Euclideon Infinite Detail

[ali31]

This Australian company claims to have achieved "unlimited detail" by switching from triangles to points.

Video Here

What do you think of the technology and if it's legitimate would Epic be interested in incorporating it into UDK?

[Blade[UG]]

This gets hashed out on virtually every game enthusiast forum at least once every 2 years or so for the past 10 or more. Until someone actually shows something off, it's pretty much snake oil.

[ADF86]

Seems you can't animate with that tech; might be better just to stick with tesselation, it'll achieve similar results and actually works for game purposes. Otherwise you've simply got rocks with as many verts as an average games' entire level.

[Graylord]

It's old news rehashed and renamed. Most people believe it to be a scam, as we never see any proof of it being realtime, they don't explain anything at all in their videos except repeating themselves for 10 minutes and there's no animations.

They've been gone for a year and all they did was put together a map. If this was truly useful technology, it would have been picked up by the big guys ages ago.

[danimal']

They're called voxels, they're old news, move along, move along...

[Crusha K. Rool]

Wow, this suddenly pops out on all gaming forums simultaneously again as if it would be such a big deal.

All I see is that the commentator is still as swanky as he used to be in the first video and that he is still harassing on other games, just that they added Bulletstorm into the new video. And if I wanted to get some funding for such a project, I wouldn't slander about potential partners like AMD and nVidia in my first appearance.

[Outstandish]

I have said this on OCN, it is redundant now. If you have seen the Samaritan video, you see that in terms of geometry the objects are very realistic, and those are done with well POLYGONS! THis causes way to much load I can predict, and it is a little to late to actually make a damn difference when Polygons are starting to look as good as they do now.....

[darthviper107]

And the issue today isn't the polygon count--it's lighting. Notice in that video, even though they were using "billions of polygons" it didn't actually look better than today's video games because the lighting was crap.

[Outstandish]

And the issue today isn't the polygon count--it's lighting. Notice in that video, even though they were using "billions of polygons" it didn't actually look better than today's video games because the lighting was crap.

Quite, lighting makes a game for me, besides texture sizes of course. Polygons are not much of a problem anymore....

[ambershee]

There are decent voxel engines out there, this isn't one of them. It's not lit properly, and it's not in any way interactive. The most promising one I've seen so far was by a small Polish developer - but they still used regular polygonal geometry for everything except the static scenery.

[DifferenceEngine3D]

In the end, they're still dependent on polygons; the modellers all work on them, and as far as I can see, they use a fancy algorithm to simply extract the surface layer of voxels to produce a pixel of colour. Nothing new here, TBH; still needs DX9/10/11/OGL 3/4/whatever, since that's what the graphics cards they display it on works with. Something like this will work out eventually, but it'll take some time.
@ambershee: I'm guessing atomontage?

[DifferenceEngine3D]

BTW: my guess about color extraction is based on the fact that no hardware in the world can run trillions of voxels in memory at once. I worked on medical visualizations for Royal Holloway and you have to find ways to extract point-cloud data efficiently without holding it all in memory.

[ambershee]

@ambershee: I'm guessing atomontage?

Amongst others. I'm fond of Atomontage because it's hybrid (voxels for terrain / static environment), and because at least they're actually honest about the issues they're running into. Instead of selling the same crap every 6 months trying to get more money for their snake oil.

I don't know why this is cropping up all of a sudden, but I've seen four threads on it today in various forums.

It's the exact same **** the same people have been showing for years now, with no noticeable improvement.

[DifferenceEngine3D]

I was taken with Atomontage, too. I mean, no-one else out there had the guts to say the procedural generation took hours for a small level (on a low-powered laptop, sure, but still). They're honest about their timeline, and as open as they can be about the technology without revealing too much detail. Plus their presentations are nicely poetic, thanks to the translation from Slovakian!

Voxel tech has been used in games before, of course (remember "Comanche"?) but it's still relatively inefficient. It has engineering, Earth Science and Medical Imaging appliations, because you need to move large amounts of 3D data with actual 3D properties in those arenas. Game tech simply doesn't need voxels; the CG companies out there use polys (last I checked, even Pixar still used them! < snark > ) and all we need is better applications of existing tech to make it look real. UE3 and other engines can do that with good enough hardware and good enough artists.

Ah, it'll recede to background noise before you know it. Only to start up again in 2012.

[Crusha K. Rool]

Even Notch warns about it:

It’s a scam!

Perhaps you’ve seen the videos about some groundbreaking “unlimited detail” rendering technology? If not, check it out here, then get back to this post: http://www.youtube.com/watch?v=00gAbgBu8R4

Well, it is a scam.

