Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
April 24, 2014
arrowPress Releases
April 24, 2014
PR Newswire
View All





If you enjoy reading this site, you might also want to check out these UBM TechWeb sites:


Unlimited Detail Engine Creator Champions Voxels, Dispels Criticism
Unlimited Detail Engine Creator Champions Voxels, Dispels Criticism
November 23, 2011 | By Frank Cifaldi

November 23, 2011 | By Frank Cifaldi
Comments
    37 comments
More: Console/PC, Programming, Art



When Euclideon CEO Bruce Dell unveiled his voxel-based Unlimited Detail technology in a YouTube video earlier this year, projecting the death of polygons in favor of an unlimited amount of atom-like voxels, his claims were met with no small amount of criticism.

Id's John Carmack chimed in, saying the technology could "maybe" work "several years from now."

"[Carmack] was unaware at the time that we were running this purely in software," Dell recently told Game Informer Australia.

"In his particular case, he and Intel had tried to go down a similar road themselves. Intel had tried making its own system in order to do things along the field of unlimited graphics. It ended up closing that avenue as it figured it was something for the distant future when computers have more power."

Minecraft creator Markus "Notch" Persson went even further in his criticism, with a detailed blog post calling the engine a "scam" and Dell a "snake oil salesman," claiming the video was impossible by way of past attempts at voxel engines.

"The examples he suggested were so incorrect you'd have to be pretty disconnected from 3D graphics to make the mistakes he made," explained Dell.

"The mistakes Notch made were so bad that, if we were less kind, we would be able to really discredit his actual understanding of a lot of stuff in general. But that's honestly not our intention: we don't want to make enemies with him. We prefer to bring him around and be nice to him."

Developers will soon be able to make their own calls when an SDK of Euclideon's tech is released. While Dell was unable to say when that might be, he hopes that it will be "a lot less than a year."


Related Jobs

2K
2K — Novato, California, United States
[04.24.14]

Senior Tools Programmer
Turbine Inc.
Turbine Inc. — Needham, Massachusetts, United States
[04.23.14]

Director, Analytics Platform Development
Gameloft Toronto
Gameloft Toronto — Toronto, Ontario, Canada
[04.23.14]

Technical Artist
Muti Labs
Muti Labs — Santa Monica, California, United States
[04.23.14]

Senior Game Engineer










Comments


David Roberts
profile image
Why are we giving this conman a platform?



If he wanted his technology to be seen as anything more than a scam then why not release a techdemo? (Of which obviously exists as per the videos). His entire outfit seems to be a vapourware scam out to extract money out of investors



"a lot less than a year." is also the same time span (used again and again) since this con first appeared, and it has never materialised.



If this technology ever does show up in any real form, they need to sack this guy ASAP. He is a walking PR disaster. On top of that they need to stop calling it 'unlimited detail' if they want an ounce of credibility. Anyone in computing who hears unlimited in any context is going to know this is rubbish

Andrew Nissen
profile image
"The mistakes Notch made were so bad that, if we were less kind, we would be able to really discredit his actual understanding of a lot of stuff in general. But that's honestly not our intention: we don't want to make enemies with him. We prefer to bring him around and be nice to him."



I'm sorry, but instead of being a real man and calling someone out on their supposed bullshit (I still support what notch has said about it all), he's trying to insult him but is too afraid to actually come out and say it. You can't have it both ways buddy, and by trying to I just lose even more respect (as if I had any in the first place) for the guy. Choose a side and stick to it. If the tech really existed and he was 100000% sure it he was correct (as he certainly comes off as when he talks), then he should have no problem disproving Notch word for word. In fact, Notch (and everyone else) would probably believe him and have more faith then him instead of being passive aggressive and not actually saying anything.

Tyson Sewell
profile image
Actually. Mr Dell is trying to be polite to someone who has achieved much despite being personally insulted by slanderous claims. There is absolutely nothing wrong with that. Mr Dell is simply trying to remain civil, something I wish I saw more of on the internet.

Tom Baird
profile image
"The mistakes Notch made were so bad that, if we were less kind, we would be able to really discredit his actual understanding of a lot of stuff in general."



Explaining new concepts is not making enemies, it's called explaining yourself. By saying 'Oh he's so very wrong and doesn't understand 3D graphics, but we won't explain why' rather than saying 'It's not an issue for us because of...' makes you look like, as used above, a snake oil salesman. Current Voxel implementations often fight against very high memory usage. They may have have found good solutions to that issue, but if your going to try to explain yourself, be constructive and use instruction and evidence, rather than relying on trust and 'oh he just doesn't understand'.

