Attached Link: Radiosity overview

First, visit the link for the facts on what radiosity is(they even provide the mathmatics for nerd types).Now I want to talk about it(radiosity) and some myths. The first myth is that of the "holy" lightdome.A light dome DOES NOT FAKE radiosity! Fact, a light dome DOES FAKE GI( or global illumination).I noticed that some guy was claiming it as radiosity faking....its not, plain and simple.Lightdomes are never used to fake radiosity, as radiosity play little in making a shadow soft or hard(lightdomes only make softened shadows nothing else).Radiosity Only involves the bounceing of refracted or reflected light(which can produce minor shadowing in some cases, but it is secondary to the ACTUAL light that illuminates the scene).The only methods that come close are Drawbrigep,Drac,Mine,and Pumeco's(except the double banding shadows).I will use Drac's method as an example.His is faking it for one simple reason,REFLECTIONS, which are tied closely to REAL radiosity.He uses low light numbers too, as the dome is NOT needed at all.But the failing of that scene is that the "fake radiosity" is still only a reflection, just blurred and diffused.It doesnt hold up under close inspection because the "radiosity" behaves like a reflection as viwed by the camera not by actual light bouncing from the light source.On the opposing side is the true amiance button.A method that some say is defunct and non working in faking radiosity. This is not true, as I provide a scene using it,with none of the drawbacks that usually creep up(much brighter scenes).But my scenes use no reflection at all, but contains the color bleed that is seen in True radiosity scenes,and reacts more like radiosity as to where the color lands and bleeds onto other objects.So who is closer to faking it you say? Actually niether of us can claim that,yet.I believe that if the scenes could be combined in a certain way, that we would finally find an acurate way of Radiosity faking,using the "bleeding colors" of true ambiance, and drac's "muted" reflections method(Drawbridgep should have some cool tricks to add as well, such as fake light reflection).I have been combining them for the last week and the tests are rendering so I hope I can provide the scene for Drac,DB, and others to check.Soon we may finaly put all this faking crap to rest and release a set of difinitive Fake Radiosity scenes for community download.Drac,DB what do you say?

Thanks for the link. Of course you are right in that radiosity is light being bounced back from a surface and picking up (to an extent) the wavelength of the surface's color. But I have one suggestion for this post: Paragraphs! (Good natured harassment intended!)

lol, I forgot about paragraphs....but I had to rewrite it 4 times!!! as every time i cliked "post message" it went back to the main Ro'sity page.The first draft had paragraphs....but I got frustrated and just retyped as fast as I could. P.S.Drac, In no way was I trying to degrade your efforts by talking about the faults in your scene,as I was just trying to illustrate why it wasnt radiosity, the same as with my scene.

I'm glad some finally said this.

IMO one of the best methods of faking radiosity ( in Bryce ) was one developed years ago ( can't remember who ) with B4. It involved the simple procedure of putting a radius light inside the radiosity object. Then make the object slightly transparent 2 to 5 percent. Set the light's color to what ever was needed ( if the radiosity object was gold, then a gold color is used on the light ). Since the object is only slightly transparent, it appears as though it has no transparent properties. But it is enough to let the light leak out, so the objects around the radiosity object collect the light. Giving the appearance of Radiosity.

And the render time isn't bad.

Huh, did someone mention my name? I still have trouble with all the terms (GI, radiosity etc) so I tend not to use the term. I've only really just started playing. My method is just to put a strong light source in the scene and to guess where the light is falling and put more lights (at a low intensity) at those points as close to the surface as I can. The bright light will wipe out trace of the visible little lights, but they still cast shadows as reflected light should. Is it real? Is it fake? Dunno, but I like it.

well I feel sufficiently intimidated..;) informative, but what you might try (an old hand from the OT forum, I saw a lot of rants get cut or lost because the reply window was open too long (especially on high traffic days)..mebbe type it up in wordpad, cut and paste, and it's only open for the few moments..you know about the

paragraph,
break, etc tags..take care..;)