They made a voxel renderer, probably based on sparse voxel octrees. That’s cool and all, but.. To quote the video, the island in the video is one km^2. Let’s assume a modest island height of just eight meters, and we end up with 0.008 km^3. At 64 atoms per cubic millimeter (four per millimeter), that is a total of 512 000 000 000 000 000 atoms. If each voxel is made up of one byte of data, that is a total of 512 petabytes of information, or about 170 000 three-terrabyte harddrives full of information. In reality, you will need way more than just one byte of data per voxel to do colors and lighting, and the island is probably way taller than just eight meters, so that estimate is very optimistic.

So obviously, it’s not made up of that many unique voxels.

In the video, you can make up loads of repeated structured, all roughly the same size. Sparse voxel octrees work great for this, as you don’t need to have unique data in each leaf node, but can reference the same data repeatedly (at fixed intervals) with great speed and memory efficiency. This explains how they can have that much data, but it also shows one of the biggest weaknesses of their engine.

Another weakness is that voxels are horrible for doing animation, because there is no current fast algorithms for deforming a voxel cloud based on a skeletal mesh, and if you do keyframe animation, you end up with a LOT of data. It’s possible to rotate, scale and translate individual chunks of voxel data to do simple animation (imagine one chunk for the upper arm, one for the lower, one for the torso, and so on), but it’s not going to look as nice as polygon based animated characters do.

It’s a very pretty and very impressive piece of technology, but they’re carefully avoiding to mention any of the drawbacks, and they’re pretending like what they’re doing is something new and impressive. In reality, it’s been done several times before.

There’s the very impressive looking Atomontage Engine: http://www.youtube.com/watch?v=Gshc8GMTa1Y

Ken Silverman (the guy who wrote the Build engine, used in Duke Nukem 3D) has been working on a voxel engine called Voxlap, which is the basis for Voxelstein 3d: http://www.youtube.com/watch?v=oB1eMC9Jdsw

And there’s more: http://www.youtube.com/watch?v=xUe4ofdz5oI http://www.youtube.com/watch?v=lEHIUC4LNFE http://www.youtube.com/watch?v=Zl9CiGJiZuc

They’re hyping this as something new and revolutionary because they want funding. It’s a scam. Don’t get excited.

Or, more correctly, get excited about voxels, but not about the snake oil salesmen.

[frostymoss]

I'd put my money on lighting tech (e.g. Dice & BF3). You ever notice how laptops and tvs these days seem to produce much richer bass tones - it's just smarter algorithms and cooler harmonious mathematics. Turns out your brain can take on its share of the rendering...

[onethought]

The main guy involved in this has been trying to get someone to pick this up for ages (2003 is when I think he first started). I remember talking to one of my coder friends (who used to be a lead at one of the largest studios in Australia) about 2 years ago and him having a rant about this guy and how he tried to convince them to pick it up.

What amazes me is that he managed to get himself not quite $2 million in funding from the Australian government to pursue this. I mean, How do you do physics with voxels and avoid a huge overhead? How do you trivially animate something? Dynamic shadows? Decent load times for Voxel data?

I really hope that the huge amount of funding he got will help develop the technology into something feasible.

[darthviper107]

The guy kind of sounds like a tool in the video anyways

[ambershee]

I really hope that the huge amount of funding he got will help develop the technology into something feasible.

It did. Years before he started. Voxels have been used in engineering applications for a long time. Voxel based renderers have also been used in games since the 90s (Delta Force, anyone?).

[onethought]

It did. Years before he started. Voxels have been used in engineering applications for a long time. Voxel based renderers have also been used in games since the 90s (Delta Force, anyone?).

I was aware they had been in use but not to that extent. In that case, this may be quite a waste of money by the government. Disappointing really.

[BlackHornet80]

Don't forget one of the last Volex based games: Outcast

[DifferenceEngine3D]

One of the biggest challenges is animating voxels, as has been pointed out before; For the medical imaging applications I consulted on, the voxel data was stored in a really insanely huge network which was then passed through (as has also been pointed out) sparse octrees to provide a surface-only representation of the view for that quadrant. The data came from MRIs, so there was LOT of it (a small scan creates several gigabytes of raw voxel data at a much larger level of granularity than this guy on the video claims to have).

The renderer was to be used for a keyhole surgery simulator; Essentially all the voxel data was done away with in the renderer and was used only as a referent for when a cut was made to determine what lay underneath the next layer, which was then dutifully reconstructed... using polygonal data! Voxel rendering, even of something a "limited" in scale and scope as, say, a shoulder joint or a heart valve, did not allow for the same flexibility as polygonal data, even then...

[frostymoss]

One of the biggest challenges is animating voxels, as has been pointed out before; For the medical imaging applications I consulted on, the voxel data was stored in a really insanely huge network which was then passed through (as has also been pointed out) sparse octrees to provide a surface-only representation of the view for that quadrant. The data came from MRIs, so there was LOT of it (a small scan creates several gigabytes of raw voxel data at a much larger level of granularity than this guy on the video claims to have).