Darcy Nelson
profile image
"The mistakes Notch made were so bad that, if we were less kind, we would be able to really discredit his actual understanding of a lot of stuff in general. But that's honestly not our intention: we don't want to make enemies with him. We prefer to bring him around and be nice to him."



Sheez, seriously? Providing examples of how his flawed reasoning wouldn't be discrediting to Notch so much as it would grant legitimacy to tech that's being looked at with a lot of skepticism. That's a missed opportunity right there.



If anything, I'd like to see something akin to Unity's Bootcamp demo before looking at this again. So far the Euclideon vids that I've seen have been pretty, repetitive graphics with no lighting and nothing to suggest it would be a great fit for games instead of say, film.

Mike Griffin
profile image
Time for Bruce to take the stage with a verifiable and realistically equipped live demo of the engine in a gaming-like implementation. Real-time interactive demo on a mortal PC configuration, or I call BS.

Johnathon Swift
profile image
Does anyone care about this guy... nope!

Harry Fields
profile image
What a crock. Anyone gullible enough to fall for this.... please contact me. I have a private tropical planet populated with beautiful Amazonian ladies for sale... cheap.

Yama Habib
profile image
I, for one, am praying that this proves fruitful.



I've been waiting for the realm of 3D gaming to migrate away from polygons for quite some time now. It's just not how the world works. As far as the world of video gaming is concerned, we haven't seen anything like this since compact discs became popular and paved the way for 3D polygon-based engines.



I feel like it's usually some of the most knowledgeable that end up most eager to discredit groundbreaking innovation. I'm trying not to fall into the "over-skeptic" category. Even if this doesn't turn out to be quite as great as it claims to be, I've no doubt that we'll see practical atom-based engines in the future. Judging from how quickly hardware continues to evolve, it might be sooner than most would think...

Darcy Nelson
profile image
We're all hopeful, and that's what makes us so miffed when he makes claims like that and can't seem to back them up. It's the great dangling carrot of graphics!

Harry Fields
profile image
Hardware, in any conceivable consumer format is just not powerful enough for this kind of engine yet. It's not even a matter of "almost there". Perhaps in a decade, it might be feasible... but then, if display technology goes 4k by then, all bets are off. True advances in software are always incremental and this conman is promising things that anyone with half a brain, unless incredibly naive, knows he cannot deliver, much less on mobile devices and the like as he claims possible.

Corbo UK
profile image
The compact disk didn't pave the way for 3D polygon based engines, they existed long before the first single speed CD drive, as far back as the 8bit era.



I'd like this technology to come to fruition and overcome the lighting, animation and other challenges they are facing. I can't help but think that if they went to a multinational such as Electronic Arts and proved the technology, they could make a small fortune and decrease development time ten fold with the resources they need. The secretism serves only to raise suspicion.

Marc Schaerer
profile image
Really, why is this guy crawling out of his cave without having anything to proof his claims?

So far he did pretty well in one thing, the thing he totally succeeded in: showing that even 12 months of time didn't make the tech progress far enough to make it production usable before 2020 at this pace ...



as long as he does not even grasp this fact, I've little hope that he will grasp on why he gets shot down like this.

Paul Szczepanek
profile image
From the interview:

<< Do you see this being compatible with existing middleware, such as Havok?

"I'm not allowed to comment on if we're already talking with people in that particular area [smiles]. But yes, it is our aim to join up with existing middleware." >>



I think this is representative of all their statements. Suggestions, unfounded claims, and hiding behind super secret sauce ingredients they can't share but will revolutionise the industry. In "a lot less than a year."



The only value they are bringing right now is entertainment. I'm not paying for it so I'm OK with this.

Nick Kinsman
profile image
I think the take-away for the time being is "Wait a year." We'll see then, no?

Paul Shirley
profile image
This remains a story where buzzwords outnumber explanation. Still don't know *what this does better* than other engines, better enough that it couldn't be added to them in a few months. Better enough that I'd care about using it.



'Infinite detail' is clearly not infinite and would appear not implemented in other engines simply because no-one really wants|needs it. Rendering massive scale ranges was solved before the 8bit days.



If they have a tool for creating or encoding extraordinary detail levels that would be newsworthy. Another low performance rendering engine with visibly poor quality is not news.

Luis Guimaraes
profile image
Stop using the words "atom" and "voxel" then. Everybody starts to think of voxel physics, voxel dynamic interaction, voxel fluids, non-empty mass volumes, and starts the impossibility claim without knowing what it's really about. Unless you're intentionally baiting for the skeptical uninformed chorus as a marketing strategy...



People are way too easy to take anything as impossible without even trying to think how to make it possible...

