[loren s]

I have the grandest question of all for 3d. In the Mac OSX even in just Panther if a user has a good enough card that can run Quartz Extreme why is there no app that lets a user use and create content in realtime with this tech?

Story: I installed developer tools and went to check out the OpenGL stuff just for kicks and what did I find, some really super fast programs that render shadows normal maps grooves, lighting, everything that takes an app like Blender MAya Lightwave Cinema and others a day and a half to render on d@m frame give or take.

Developers: But the issue is that these programs are just developer tests and nothing more. Why ? Why is there nothing that can let a user create and animate in this ?

Artists: Would it help any at all to have tools that can render in real time frames compared to something that just chugs away time rendering one frame after the other forever...... ?

[Chuckit]

Not again...

[Deimos]

That's because you're comparing OpenGL renders to vastly more complex renderers such as Mental Ray/Renderman/Lightwave's, etc.

That's just one reason.

There's also the matter of geometry in which the typical OpenGL scene will have low-poly count models. So, taking the matter of a complex 3D scene in which you are using far better, and more complex rendering algorithms to work on vastly more complex geometry (Polygons, NURBs, Subds, etc), combined with the rendering calculations you require (multiple light sources of various kinds, photons bouncing everywhere, depth maps, normal maps, etc.), then you can imagine why there is such a disparity in rendering times.

OpenGL has its place in 3D, but for final output, it's not really an option if you're aiming for the kind of quality you see in movies.

Believe me, if real-time rendering of very complex scenes was here today, we'd know about it. Some try to do it in hardware (ART), or in software (RenderGL, Worley's FPrime), but they're not real-time, and if close, the quality and complexity of the scene is probably slight.

[loren s]

Originally Posted by Chuckit

Not again...

??? if I posted before then there is a reson I am still after the info.

Originally Posted by Deimos

That's because you're comparing OpenGL renders to vastly more complex renderers such as Mental Ray/Renderman/Lightwave's, etc.

That's just one reason.

There's also the matter of geometry in which the typical OpenGL scene will have low-poly count models. So, taking the matter of a complex 3D scene in which you are using far better, and more complex rendering algorithms to work on vastly more complex geometry (Polygons, NURBs, Subds, etc), combined with the rendering calculations you require (multiple light sources of various kinds, photons bouncing everywhere, depth maps, normal maps, etc.), then you can imagine why there is such a disparity in rendering times.

OpenGL has its place in 3D, but for final output, it's not really an option if you're aiming for the kind of quality you see in movies.

Believe me, if real-time rendering of very complex scenes was here today, we'd know about it. Some try to do it in hardware (ART), or in software (RenderGL, Worley's FPrime), but they're not real-time, and if close, the quality and complexity of the scene is probably slight.

Deimos: You are very correct about Lightwave and menta ray and others on having a better render engine. But some of the stuff I have seen from todays Open gl just kicks the pants off of some renders that can take ages to render. And what of Preview renders in these basic cpu render apps? If you set the preview render in an app today to be the fastest it can be you a: lose lots of info about the final image and B: still have to wait for it to render at the very least a second for the render but increaseing items like shadow and lighting GI AO will increase those test renders to half minutes. These add up so much when just doing basic work.

Another item is Open GL uses low poly meshes, but Normal maps and bump maps fix the low poly mesh issue. So detail is not lost in open gl renders. Though adding more and more bits like multi sampleing will increase the animation render time it will still beat any basic render animations time.

I disagree with the idea that if it was possible that we would see it already, Case example Photoshop CS 2.0 is coming out just after Tiger. One would have thought they were going to use afew of the newer techs in Tiger core image, but zero word of such speed increases have been heard of even though they have had the time to intergrate the features into the app for near a year now.

Companies will not movve on anything major unless they really see someone as a threat to their business. I fear and hope that Apples secret app code named Fantasia is a 3d app that does just this, as they tend to pick up the slack when developers do not see the method of user need in apps. I fear it because it would prove my point and thoughts for all of this.

