New Unlimited Detail Videos

You've probably read about this and (Minecraft) Notch saying that they are lying about the way the technology works. I'm not qualified to say what's wrong or right, but it does sound too good to be true.

Anyway, I love the first video and how detailed everything looks and I thought I'd share it with people who haven't seen it yet.

The second video is pretty entertaining, the creator of that technology (or not, who knows) talks about Notch and confutes all of his accusations (at least it seem that way, but for some reason most of it doesn't sound very trustworthy to me.. I'd love it, if all of what they claim it to be turned out to be true though).

Several jump cuts in the 2nd video...

I'm not nearly as cynical as the average internet user, so I'm being cautiously optimistic on this project.

Oh god that voice again.. Makes me want to punch that guy in the face. Not expecting to see a big voxel engine (or whatever it's called) for some years.

Also the Carmack vs Notch statement is pretty bad. They obviously mean different things.

And next year, they're going to announce a perpetual motion engine.

The "interview" itself wasn't particularly good; but overall, it was actually pretty informative. I look forward to seeing a 2560 x 2048 pixel png screenshot.

@BeachThunder: Yeah, well.. the question is if what they are saying is true. They all sound like actors to me, kinda strange.

I heard about this years ago while i was connected to a project at EA. Suffice to say the technical guys there considered it a source of good jokes.

my verdict: smoke and mirrors.

@Jadeskye: Yep, I guess when games start looking that detailed we'll be really really old.. sadly. :)

Never trust a red head. They have no souls

Never trust a red head. They have no souls

It sure would be nice if they showed something animate in that engine. Is it too much to ask to get the trees to at least sway a bit?

Show me some character animation.

It sure would be nice if they showed something animate in that engine. Is it too much to ask to get the trees to at least sway a bit?

I don't believe them, that's all I'm going to say.

I know I'm bumping an old thread, but if this wasn't true and they were just asking for money then why are they:

• Publicising the head of the company, who will be incapable of getting any credible jobs if this is fake.
• Openly admitting that what they haven't shown isn't finished yet.
• Getting millions of dollars from the government.
• Declining all types of investments
• Wasting years of peoples life's, and starting a company, for apparently nothing.

It definitely does sound too good to be true. That's the point.

They've actually said that next time they 're-appear' they'll have a public demo that people can play, and that it will be fairly soon.

I read the recent Edge magazine article and it made me a little more convinced about this. Things like the engine only rendering what is on the screen and not the entire world. He is also of South African descent which explains his odd accent :P

There is going to be an interview with him on Good Game (an Australian TV show so probably not accessible to other countries) this week so maybe we'll know some more about it soon.

It definitely does sound too good to be true. That's the point.

@Grumbel said:

@benpicko said:

It definitely does sound too good to be true. That's the point.

The thing is, it can be true and a fraud at the same time. Looking at their tech demo it is easy to notice that the game world is blocky as hell, each of those blocks has a ton of detail of course, but the general world layout is no more detailed then Minecraft. This means they can reuse the same data- over and over again, which cuts down the memory need a lot., one of the things that would likely kill the engine when used in a real game. That's not to say that the engine is useless. Back before 3D acceleration there have been a few games build around voxels. Outcast, which I just finished, for example seems to be build on pretty similar tech and it can pull of quite a few impressive tricks for a game from 1999, so over ten years later it's not hard to imagine that you could put a little more detail in the voxels. But then, even doing so doesn't give you a drop in replacement for currently popular engines, it gives you a interesting tech demo with some pretty troublesome limits and the way from there to a finished game can be very long.

Yeah, but assuming he wasn't lying in the interview, that's not a problem with the engine, it's because they don't have enough artists.

Yeah, but assuming he wasn't lying in the interview, that's not a problem with the engine, it's because they don't have enough artists.

I'll believe them when a game is released with the tech and runs at 60fps.

I haven't watched the full interview yet, I guess it's way too early to say if it's a fraud or not yet, but @benpicko's points really don't say anything one way or the other either. It's probably common practice for both frauds and actual projects to be marketed this way. Something to keep in mind when it comes to frauds is that they, for example when it comes to perpetual machines, cold fusion, zero point energy, and all that often is "we're not finished yet, but the breakthrough/big announcement/unveiling is just around the corner". That they put their real names out there isn't an issue (Harold Camping and Andrew Wakefield come to mind as people who've been disproven again and again, but keep attracting followers), Patrick Klepek mentioned confirmation bias a couple of weeks ago, and this stuff generally relies on the same thing - people remember the "hits" and forget the "misses", in this case this would be important when it comes to media attention - announcing an extraordinary project, experiment or find will always get more attention than when it's disproven or found to be fraudulent, as "Yeah, that arsenic microbe life form we had a lot of coverage about 6 month ago turned out to be announced way prematurely and probably was wrong" is not as newsworthy as "ELVIS SIGHTED IN KENTUCKY, SCIENTISTS BAFFLED!", and then people remember the big announcement, but not how it turned out to be all BS. Even if established businesses understand and remember that something doesn't work, upstarts could still be tricked into spending money for something that would seem to give them an edge against the established competition, and the guys behind this project could just keep saying "No, we're still working on this, but it's going to be finished very soon!".

Let's wait and see for them to release the SDK or more detailed information, until then, it's not impossible that they're on to something, and for us as consumers it probably doesn't matter either way, but upstart game developers should probably be a bit cautious in spending their budget on a license for this tech.

It's just him doing the same bulllshit again, saying all of this shit and not backing any of it up. Mean while, the very technologies that Notch mentioned that this guy is saying are bullshit are actually showing their shit off, being transparent about the limitations and coming up with ways to get bast the limitations. We have yet to see anything with lighting, post processing, or interactivity beyond moving a camera. Not to mention how much tiling exists in EVERYTHING they have shown of this "unlimited detail" bullshit. You can't repeat the same set of objects in exactly the same rotation just translated around and say "LOOK WE HAVE A SHIT TON OF OBJECTS IN REAL TIME." Even with all of the "infinite detail" their shit looks worse than just about any game coming out. Hell, I'd take an Xbox or PS2 game's engine over my games looking like the shit they show off. Even if the things they have shown are really as easy to render as they say, there is no reason to believe it'll be possible to actually have good animations generated by artists, or that lighting will work properly, or that physics will be even remotely possible, or that this technology will every be useful for video games. Could be cool for the medical field or something, but until they show me something like the Samaritan Demo, I'm not going to buy the bullshit, and just saying it's not bullshit and coming off like a huge douche isn't going to get that guy or his company anywhere. Show the fucking technology, be transparent, prove that it works. You don't have to give away your secrets, just build a game with the tech, and make sure it matches or bests modern engines in everything. If you can show me infinite detail with real time lighting, real time reflections, object based motion blur, maybe some nice AO or even better global illumination, then maybe we can talk. But until I see a video game, this shit means nothing. It'll maybe be cool to run Doom in your browser if it really can do what they say in terms of software rendering and maybe some basic animation, and just make everything really smmooth and weird looking, but probably better than doom/quake era visuals, but that's all I see out of this technology right now.

Please leave the poor horse alone.

There is nothing impressive about this at all, for one they are using repeated models in that world, the world lacks any sort of gameplay, collision detection, dynamic lighting or physics, also notice the enviroment is made up of a alot of right angles. I just find this company crazy there arn't applying any gaming principles into the engine all they talk about is how many things are on screen. They don't talk about can this engine run smoothly at 60fps or about the texture memory needed for the high res textures. They have a very single minded focus, they seem to be oblvious to what actually goes on in the games industry and the different aspects that make up a game.

If they come out with the engine running smoothly with gameplay mechanics then fair game to them but at the moment, this is a joke.

@Legxend said:

There is nothing impressive about this at all, for one they are using repeated models in that world, the world lacks any sort of gameplay, collision detection, dynamic lighting or physics, also notice the enviroment is made up of a alot of right angles. I just find this company crazy there arn't applying any gaming principles into the engine all they talk about is how many things are on screen. They don't talk about can this engine run smoothly at 60fps or about the texture memory needed for the high res textures. They have a very single minded focus, they seem to be oblvious to what actually goes on in the games industry and the different aspects that make up a game.

If they come out with the engine running smoothly with gameplay mechanics then fair game to them but at the moment, this is a joke.

Yeah, as for the technology itself, it's not impossible, though I have a hard time believing their description of the world being built up by individual atoms. Let's assume they aren't making everything up. Part of their description kind of sounds like a classic raycasting engine (like Wolfenstein or Doom), where the "camera" shoots out a ray for every pixel which then hits walls in the world and returns with what should be displayed in that pixel. I'd guess they might be using high res voxels for the objects in the world, maybe some kind of tesselation to smoothly remove detail as the object is farther away. This obviously would mean that there isn't unlimited detail, as there is some point where voxel graphics are just voxels, three dimensional pixels (also, I'd think that infinite detail would also mean infnite file sizes, as the detail data has to be stored somehow). They make a lot of comparisons to what kind of computer graphics are used in medical applications, and as far as I've been able to dig up, the most common medical and scientific visualization techniques are VoluMedic and Volume Rendering, the first which is a combination of polygons and a proprietary format that seems to be voxel based, built using sequences of 2D images, and the other being purely voxel based. Neither have unlimited details.

@selbie said:

I read the recent Edge magazine article and it made me a little more convinced about this. Things like the engine only rendering what is on the screen and not the entire world. He is also of South African descent which explains his odd accent :P

There is going to be an interview with him on Good Game (an Australian TV show so probably not accessible to other countries) this week so maybe we'll know some more about it soon.

Engine rendering what is on screen and not the entire world isn't something new, in fact its been going on for quite a while in game engines.