Pablo Puente
profile image
Where does all this hate come from?

Let me quote the article:

"It's hard then not to be enthusiastic about what Unlimited Detail could mean for the fidelity of graphics. I must admit that I was very impressed with what I was shown, and even though I poked and pried as much as I did onscreen, I couldn't find any telling cracks to suggest a facade."



There's a difference between being skeptical about something and calling someone a scammer outright. I'm aware notch has a lot of fans around here, but his "analysis" was simply flawed and he deserves to be criticized in words no less harsh than those that he picked himself.

Tyson Sewell
profile image
I agree. Notch's attacks on Dell were practically personally. Rather pointing out issues with with what Bruce proposed, he rather overstated his skepticism. One of his claims has been withdrawn I believe, but the others still stand last I checked. No matter, time will reveal all one way or the other.

Maria Jayne
profile image
Is it possible he can't explain how it's being done because it might reveal too much of the secret, before it's marketable?



Even if it isn't possible right now, everybody seems to think it will be possible, being the one to patent it and bring it to market will be a huge return if they manage it.

Paul Szczepanek
profile image
Is it possible he can't explain how it's being done because it might reveal it's not actually feasible?



How far can patents go? I fully expect some company to patent free speech at some point.

Dave Kay
profile image
On the (very big) assumption that this is all legitimate... you're absolutely right. They might have to hold back on a lot of the detail to maintain 'trade secret', if you will.



They still have to talk about it, though, because they have to try to generate investment money to continue development.



It becomes a fine line to walk... generating enough hype about your product to hook investors, but keeping enough secrets that nobody else can scoop your technology or product from under you.



However, 'snake oil salesmen' know this formula well, and it's exactly what they claim they're doing when they're really just spending all your investment money.



So, in the end, only time will tell.

Denis Nickoleff
profile image
I don't see why this is such an emotional issue for everyone, if it works and it's proven, awesome. If it doesn't and it's not, then no one in their right mind would look much into it. I've got more reading to do on this but from what I can see so far; Bruce Dell is too cryptic and needs to show a tech demo to some people in person, and not just YouTube, and Notch needs to take some Xanax.



So far the only level headed person seems to be Carmack with his ' I think it's possible, but we're a ways off' assessment.

Tyson Sewell
profile image
I think it's because some people are worried Mr. Dell is ripping the government off for millions of dollars, which would certainly warrant outrage if it were true. So people get vicious in their comments and crits of Dell, his company, and his software. And If you have one side responding in a personal manner, usually the other side will too.

Kostas Yiatilis
profile image
I guess you guys haven't seen the video he did with a reporter controlling the software. The demo might not represent the tech 100% and that is why they want to address stuff before they release one.



Half the comments in here use Notch's comments (even the wording) to somehow disproove this guy.



First off he didn't ask for your money, so what's your problem.

Second he didn't ask for any money, his project got a goverment grant before he went public with his demos, they also made very clear that they are not looking for more funding.



So instead of getting into a tangle let the guy do his thing. It's not like you have to do anything, he posted a video, you believe it great, you don't great. Can you do anything about it? NO



I actually think it's his accent that makes everybody so huffy-puffy, hehe.



Anyway wait (no effort required) and see. They are doing all the work, all you guys are doing is being drama queens for no good reason.

Joe Wreschnig
profile image
"First off he didn't ask for your money, so what's your problem."



If you live Australia, he did ask for your money, and he got it.

Kostas Yiatilis
profile image
Here: http://www.youtube.com/watch?v=JVB1ayT6Fdc

Michael Joseph
profile image
and after that, watch this...



http://www.youtube.com/watch?v=P_6vDLq64gE



lol. i couldn't resist.



I actually watched a good bit of that video and he talks about everything except the stuff that people want to know. He also makes some statements which makes me question whether he has anything more than a rudimentary understanding of 3d graphics programming. Red flags galore.



EDIT:

"In 2010 Euclideon was the recipient of approximately $2 Million, the largest grant awarded by the Australian Federal Government under its new Commercialisation Australia initiative. The funds provided by the grant will support the implementation of multi-platform functionality to Unlimited Detail. Thus allowing the technology to operate on a variety of hardware, including mobile phones and game consoles"



So good for him, he's gone from working collecting dead bodies at mortuary (he says so himself in the video) to getting a nice pay-day.



EDIT2: The more vids you watch of this guy, the more convinced you will be that he's a scammer.

Arnaud Clermonté
profile image
I can see why people would be suspicious of this guy.

With his "Game Boy DS" and his "Level of Distance", and his constant joking and laughing at his own jokes...