The renderer was to be used for a keyhole surgery simulator; Essentially all the voxel data was done away with in the renderer and was used only as a referent for when a cut was made to determine what lay underneath the next layer, which was then dutifully reconstructed... using polygonal data! Voxel rendering, even of something a "limited" in scale and scope as, say, a shoulder joint or a heart valve, did not allow for the same flexibility as polygonal data, even then...

MRI software consulting/programming -- that's the business I want to get into some day

[flota113]

I'm really very curious how authors of unlimited details would anserw ;-D

[darthviper107]

Here's his response:
http://kotaku.com/5827192/euclideon-...-is-not-a-hoax

still extremely vague. I highly doubt his claims still. He doesn't actually counter anything other than saying "they're wrong!" and "We're working on it"

[MonsOlympus]

I dont think funding from the Australian government is any kind of guarantee, our last ummz, lets say 6 for arguements sake, governments have all made promises they broke. They also gave money to Auran, a well developed studio with some actual game design experience and they ended up going under.

The simple fact of the matter is the renderer is moot without any game to play on it and until they recognize that they wont get anywhere. Gaming is about alot more then pretty graphics (despite what end users would have you believe), the clips they show dont have any gameplay what so ever.

Let me put it to you this way, how many times have you stopped in the middle of a battle to admire a grain of sand or a rock? It doesnt really matter a whole lot, does it? Now another thing is why we can see polygons so well, one reason is we have anti-aliasing which gives us nice clean edges, atoms on the other hand give us blocky looking edges which would only look worse with AA turned up. Ontop of this is the clip doesnt show any DX11 demonstrations what so ever, so a high polygon model as converted from voxels (such as the elephant) isnt actually as many polygons as he says, its just a poor example because he doesnt understand polygon methods.

To finish up Id like to mention about particles, something that is used volumetrically in prerender quite alot, the results of which are baked out on textures for use on larger particle systems in games. Could you imagine if there was alittle bit of wind in their test map? Not only would there be dust kicked up, volumetric dust at that, the tree's would sway and as people have mentioned its difficult to animate voxels/atoms. Its just a huge leap from the demonstration of this infinite detail technology to an actual playable product. I for one dont know many middleware providers who dont use their own technology so until they have something more than a renderer and a polygon conversion toolset you can count me out.

Theres plenty of other people who could move to this technology and use it themselves without the need to license dreams.

Thanks

[Crusha K. Rool]

His library example is just what octrees and binary trees are all about, so Notch isn't so far away with his guess.

[DifferenceEngine3D]

Oh, as long as he swears it's not a hoax...

His "layman's" explanation confirms sparse octree data, with a reconstruction algorithm based on closest-point data. Essentially, the renderer "guesses" what to display based on what's nearby. Nothing revolutionary there. His claim of "trillions" of atoms is therefore utterly bogus. The algorithm may have the ability to arbitrarily reconstruct the equivalent of trillions, but in reality, there are probably only a few gigabits of sparse data available, with lookup tables supplementing the renderer to create the "grains of sand" he claims were there all along.

Nothing to see here... Move along...

P.S. That he could hoodwink the government into giving him a tech grant is proof of precisely nothing. I could show you a thousand grants given in the U.S. alone on the basis of suspect, undeliverable technology...

[Crusha K. Rool]

And Notch goes more into the details of his disbelief.

“But Notch, it’s NOT a scam!”

I’ve been getting a bunch of feedback that my last blog post is wrong for various reasons, and I’d just like to say that I would absolutely LOVE to be proven wrong. Being wrong is awesome, that’s how you learn.

If you want to read my reasoning behind various assumptions, click “read more”.

Why I assume it’s voxels and not point clouds:

* Voxels store only the information about each point, and their positions are implicit in the location of where the voxel is stored. Point cloud data stores both the information about each point and the position of each point.
* They mention “64 atoms per cubic millimeter”, which is 4*4*4 points per mm^2. While it’s possible they only refer to the sampling frequency for turning polygonal structures into point data, the numbers are just too round for me to ignore as a programmer.
* All repeated structures in the world are all facing the same direction. To me, that means they aren’t able to easily rotate them arbitrarily.

About the size calculation:

* I was trying to show that there was no way there was that much UNIQUE data in the world, and that everything had to be made up of repeated chunks.
* One byte per voxel is way lower than the raw data you’d need. In reality, you’d probably want to track at least 24 bits of color and eight bits of normal vector data per voxel. That’s four times as much data. It’s quite possible you’d want to track even more data.
* If the data compresses down to 1%, it would still be 1 700 three-terrabyte hard drives of data at one byte of raw data per voxel.