So the question still stands even if a scene is a basic one object item, where is the app that can let an artist tinker with a low poly model with normal mapping and add things like a few more low polys and stick in real time GI (even if the Gi is less quality) . The app would not even have to worry about cards since there are so few for mac in the first place and this would only be a mac osx app, no need to worry about windows support.

[Millennium]

Originally Posted by loren s

Deimos: You are very correct about Lightwave and menta ray and others on having a better render engine. But some of the stuff I have seen from todays Open gl just kicks the pants off of some renders that can take ages to render.

It doesn't matter. Lightwave and its ilk use completely different methods of rendering from your average 3-D card. They just plain don't translate straight across.

Another item is Open GL uses low poly meshes, but Normal maps and bump maps fix the low poly mesh issue.

No, they don't. You can't just convert normals and bump maps to meshes like that. The mathematics involved in the different styles are completely different, and 3-D cards aren't made to handle the type of mathematics used in the engines you mention.

I disagree with the idea that if it was possible that we would see it already, Case example Photoshop CS 2.0 is coming out just after Tiger. One would have thought they were going to use afew of the newer techs in Tiger core image, but zero word of such speed increases have been heard of even though they have had the time to intergrate the features into the app for near a year now.

Apples-and-oranges comparison. Using a new API is relatively easy. Converting between completely different mathematical models is not. What you're suggesting is not unlike representing integral calculus using eighth-grade algebra.

[loren s]

Originally Posted by Millennium

It doesn't matter. Lightwave and its ilk use completely different methods of rendering from your average 3-D card. They just plain don't translate straight across.

No, they don't. You can't just convert normals and bump maps to meshes like that. The mathematics involved in the different styles are completely different, and 3-D cards aren't made to handle the type of mathematics used in the engines you mention.

Apples-and-oranges comparison. Using a new API is relatively easy. Converting between completely different mathematical models is not. What you're suggesting is not unlike representing integral calculus using eighth-grade algebra.

? ? forgive me but I think what your trying to say is that developers just dont want to try ? Or can't ?

It can be whatever math system there is in the world I don't want to go to deep into the details that I can never know or understand, I just see the raw power that is sitting dormat and look over to the side at the all powerful developers that can harness and tame the stuff to give to users. But I am seeing those that can control this power just picking their noses and arguing with each other instead.

I dont want to make anyone mad and I know some of my first words are always a little rude or confusing but I need to get a deveolper that can and will work at new ideas. I dont know why I see this and understand all of this in different ways but I HATE when an app is finally released that I and a others begged for way before it came out. But also happy when it does ..

[Chuckit]

Loren, how many threads are you going to make lecturing developers about technologies you clearly don't understand? If you're such an expert, make it yourself. If you admit you don't know, why are you arguing with those who do?

I really don't want to be rude, but threads like these are just ridiculously arrogant. Yes, everything looks simple when you don't know anything about it. That doesn't mean it is.

[loren s]

Originally Posted by Chuckit

Loren, how many threads are you going to make lecturing developers about technologies you clearly don't understand? If you're such an expert, make it yourself. If you admit you don't know, why are you arguing with those who do?

I really don't want to be rude, but threads like these are just ridiculously arrogant. Yes, everything looks simple when you don't know anything about it. That doesn't mean it is.

I have not made a post in near two months for any of these topics.

Why quest is just a basic need and search. If I or others do not ask and search then how would you developers even know what someone wants? Or is it all just... eh I have no idea...

what is the harm of asking? Why am I a villian if I am asking? What can I do to stop wasting time asking, and actually get some tiny bit of the app started? I can not code, I do not wish to code, I would and have wasited to much time trying to code a subject that I am in no way proficient in. Instead I can create art and design. Why can I not ask for better faster tools? Why must I rely on big fat corparations to deside what tools I can and can not have and aat what speed and design of interface I must use?

I am full of questions and empty of answers. And all I seem to ever get from 99% of developers is "it can't be done" When it is right there somehow even if it is in peices I see it I have used bits of the tools in little developer school papers and test projects.

I have seen all real time Voxels of clouds, smoke, rain water caustics, particules of all sorts. Distortion mesh softmesh that you can poke and reform like jelly. Springs, ode, hdr lighting Gi, AO, radiocity, raytraceing is still very very slow, normal mapping, displacement mapping, parlex mapping, painting, gui creation, and so much more. But all broken up in little apps that only have run time and no editing maybe some free cursor control spinning and a little mouse control click and drag....

If you want to hel[p please do. If you dont, then please stop attacking me and my thoughts. What sounds foolish and impossible to you is just a step away for me. Maybe you can or can not code it, but no matter, if you dont try then it will never be created by you instead someone else will do it one day but I want that day much sooner.

Nothing is over night but why mock it and call it impossible? You can not see the M or string theory of the universe but that does not dare stop scientist from trying. I dont know I quess I am just fishing in the wrong fish ponds...

[Chuckit]

Originally Posted by loren s

Why quest is just a basic need and search. If I or others do not ask and search then how would you developers even know what someone wants? Or is it all just... eh I have no idea...

what is the harm of asking? Why am I a villian if I am asking?

You are not just asking. You ask a question (why something hasn't been done) and then accuse people of being lazy and/or wrong when they give you the answer.

Originally Posted by loren s

I am full of questions and empty of answers. And all I seem to ever get from 99% of developers is "it can't be done" When it is right there somehow even if it is in peices I see it I have used bits of the tools in little developer school papers and test projects.

It sounds to me like you have your answers and aren't looking for any from us. That's my problem: You don't know this stuff, but you argue as though you do. You see Sketch and think, "Gee, I'd like to see an HTML authoring app based on that," but that's just superficial similarities that have nothing to do with the implementation. But since you don't know anything about the actual implementation, you think that's enough, and accuse those who know more of being lazy or unimaginative when in actuality you just don't know what you're talking about. That's what bothers me.

To be honest, I don't know that much about 3D graphics implementations, which is why I'm not directly chiming in. But it's the same kind of "But it looks the same!" logic at work here, the same stubborn refusal to discuss the actual technology involved because you don't know about it.

And just to reiterate, I'm really not trying to be mean. I just don't like these aggressive accusations with no intelligent discussion.

[Deimos]

Originally Posted by loren s

So the question still stands even if a scene is a basic one object item, where is the app that can let an artist tinker with a low poly model with normal mapping and add things like a few more low polys and stick in real time GI (even if the Gi is less quality) . The app would not even have to worry about cards since there are so few for mac in the first place and this would only be a mac osx app, no need to worry about windows support.

Millennium succinctly answered the other part of your question, so I won't add to it. What I'd recommend, is to have a glance at the OpenGL forums, familiarise yourself with the technologies that make up OpenGL, then have a peek at the numerous Siggraph papers (where so much rendering tech. is proposed long before application vendors implement them.

You'll see a gulf in difference between what a complex renderer, and OpenGL are capable of, and their intended use.

To answer your question above, it's already possible to have near real-time rendering on current hardware/software. It's mainly implemented through a technique called "iterative rendering"; the image is built up over time and each successive pass refines the accuracy and quality. Of course, for low-poly scenes this is quite nice, and it's basically what you're wanting above. The more complex the scene gets, the longer it takes to reach the highest quality preview; but if you're needing instantaneous results, then even the first couple of passes will be good enough to see if something is working, or not.

Where's OpenGL in this? Nowhere, it's not needed. The type of renderer I mentioned above will not only give you very quick results, but more importantly, they will be closer to what you're probably intending since you're using the same renderer as that of your final output. In this scenario, you can throw in a cube, subdivide it, texture it, even add caustics, GI, etc. You will get instant results that actually aid you in what you're attempting to achieve. OpenGL will not do this for you, unless you're looking for game-type quality.