And this video of a demo of a completely static and mega-repetitive landscape with no collisions...



Compare that to what Notch delivered, and it's clear which one of these 2 guys is more trustworthy.

Kostas Yiatilis
profile image
Plus why post a demo when someone can reverse engineer the tech and beat them to it by investing even more money? I wouldn't.

Matthew Duhamel
profile image
Because they may be looking for someone to invest.



Or they could just be stupid.



Or less paranoid than you.



Or all three.

Mark Bamford
profile image
I think it could work and here's why: http://volumetricworlds.blogspot.com/2011/11/my-thoughts-on-unlim
ited-detail.html

Michael Joseph
profile image
it's the intelligent design algorithm. the divine equation realized on the PC. brillant! This is nobel prize material people! Forget gaming, think of all the medical applications. Governments around the world are going to be bidding for this technology.

Kenneth Blaney
profile image
It makes me happy to see so many here are skeptical of Euclideon (incidentally, if I was going to make a joke company dealing with geometry, that is what I would call it).



All in all the major criticism that Notch brings up is valid no matter how much or little experience you have with voxels or polygons. That is, simply storing the information associated with a map the size they claim would take Petabytes of HDD space. The claims are quite obviously overstated at best.



However, I can conceive of a way to provide unlimited detail mathematically speaking without the petabytes of data and make it realistic on current computers: fractals. Simply put, with a 3D fractal, one would be able to create a world of infinite depth and refined detail at that depth using a single equation. Create one fractal for each texture/shader (one for bark, one for leaves, one for statues, etc.) and you can create quite a lot of detail pretty quick. Then tie what textures are shown at what zoom level and you can get the same equation that used to define rocks to define gains of dirt. Additionally, this would give a certain sense of procedural generation of a landscape. In fact, this is what I think Euclideon may have done. Why else would there be highly detailed statues in places no one is likely to see? (Because no one consciously put them there.)



The major disadvantage to using fractals in this way is that it becomes costly to change anything and impossible to change things independently. That is, to make a tree sway in the breeze or a boot print in the dirt, the whole equation governing the landscape would have to change. In my research, I do not believe the behaviors of fractal equations are well understood enough to generate an equation for something like an arbitrary boot print on the fly. That's a problem with out understanding of the math, not the technological level of our computers.



So, if they are indeed using fractals, and they are doing it in a really clever way, then maybe they'll be able to licence this technology to someone like Epic who can turn it into a highly detailed supplement to BSP geometry. However, the claims that this will be a fully interactive game environment where everything down to the grains of sand at the beach can be fully custom created and placed are far out of the ball park.

Chris Remo
profile image
I think they've actually been showing this stuff for years now--much earlier than 2011. I remember first seeing one of their videos ages ago. Wikipedia suggests they first unveiled it in 2003. And in all that time, they've still failed to provide an SDK (which perpetually seems to be coming in less than a year) or any more concrete technical papers.

Jeremy Alessi
profile image
I think the big issue for Notch is that he's assuming that this engine features deformable volumes as we typically see like Minecraft. The video makes no mention of real time deformation which means extrapolation could be used to expand patterns of volume without the memory hit Notch describes. Their goal is to improve on polygon rendering detail not to simulate the physical properties of real volumes (like Minecraft). Thus, the goal is to use the occlusion features of a voxel engine only and not the deformation features. As the view is zoomed in the number of voxels probably remains constant and they simply change the density of voxels per unit of space. The assumption is that for any vantage point there are ever only so many visible patterns which can be broken down into loops of patterns that can be described with algorithms instead of explicit memory. So the content creation process doesn't have to be procedural but the engine converts the non-procedurally generated content into procedurally rendered content. Usually the two are mutually exclusive but there's no reason that has to remain true just like there's no reason that voxels have to represent deformable volumes.



Anything's possible with a little imagination no? Who knows ;)

Samuel Batista
profile image
Finally, someone that isn't so blinded by the whole "Noch" controversy and give the method some credibility. I for one don't think it's bullshit, I've seen some impressive things done with Voxels in great quantities. Everyone keeps saying it's the holy grail, but I highly doubt it. I also highly doubt its application in modern gaming, at least in the near future, since there is a whole slew of things that are necessary for modern gaming that wasn't at all demonstrated (lighting, movement, shadowing, etc).



We'll see what this will turn into, but really, you can't blame the dude for mentioning Noch and trying to discredit his claims. After all, Noch called him a Snake Oil Salesman, and Noch isn't a tech expert, he's done some Voxel stuff sure, but I can see how Noch might be too rooted in his "good old ways" of voxel rendering to be able to appreciate this new "method".


none
 
Comment: