest to that but the effect is not very good for refraction. */ Yeah, I tried that too… didn't come out too well. /* i am hoping that the next ver. of 3ds will offer raytracing or at least selective raytracing for refractions(like shadows in r3). we will see what happens. */ Same here. Although a fixed map and an object "mask" capability would add some incredible capability too. Thanks for the feedback ara. -dave
#105959From: ARA GREGORIANJun 6, 1994 12:08 AM
you are wellcome ara
#106264From: david W. mennenohJun 7, 1994 10:20 AM
Dave, Here's the technique I was describing and for a magnifying glass effect it works very well. The trick is to place a copy of the paper both above and below the magnifying lens. The lens is assigned a 100% reflective material and relfects the image above it. The top "paper" object has it's normals flipped so it's invisible to the camera yet the "lens" can still relfect it and make it appear that it's magnifying the image below it. Make sure the top image is closer to the lens than the bottom one so it enlarges it. The curve of the lens surface make it look like it's refracting also. Let me know what you think of this. Dave Mennenoh
#106285From: dave aguilarJun 7, 1994 11:46 AM
/* Dave, Here's the technique I was describing and for a magnifying glass effect it works very well. The trick is to place a copy of the paper both above and below the magnifying lens. The lens is assigned a 100% reflective material and */ Oh wow. David this is a great tip. I get it. Totally! The lens reflects the inverted "above" paper image, and the image (having its normals flipped) is invisible to camera. Bravo. Excellent. Thank you. I'll give this method a whirl today. I've had fair success with the projector light method and just figured out some VPost capabilities to achieve this too, but you method sounds like the laziest (and most effective) yet. Thank you! -dave
#106377From: david W. mennenohJun 7, 1994 7:04 PM
Dave, Wow, thanks much for the words. I'm glad you like the technique. I uploaded a file called MAG.ZIP – its a 3DS project file that illustrates what I was talking about. It's not a great example as it needs some tweaking but it does demonstrate what I was talking about. Using the optimize IPAS sure is a cool way to see an objects normals – I can't wait to get the real thing. I hadn't done this trick for a while and forgot (until I used optimize) that you can't just do an object flip, you need to flip individual faces. If you simply flip on the object level the bottom ones flip up and you end up seeing the bottom. Anyway, I am planning on experimenting with your technique also – I think it would work better for doing things like a glass vase or just a glass. Good luck, and if you do download that file let me know what ya think. Like I said it isn't perfect (but it did only take me about 20 minutes from start to finish). Bye Dave Mennenoh
#106438From: dave aguilarJun 8, 1994 2:50 AM
/* Dave, Wow, thanks much for the words. I'm glad you like the technique. I */ It was deserved and the technique was strikingly straight forward. 🙂 /* cool way to see an objects normals – I can't wait to get the real thing. I hadn't done this trick for a while and forgot (until I used optimize) that you */ I checked out the demo Gary uploaded… very practical routine. Individual FACE flipping is a bear. /* bottom. Anyway, I am planning on experimenting with your technique also – I think it would work better for doing things like a glass vase or just a glass. */ It's the closest thing to an "object mask" capability I can find… we're doing some real weird stuff, and the objects actually "reveal through" these maps frozen in space. Tesserac type deal… three dimensional object spits out of 2D slot in 3D space… very strange effect. /* Good luck, and if you do download that file let me know what ya think. Like I said it isn't perfect (but it did only take me about 20 minutes from start to finish). */ I'd be happy to. It is quick. Very quick. As quick as projector method. -dave
#106550From: david W. mennenohJun 8, 1994 7:02 PM
Wow, sounds like you're doing some very cool stuff. I'd really like to see a more detailed sample of what you're doing. Any chance? I think I could learn some good techniques from it. Well, see ya. Dave Mennenoh
#106571From: dave aguilarJun 8, 1994 9:07 PM
/* Wow, sounds like you're doing some very cool stuff. I'd really like to see a more detailed sample of what you're doing. Any chance? I think I could learn some good techniques from it. Well, see ya. */ Maybe when we get it finished (big project) I can post some of the images from it as GIFs here. I'd have to get approval from the other frogmen though first. 🙂 -dave
#106775From: David J. MarksJun 9, 1994 8:21 PM
>> /* Wow… …Well, see ya. */ Dave, please PMJI here, but I think I'm beginning to understand the point that EvanY was making in his earlier messages to you. About 6 months ago, I was involved in a discussion with someone (John Foust, I think) regarding quoting. At that time, I made a conscious decision to cut back the size of my quotes. In the previous post you quoted the writer's words completely! Even if you cut back the size a LOT, you could still remind people of what was said. >> a more detailed sample… Would have very simply referred back to what was being discussed without the overkill. I am not trying to be an adversary here, I'm just trying to make a suggestion… — Another Dave
#105892From: david W. mennenohJun 5, 1994 2:22 PM
Dave, See my reply to Michael about exactly what you're asking. I suppose projector lights could work but I think my technique is easier. To make a map fixed in space just flip the surface normals of the object it's mapped to so the camera can't see it. Easy. Dave Mennenoh
#105907From: dave aguilarJun 5, 1994 5:13 PM
/* Dave, See my reply to Michael about exactly what you're asking. I suppose projector lights could work but I think my technique is easier. To make a map fixed in space just flip the surface normals of the object it's mapped to so the camera can't see it. Easy. */ Dave, forgive me if I haven't been clear on the capability I'm looking for… what I was wondering is if an object could be used as the mask itself of a fixed texture map in space. Ara's "movie within a movie" approach works if you can synchronize the motion of object with motion of camera in two passes, the projector light works if the projector image is high resolution and full strength… but I was wondering if a map could be locked in space and an object could reveal portions of it as it moved through the "map space" I don't understand the technique you describe above. The object will move around, over another surface… the "magnification" must reveal what the object is directly over. How does normal-flipping control motioned visibility if normals are a function of mapping and the map can't move? -dave
#105915From: John EllisJun 5, 1994 6:23 PM
PMJI Dave, >> what I was wondering is if an object could be used as the mask itself. << I've following some of this but not all, but by rendering an object as a "matte" object it can be used as a mask. Then in VP you can composite any image into that "mask". This would allow you to have a fixed map that would be composited into the "mask". To set up a matte object go into object attributes and click on "Matte Object". Read up on VP compositing and Alpha Mask. This does what you are asking but may not be what you are really looking for. Check it out. -JE
#105927From: Richard J. SmithJun 5, 1994 7:55 PM
Do you think an IPAS for raytracing is a good idea?
#105933From: John EllisJun 5, 1994 9:17 PM
>> Do you think an IPAS for raytracing is a good idea? << Richard, I think it might be a good idea for some people and I suspect there would be enough of a market to make it worthwhile to develop, but personally, I really like the look and speed of 3DS the way it is. If I had a budget which could afford raytracing, I could probably do it all in 3DS and do it more cost effectively. There are probably some exceptions but I haven't found myself in that situation. I think for instance Ara's suggestion for refraction would be faster to implement and render; even though it requires two renderings, than ray-tracing. -JE
#106555From: Richard J. SmithJun 8, 1994 7:16 PM
I know that most raytracing type of effects can be fuged but, I would like an IPAS to raytace an object if I wanted without leaving 3ds. Also it might take longer for me to fake it, if this fourm wasn't around.
#106590From: John EllisJun 8, 1994 11:03 PM
Richard, >> I would like an IPAS to raytrace an object if I wanted without leaving 3ds. << There have been several discussions about this and interest as well. I don't feel a need for it, so I'm probably not a good person to ask. As a production tool 3DS is really fast. If you need raytracing for a project you might look into Accurender which is reasonably priced. I prefer "fugging" <g> because I like the "look" of 3DS. I've had people ask if my work was ray-traced, because they can't tell the difference. So, its not a big issue with me. For me the name of the game is speed with good design. 3DS has the right look and performance to get the job done, and beautifully I might add. -JE
#106608From: James Coulter[Mindscape]Jun 9, 1994 1:11 AM
>> I would like an IPAS to raytace an object if I wanted without leaving 3ds. << I suggested the RXP (renderer external process) for r4. This would allow third party add-on shaders that would be selectable in the materials editor. Ray-trace could be one of them, and others like Fuzzy and Bubbly. Use of these shaders could be user selectable so that they could be applied to individual objects (as is Phong, Metal, etc. now). I just wanted to plug this idea again since you brought it up. I really want this feature! — James — MAP —
#106761From: Richard J. SmithJun 9, 1994 7:19 PM
Hi James good point.
#105936From: dave aguilarJun 5, 1994 9:30 PM
/* I've following some of this but not all, but by rendering an object as a "matte" object it can be used as a mask. Then in VP you can composite any image into that "mask". This would allow you to have a fixed map that would be composited into the "mask". To set up a matte object go into object attributes and click on "Matte Object". Read up on VP compositing and Alpha Mask. This does what you are asking but may not be what you are really looking for. Check it out. */ Will do. Thanks for the feedback, John. -dave
#105944From: John EllisJun 5, 1994 11:05 PM
>> Will do. << Let me know if that's what you had in mind. -JE
#105960From: dave aguilarJun 6, 1994 12:10 AM
/* Let me know if that's what you had in mind. */ It is. Apparently there are a lot of "methods" to simulate refraction. I uploaded a file LENS1.ZIP (.FLI) tonight showing the effect achievable through the down-and-dirty projector light methods… this is definitely the laziest "fixed mapspace" I've found yet. The movie-within-movie works well too if all motion is predefined… camera, magnifying glass… VP seems limitless in its uses, your technique makes sense and I'll attempt the same excercise again using it… no doubt it will look better. Thanks for the interest on this John. It would be great to have this capability built in, but I continue to be amazed at the ingenuity I see in this forum. -dave p.s. when LENS1 becomes available (340K I think), I'd appreciate any feedback you'd have on it.
#105973From: John EllisJun 6, 1994 3:35 AM
Dave, >> It is. << Excellent. <g> >> when LENS1 becomes available… << I'll download it and have a look and be glad to comment. >> I continue to be amazed at the ingenuity I see in this forum. << Ah yes, give a bunch of zany animators the tools and flexibility to find their own solutions and stand back! <g> Isn't it great?! 😉 And even better to have a forum where we can share them, Globally! I think about the fact that none of this was possible ten years ago, and here I am in my element. Its terrific. -JE