[Millennium]

Originally Posted by loren s

? ? forgive me but I think what your trying to say is that developers just dont want to try ? Or can't ?

There are certain types of problems in computing that no one has yet been able to solve. This is not due to lack of trying; many groups have been trying to solve them for many years, and millions of dollars in prizes are set to go to the research teams who can solve them. But as far as we've been able to tell thus far, even with all of the research that has been done over the last 70-odd years, they are beyond the capabiliities of current computers. It's not a matter of throwing more effort or computing power that's the issue: the fundamental models of computing that we currently use are actually not capable of it. In order for this to happen, something is going to have to change at a level so low that much of current computer-science knowledge won't even apply. What you have suggested is such a problem.

This is not to say that such a revolution will never occur. There are people working on new models of computing all the time, and though I don't know of any that have this specific kind of problem in mind, they might be capable of it. A DNA computer has actually been constructed, for example; although it isn't programable at this point in time, it cannot be beaten in Tic-Tac-Toe (though it has been played to a draw before). It also doesn't resemble our current computers at all; it exists as a series of chemical vats. More recently, someone has managed to build a crude computing device out of light which has been slowed down so much that you or I could almost outrun it. You may have heard of quantum computing; this has somewhat less promise for what you describe, since it uses the same models we use nowadays, but it promises parallel computing on a scale equivalent to many thousands of processors.

But these things are in their infancy. None of them can even be programmed yet; they are built to do one specific thing and can do nothing else. Of course, our own computing models were at that stage once, a long time ago, but it took many years just to get them to that point, and many more to get that technology to the point where it is today. Twenty years ago, the entire Internet could have fit on a single dual-layer DVD, had such technology existed back then.

It can be whatever math system there is in the world I don't want to go to deep into the details that I can never know or understand, I just see the raw power that is sitting dormat and look over to the side at the all powerful developers that can harness and tame the stuff to give to users.

That's just it; it's not a matter of computing power or programmer effort. The fundamental model of computing that we use isn't enough for this kind of task.

Many of the most ancient mathematicians did not know anything about the concept of zero. It is a simple thing to think of today, but people back then could not even conceive of the idea. Even the ancient Greeks and Romans, advanced as they were, did not have it. They did many great things without zero, but its introduction was nothing less than revolutionary. It enabled many new discoveries that simply could not have taken place beforehand. Some people even committed suicide over it; their worldview was that disrupted by the idea. And yet now, to us, it is nothing.

What computing needs, in order for your idea to work, is a concept as revolutionary as zero was. Such a concept may exist, somewherer in the universe, but I know no more about it than Pythagoras knew about zero.

But I am seeing those that can control this power just picking their noses and arguing with each other instead.

We can control this power, but the power itself has limitations that we can do nothing about. The only way out is to find new powers, and although there are people working on that it will be many years before their work yields usable fruit.

As a tech-head myself, it is frustrating to admit this. But computers can't do everything; not yet.

[loren s]

Wow Millennium and Deimos, ... This is why I ask around, i seek detailed answers like these. I cant really quote anything since it is so full of very useful examples and current works that I have to sit and think about for days.

But I do have to say I never heard of iterative rendering before or have misplaced the exact method. I must look into this, but since it does exsit what is the chance of seeing it anywhere, even in basic papers tests?

Chuckit I know you mean well, I just guess we have conflicting repore with each others thoughts and how to express the written word.

As for anything else I guess that concludes my post, it looks like I will not find any developers since lots of stuff is still missing for it all to happen. Though I still wish people would either make little apps for 3d that could allow a little control in just one thing.

Example: itunes has very nice visual feedback music generators. If there was just one little addition that could let a user control these with the mouse of sliders and it could render out frames as it apperas on screen. A simple clean app ( no clue how simple or hard it is to make) that has one purpose visuals.

Then another app could just do another one job system. This way very experimental techs could be dabbled with.

Either way I am left with answers, and nothing else. It is what I seek therefore I am happy somewhat. Thank you all.

^v^