A Newer video about http://www.hardocp.com/article/2011/08/10/euclideon_unlimited_detail_bruce_dell_interview.

40 or so minutes long with basically FULL Access to the company and what they are about, and how they explain That They Are Not A Hoax.

I don't think they are, I think they are just so far ahead of everybody else people can't get their mind wrapped around what they are doing.

I have an idea about how they are doing it based on what Bruce says, I could be wrong, and probably am, but he says it several times, about an atom tied to a pixel being a key part of what they are doing in how the render graphically.

And in doing it that way they can "Sample" the render in a much faster and detailed way.

Well to me as a Retired US Navy Sonar Technician, it sounds allot like FFT or http://mathworld.wolfram.com/FastFourierTransform.html, which in my previous field is used as a data sampling algorithm.

Which is essentially what Euclideon is doing, by converting the standard polygon in to "Atom" sized particles and then taking a sample of them based on pixel size to render the graphic, allowing them to have trillions of sample points.

Allowing the to render objects any object in Very Great Detail.

Which is what the FFT algorithm does for raw data sound coming into the processing equipment, it samples it and then send it to the screen to be graphically displayed in many ways.

I think they are adapting samlping method for what they are doing, I don't know but it sure looks familiar to me.. lol

Maybe some math wizards out there that know what I am talking about can watch the video and see if I am maybe onto what they are doing in some backward way applying what I know.

Fun stuff to talk about though, I think so anyway, what do you think?... :goodjob:

And what about the storage requirements for this? Their claims still scream "HOAX".

Did you actually watch the 40 minute video?

Did you understand what I said in my post?

Because I told you how it Might Be possible.

I'll believe it when Notch does.

Because Notch is great at efficient programming.

To be fair, he made a game good enough to make him richer than any of us will most likely be.

as an australian, this is fairly old news to me.

I dont think its a perfect system, but there is no reason it cant work. People complain that it cant be animated....or shadows wont work......how do we know? the tradional method for these things may not work, hence why it takes so long for them to make it, they need to rethink everything. It couldb e a method for the future.

http://uk.pc.gamespy.com/articles/118/1187338p1.html, which is also recent, is quite telling for the number of things that aren't finished yet.

My guess is that the technology is real, shows promise, but needs an absolute fortune spent on development before it's worth game developers using it. Remember also that it will not be an industry standard, so there'll be a lot of incentive for developers to ignore it in favour of DirectX or OpenGL, even if it's technically superior. It would need to be technically superior and have a fantastic set of development tools to be worth the risk of adopting. Developing tools like that, in competition with companies that have been in the field for many years and have mature toolsets with many millions of dollars investment behind them is going to be a very hard task.

A brand new company with investment of only a few million, with only one graphics artist? However great the potential of the technology, if they can't attract the investment to realise it, and do so fast, I suspect they don't have a chance.

Guys voice really annoys me.

I think it's something that's awesome and should be researched more but I don't think we'll see it anytime soon. In any case, I'm happy with my polygons for the time being...

It might be the way of the future or it could just be a waste of time. Somebody should attempt a small game with it and see how it goes.

To be fair, it wasn't even an original game idea. He simply took an existing game that had flopped and improved on it by making it so the world wasn't a limited size and added more objects and such. It isn't hard to make a game like Minecraft. However, he is obviously better at his lighting than I am since that is the only thing that was slow in my little test Minecraft engine. So slow I just replaced it with deferred shading :V

At 22' he's playing the demo in realtime on his laptop. Hard to believe that laptop can store terabytes of data.

Yup yup the engine seems to be working superb , also please check the 40 min interview at around at start of 27:10 when he starts talking about lvl of detail and draw distance to other engine compared to this one you will be very surprised its amazing!

When I first saw the demo when it came out a while ago, I was hugely skeptical, but hopeful. After this interview, I'm a believer. That laptop running at 15-25fps and they haven't touched the GPU power? Bring it on.

I do question how they're going to do humans. They can scan in objects, so I assume scanning in a human would work the same. But where would you find someone built like a Gears of War guy or any chick from a JRPG....

If he was. He wouldn't be using java..

this would be a nice way of programming for 360 and ps3, letting the games run at 1080p instead of 720p like moste of them do at present time.

I believe he said something about modeling objects in say, clay, then scanning those in.

It's more original than a lot of games I've seen lately :toughninja:

Where did you see anything in any of the videos that they have released that says anything about how much Hard drives space is needed to do anything?

If they said I sure missed it.

I think you are missing what is going on here.

The Polygons have been reduced to what the call "Atoms" as they have been converted using "Their" software.

Part of that process is sampling each one of those converted polygons (Atoms) to make the object render.