Message edited on: 07/27/2004 20:36

Dude, definetely, if this will help people understand radiosity better, i say tear the whole place apart. however, i would like to mekely state that the guy who thought out the light dome for bryce defended himself pretty well: "Global Illumination not radiosity? No, I don't think that I am confusing something, at least not as far as I can understand how radiosity works. I already pointed out that the 'semi dome lighting method' isn't radiosity in its proper sense. There are no light rays calculated depending on material properties. Therefore you can't simply put any type of light somewhere in a virtual room and expect hidden areas to be lit. Remember the tutorial is named 'simulated radiosity' or you can call it 'faked' if you like, 'not' true radiosity in Bryce. As far as I can understand it global illumination is one aspect of radiosity. Imagine you are on an infinite desert like plane (it's a beautiful sunny day) - there is only a single pillar and no other objects around. Go to the shaded side of the pillar. Do you think that it is absolutely dark behind it? It would even not be absolutely dark if you would think of a ground material which absorbs 100% of the light it receives. Well, there are still particles in the air which reflect light as well as of course the sky dome (clouds etc) itself. I think that this is the reason why some images made with this method (no other lights but the semi dome as a light source) have the mood of cloudy days - simply because you have almost only ambient light with no distinct shadow casting sunlight. Androo's image is a good example for that. Bla, bla, bla ...... sorry! What I am only trying to say is that when I say simulated radiosity I mean the very effect yield by one aspect of radiosity which normally would be calculated. In this case it is only an array of (mostly) uniform lights which cast light on objects with no regard of their material properties - basta!" drac

Very cool, I look forward to the results of your tests! And pakled is right,Notepad is your friend! Now I automatically open Notepad whenever I write more than a paragraph, and copy & paste from there to the reply window. I haven't had the frustration of having to rewrite, or trying to remember the points of what I wanted to express before the @&%#!! kicked me out to the main forums page since I started doing this,lol.

Ok here is a test of Drac's & my settings combined.The bump needs obvious tweaking.The spheres are in a "tunnel" with the lightsourse "outside" and slightly above.The tunnel ends behind the camera.It fakes light "Flooding" quite well. @SR-I was not aware of that method,it is very promising,I will add it to the scene and do more tests. @DB-I was refering to your awesome light reflection workaround(a few posts ago) as it adds higher realism to the lighting methods, yours included.Also, for not knowing all the "terms" , your imiges display careful real world observation(something lacking from alot of self proclaimed artists here).

wait... who shot who in the what now? @SR, @DB?

@Drac- you are right on GI as being a "part" of radiosity,also I was not reffering to you and your definitions but others that say its radiosity(using light domes).GI provides only part of the solution in radiosity(for the shadows) but then comes in light bouncing and the colors that are refracted, not addresed by a light dome whatsoever.Also, I say light domes are innefective, but only in a bryce environment(outside, not an indoor setting).Also,the rendertimes can be horribly long,for an effect that is produced in shorter times with diff methods.In other apps,a light dome can take 20 min in a full scene,with all the reflections and soft shadows you can shake a stick at.In Bryce(as of now....5.5 will hopefully fix the render speed)it becomes pointless for high amounts of radials.Anyway back on topic.

DB=Drawbridgep
SR=Stephen Ray
@=at, or directed at Drac...are you messing with me? I am too nerdy for jokes that dont involve boobies,LOL.

Message edited on: 07/27/2004 21:41

completely true =)

AH! Reflecting off of mirrors? Using windows instead of mirrors and duplicating the scene? And where are the bots? I've had none for hours. Drac stop messing with people. You can get in trouble that way. Goodnight fellow inmates.

Oh and one more thing. Stephen Ray! That is so simple and under the right circumstance works really well (I just tried it). Never heard or thought of that before.

Using windows instead of mirrors and duplicating the scene.....its classic I tell you,CLASSIC! OT: for all the bryce images in the gallery,maybe 10 out of 100 come to the forums and post stuff(anything for that matter), whats up with that? I say this because of Dracs recent pic with fairy breasts..lol ,Do people just read stuff and not respond or ask stuff?

TB: pretty much. that pic is classic. 500+ views. I know what I'm making all my thumbnails from now on :D goes to download some more pr0n

I made some pretty balls too!

For the window instead of mirror go see my gallery.
For the boobies, also go see my gallery.
For the posting anything, see here.

Message edited on: 07/28/2004 01:59

But.. .at the end... is it realy so important? light domes, true ambience, reflecting surfaces.... We are not Maya of Lightwave freaks who spend more time in math calculations than image compositing. THe only important think is the result, if the image looks convincing it's ok, if the image has some art inside it's ok, if the image has some boobs inside better! Twisted your link is really cool for the guys who didn't know that's radiosity... but your text and what you're trying to point out is not so easy, moreover adding some square lights or piramd light (ore radial light like rochr does) in the right places works better than 2-3 days renders with light dome and true ambience. just my 2 cents (Good natured harassment intended!) we can just add a new challenge: "Best fake true radiosity boobs" (fake is the radiosity not the boots)

Attached Link: http://www.renderosity.com/viewed.ez?galleryid=694682&Sectionid=8

hmmnnn...a lot of what you talk about is way over my head when you-all start talking about radiosity :o( The attached (the first pic I came across in the Photography section) illustrates what I understand as radiosity - the light bouncing back into the shadows of the cloth material. Seems to me what would be very difficult to fake (I do loads of life-drawing, so I'll have to use a life-drawing example) is the way...if you have one light-source, a window or whatever, light falling on the side of a body, and the light is bouncing back from the rib-cage to the inside of the arm. So the outside of the arm is illuminated, the rib-cage is illuminated, there's a dark area running down the centre of the arm, then reflected light flooding back from the rib-cage to the inside of the arm. I notice that professional photographers use a sheet of white polystyrene to flood light back into the shadows. Don't know if I'm adding anything to the discussion, or just mouthing off.

At work now and very little time to write. Suffice to say for now that this interesting thread contains myths and non-facts a plenty. "The first myth is that of the "holy" lightdome.A light dome DOES NOT FAKE radiosity! Fact, a light dome DOES FAKE GI( or global illumination)." This is not so. If a light dome can't fake radiosity (which it can't) then it can't come anywhere near close to faking Global Illumination. Global Illumination is a 3D term that refers to any combination of rendering methods that duplicate every aspect of real light in a real scene. This will include radiosity, specular reflections, caustic reflections and transmissions, etc, etc, etc. The word 'global' in Global Illumination does not relate to anything of a spherical nature (such as a dome); it simply means 'all encompassing'. "The only methods that come close are Drawbrigep,Drac,Mine,and Pumeco's(except the double banding shadows)." Pumeco's method does not produce "double banding shadows" - you've simply misinterpreted his render/s. I feel that you've missed quite a few methods that do a reasonable job of faking radiosity.

The simplest, or you might say, the crudest, is just setting the ambience to a respectable value. This counters the problem mentioned above of a column in the desert. Render this with one light and no ambience and you get the totally black shadow. Put in a little ambience and the problem goes. It may not be mechanically like radiosity, but it has a similar effect in terms of getting everyday scenes to look that bit more natural.

oh well.....heres some nudity.

Message edited on: 07/28/2004 21:02

great finaly nudity always better then paragraphs imho.:P why not add that rays tip into the mix just incase not enough fake light is generated just put a lamp inside that partly transparant object in the desired color that example picture rocks. I agree with vaques about picture being more important then maths. So plz now back to how to make brickwork walls in the dte with or without fake radiosity. btw radiosity looks suspiciously much like renderosity lol

"lol, I forgot about paragraphs....but I had to rewrite it 4 times!!! as every time i cliked "post message" it went back to the main Ro'sity page.The first draft had paragraphs....but I got frustrated and just retyped as fast as I could." It is ALWAYS useful to type your posts into another editor (notepad works)... then "refresh" the Renderosity page, paste in your post and away you go.

twisted the light in your experiments, from a light-feticist point of view is really good, but why has it that grain?

We are not Maya of Lightwave freaks who spend more time in math calculations than image compositing. Hey! Keep us LightWave users out of this. The software does the the calculations, not us. And I think you might mean image composing, not compositing. There's a big difference.

@Vasquez- the grain is due to the RPP(rays per pixel), its only at 9 in the picture.RPP of 64 and up is good, but this is only a test to see what happens when all the methods are jammed together.Oh, render time was like 45 min at 1024x768,at 64 RPP it estimated 1hr 55min.......super fast in my opinion. Note:all effects but DOF was on.

Phantast wrote: "The simplest, or you might say, the crudest, is just setting the ambience to a respectable value." This is, as you say, simple and crude. It's very inexpensive in render terms, but the results usually leave a lot to be desired. The areas illuminated by just the ambient aspect of the material have a completely dead look, with no real shadows or contrast. However, using ambient materials but with the overall ambient color set to black whilst using True Ambience will result in very real looking indirectly lit areas. Perfect black shadows where appropriate; a proper gradient from light to dark; and, of course, radiosity like diffuse light and colour bounce. And not forgetting relatively horrible render times. ;-)

Here's a real blast from the past. This is pure blurry reflections - no True Ambience. It's quite a powerful technique for faking radiosity and suitable for many scenes. Again, just one light with soft shadows. The major downside with blurry reflections is when rendering smooth surfaces. The reflections aren't quite blurry enough at their maximum setting, and it shows. Shadows still look like reflections. Rough surfaces, like here, hide it fairly well. But there are none of the problems with True Ambience, except for long render times.

^_^

@.@