Animated voxels:

* Holy crap, people sent me videos of this actually being done!
* I was wrong!
* http://www.youtube.com/watch?v=tkn6ubbp1SE
* (But please note that just that single animated character runs at 36 fps)

Why it’s a scam:

* They pretend like they’re doing something new and unique, but in reality a lot of people are researching this. There are a lot of known draw-backs to doing this.
* They refuse to address the known flaws. They don’t show non-repeated architecture, they don’t show animation, they don’t show rotated geometry, and they don’t show dynamic lighting.
* They invent new terminology and use superlatives and plenty of unverifiable claims.
* They say it’s a “search algorithm”. That’s just semantics to confuse the issue. Sparse voxel octrees is a search algorithm to do very fast ray casting in a voxel space.
* They seem to be doing some very impressive voxel rendering stuff, which could absolutely be used to make very interesting games, but it’s not as great as they claim it is. The only reason I can see for them misrepresenting it this bad is that I assume they’re looking for funding and/or to get bought up.

If these guys were being honest with the drawbacks and weaknesses of their system, I’d be their biggest fan. As it is now, it’s almost like they’re trying NOT to be trustworthy.

All this said, voxels are amazing. So is raytracing and raycasting. As computers get more powerful, and storage gets faster and cheaper, we will see amazing things happen.

And a final word to the engineers who worked on this: Great job, I am impressed! But please tell your marketing department to stop lying.

But yeah, maybe I should stop now to bump this.

[DifferenceEngine3D]

That's a nice video in Notch's post, but that's a lot of data for an only-okay walk cycle that any polygon-based game engine could achieve at power-of-ten-better frame rates... Which, of course, proves Notch's broader point about the current impracticality of pure-voxel solutions.

And, yes, I hate myself for bumping...

[MonsOlympus]

Time based renderers ftw, gah bump <slitwrists/>

[chester1997]

people its not fake they have proof now http://www.youtube.com/watch?v=JVB1ayT6Fdc

[Crusha K. Rool]

They still don't proof any of the points mentioned here.
How do you know that the "interviewer" is not a friend of their company? He seems like quite brown noser of the CEO to me and does not ask any of the detailed questions that the critiques have.
The only critical thing they spoke about was "Can you do animation?".
The answer was "Yes, we can." Instead of a proof for that claim does he only show animation from another seven years old voxel engine.

Besides that the guy is constantly laughing at other technologies during the entire interview and behaves in general quite childish. How are you even remotely supposed to take such a guy serious?
He also says that LOD would mean Level of Distance.

Could we now forget about this stuff please? Move along, nothing to see here.

[ambershee]

They have an interview where their CEO throws a few marketing terms around and childishly slanders existing proven products witht he simple 'I'm better than you and don't have to prove it' style rhetoric. Nothing new here.

[Crusha K. Rool]

Another nice post that wraps up all the problems:
http://www.overclock.net/video-games...ne-viable.html

[DifferenceEngine3D]

Well, I'm sold. After all, they said they can do animation, and they said they can overcome all the problems that every other voxel-based renderer has struggled with, so what's not to believe?

Just because they have no working examples of anything they said they can do is meaningless. After all, remember that cold-fusion thing a while back? They didn't have to prove a thing, and now look! Every power-station in the world is running on cold-fusion, and my cold-fusion car gets a million miles to a gallon of water! I say bring on the unsubstantiated gossip and bold claims of revolution without proof! It's always a paradigm-changing event every single time!

And as for you naysayers: have a little faith; if I had a dollar for every time a bold claim didn't pan out, I'd have, well, lots of money. But that proves nothing. After all, a 41-minute infomercial that shows nothing more than what they've already shown is proof enough.

In the words of one old-time Youtube sensation:

"Leave Euclideon AAALLLOOOONNNNE!!!! Waaaahh!!!"

< unnecessary snark tag >

[onethought]

The interview has removed a lot of my doubts, but I would still like to see a bit more in depth. I guess we will have to wait.

FYI, here's the interviewers LinkedIn profile. I think there must have been some form of a connection to start with, hence why he managed to get an interview.

[ambershee]

From what I can gather he is presently employed to promote their product.

[DifferenceEngine3D]

Eh, let those who want to believe, believe. To be perfectly honest (and this is an artistic preference, not a value-judgement) the atomontage engine is visually more impressive and, as has already been stated, at least more honest about the issues. Until it actually arrives and can be dissected, there really is nothing to see. Most new technologies (or new applications of old ones) don't need anything like as much fanfare and trumpet-blowing, which is what gives those of us who follow these things that itch on the backs of our necks...

[MonsOlympus]

What is memory compaction?