So that in itself is a quantum leap forward on what you need to store, because you are working with millions of smaller, "Sampled" objects, and Not objects that are rendered in polygon size and left sitting there.

That is why there is no scaling when moving toward and away from an object, because that object is "Sampled" as it is rendered, and rendered according to the distance from it in real time.

That is why in my post I provided a link to a way that can be done, using the FFT (Fast Fourier Transform) sampling algorithm, so that it does not have to store all of that.

Once it is rendered and in the world space it is sampled at pixle size as needed in real time.

Or at least that is the way that I have proposed in this thread on what is taking place and how they are doing it, it is all in the conversion of existing large size polygons into what they call "Atom" size.

And once done, Sampling those as needed to make the render.

Home many pixels are on an HD monitor?

http://www.fourmilab.ch/documents/howmanydots/, with a Wide (16×9 or 16×10) Screen Aspect Ratio at 7680×4800 it is 36,864,000.

Is there any storage there?

The answer is No.

It is No because those pixels are in fact sampling points for Data to be displayed, and have Nothing to do with where they are at, they just display what has been sampled/scanned on the monitor.

So a Screen Aspect Ratio at 7680×4800 with 36,864,000 pixels has Nothing to do with Storage, and that is what makes this technology a Quantum Leap forward in what it is going to eventually do.

It is all about rendering and sampling objects, In Real Time once they have been scanned into the world space and then displaying that in Real Time.

This for me as a Sonar Technician in the US Navy was done with Audio data using the FFT (Fast Fourier Transform) sampling algorithm with 4 Intel 8088/Motorola 68000 chips with late 1970's early 1980's technology.

So could today's powerful laptops run that same process, you bet they can, they can run circles around those old chip sets.

Now, that being said, did anything need to be stored for that process, No is again the answer.

Because then as I suspect it is now with what the are doing it is all about sampling data and displaying it.

Now I suspect they have taken that idea and turned it on it's head and are using the same kind of idea to sample, render and display converted atom sized polygons.

Do I Know that to be the case?

No I surely do not, what I have posted in this thread is how I think they are doing it and I provided a Way that I think they could be doing it with.

I could be Completely wrong, but what I have said fits, I think for sure.

And it really makes sense if you understand what FFT does, which is indeed an Very Old idea if you look at the link I provided on it.

When I was taught FFT in my Navy school, the instructor started that portion of the class training by saying, ok this is a part of the class that many of you will grasp and understand and many of you will not".

If you do not grasp what it is I am going to have to ask you to just push the I Believe Button, well I understood how it works, not the Math behind it but conceptually I understand it.

And because of that I understand what he is talking about with sampling.

I Think.

Because I don't know for sure, nobody does except the folks working there, I have just talked about how I think it can be done, and what I have said fits.

If he's a scammer, that'd almost be more impressive, really.

Nah i doubt he is a scammer, i mean the Australian government paid him and his team 2 milion dollars( i forgot how much was it) to keep developing this engine , i mean i am pretty sure its real seeing the 40 min interview and seeing the real time demo on that map.

I agree with you for sure.

And if you look back in history, there was just this same kind of talk, before Telegraphs, Airplanes, Telephones, Silent Movies, Talking Pictures, and Computers.

I know I am leaving something out but you get the idea.. lol

It's not a scam, it's real.
The catch however is that this technology is great at rendering STATIC environments, but in games we want environments to be as dynamic as possible.

With polygons, every object you render on the screen is also treated as an object in the game code.
The total amount of objects that are rendered on the screen at the same time is called the granularity of the environment.

In a simple 3D engine, a tree has a granularity of 1. In games like Crysis, the tree has a granularity of more than one, so it can break down into smaller pieces.
With "infinite detail" every pixel on the screen can represent 10.000 objects or more... in other words, the granularity is infinite.

If you have a tree make out of 1.000 polygons and you want to make it fall down, you have to do a maximum of 1.000 calculations for every frame.
A smart 3D engine only calculates the part of the tree that's visible for the player, so it can easily be cut in half.

With "infinite detail", the granularity is orders of magnitudes larger and you would have to do at least one calculation for every pixel on the screen.
So if it's 300 pixels wide and 800 pixels tall the engine has to do 240.000 calculation for every frame.

I don't see why not, Captain America was filmed this way.

Also they said in the interview that some high respected people in the game and engine developing industries that they hate them, i mean if this is a scam or some dumb engine why would they hate them, this engine is going to be the future and that is why those people hate this company cuz they will loose the engine they are selling( witch is crap compared to this anyway) oh yeah and i lost all respect i had for Notch after seeing that interview and see how much he trashed talk about them, i mean he made a game with very ugly graphic but with great and fun game play and all of the sudden he is an expert about computer graphic without even making a game with over the edge graphic? give me a break..