[Theory] Unraveling the Unlimited Detail plausibility

Started by
166 comments, last by Ben Bowen 11 years, 10 months ago

Well, I'm only nitpicking on you :) This stuff is way beyond me.
I only picked on two things: that the colors of the world is mostly only depend on the surface topology, and that it's easier to make art with this voxel magic.

The two parts of my reply were separate. Only the first part was directed at you :P


That's not true. In the video above, John Carmack is completely honest about the abilities of his technology. He could call it "unlimited detail" if he wanted to, but he doesn't. That's because he's a person with enough integrity to tell the truth.

The fact that you so easily give a pass to lies because they're made in order to market a product says a lot about you. You might have a future in business, your ethics are slimy enough for it.

John Carmack also isn't marketing his product in most of his demos. He's explaining it. There's a stark difference. I have heard other people from Zenimax marketing it and it sounds pretty much the same. Similar things have also been said about CryEngine and Unreal engine in different aspects of the engine; IE: "X is 10000 times better at Y than everything else."


@way2lazy2care: Marketing has it's limits too. I haven't heard many marketing people claiming "100000 times better graphics" and feeling sorry for poor ATI/AMD and nVidia for pouring millions upon millions of dollars on the ugly triangles...

I've heard similar things from almost every major publisher and manufacturer.

Also, mathematics put a limit at the compression ratio when you do lossless compression (see information entropy). Replicating the same elephant and tree 1000 times does not increase the entropy much. If you want to replicate a real (not procedurally generated and with instancing all over the place) square kilometer island at milimeter detail, however, you are hopelessly screwed no matter your compresion techlonogy.
[/quote]
That's a valid thing to be curious about, but we haven't seen how their data is compressed for the scenes we've seen or what could be done if an artist were allowed to just go crazy with it. If I had to model an area that size down to the pebble I'd probably reuse a ton of stuff too just to save time. This is something we'll have to wait and see more on whether they are reusing the same data or using copies of the data existing in different places.
Advertisement

John Carmack also isn't marketing his product in most of his demos. He's explaining it. There's a stark difference. I have heard other people from Zenimax marketing it and it sounds pretty much the same. Similar things have also been said about CryEngine and Unreal engine in different aspects of the engine; IE: "X is 10000 times better at Y than everything else."


How is what John Carmack is doing in this video different from what Bruce Dell is doing in the last video posted? It's the exact same thing. You just say anything you can to win an argument, without caring about truth or validity, don't you? What's the stark difference? You should be ashamed of yourself. Do you work for Euclideon or something?

And you're lying about Zenimax marketing MegaTexturing in a similar way. Show us.

Also, mathematics put a limit at the compression ratio when you do lossless compression (see information entropy). Replicating the same elephant and tree 1000 times does not increase the entropy much. If you want to replicate a real (not procedurally generated and with instancing all over the place) square kilometer island at milimeter detail, however, you are hopelessly screwed no matter your compresion techlonogy.

:blink: Wait you think they're using a lossless compression algorithm? You realize that no computer game uses lossless compression for textures normally right? I guess this whole time when I was imagining their system I was picturing their compression guy probably has thought about this 100 times more than myself when it comes to compression and come to the conclusion that a lossy technique would possibly lose very little data while allowing for a much more compact format.

Then again a nice lossless format would be epic. :)

[quote name='D_Tr' timestamp='1313082336' post='4847772']
Also, mathematics put a limit at the compression ratio when you do lossless compression (see information entropy). Replicating the same elephant and tree 1000 times does not increase the entropy much. If you want to replicate a real (not procedurally generated and with instancing all over the place) square kilometer island at milimeter detail, however, you are hopelessly screwed no matter your compresion techlonogy.

:blink: Wait you think they're using a lossless compression algorithm? You realize that no computer game uses lossless compression for textures normally right? I guess this whole time when I was imagining their system I was picturing their compression guy probably has thought about this 100 times more than myself when it comes to compression and come to the conclusion that a lossy technique would possibly lose very little data while allowing for a much more compact format.

Then again a nice lossless format would be epic. :)
[/quote]

I actually had the geometry in mind when talking about lossless compression. They almost surely perform lossy compression on color and normals (assuming they do not calculate them on the fly). I automatically assumed, however, that they do not throw away any geonetry information when converting from polygonal to octree, because geometric detail is supposed to be a strong point of voxel technology.

I actually had the geometry in mind when talking about lossless compression. They almost surely perform lossy compression on color and normals (assuming they do not calculate them on the fly). I automatically assumed, however, that they do not throw away any geonetry information when converting from polygonal to octree, because geometric detail is supposed to be a strong point of voxel technology.


This is true. But for storing geometry information you only need 1.5bit per voxel which is really really small (and nice). Material information (color/normal/specular/emissive) can be stored in seperate textures that are strongly compressed both on disk and in graphic memory (DXT1 format for example).

In this thread I've tried to explain how you can get such good compression:
http://www.gamedev.n...36-disk-octree/

It is actually quite simple.

EDIT: In previous posts you had a fight about color inheritance in child nodes. This kind of inheritance is completely unnecessary because if color is stored in DX format texture, it will always be very well compressed - there is no need for bother with inheritance (will produce messy code). Monochromatic surfaces will be compressed very well because of DX format inner workings.

How is what John Carmack is doing in this video different from what Bruce Dell is doing in the last video posted? It's the exact same thing. You just say anything you can to win an argument, without caring about truth or validity, don't you? What's the stark difference? You should be ashamed of yourself. Do you work for Euclideon or something?

And you're lying about Zenimax marketing MegaTexturing in a similar way. Show us.


What do you not get about Bruce Dell being a marketing guy and John Carmack being an implementation guy? They talk to the press differently. When you read MSDN blogs they read differently than when you listen to an interview with Steve Balmer, because Balmer is marketing and MSDN blogs are explaining. If you want to see it for yourself go to GDC and walk around the show floor. You'll hear it from EVERY representative on the show floor; they sound just like Bruce Bell. You can even go back to the conference away from the show floor and hear people that work at the exact same companies on the exact same products talk to you the way John Carmack explains stuff.

In fact I'm fairly sure I've heard iDtech 5 pitched with the line, "unlimited texture detail."

Thanks for keeping it civil though.

What do you not get about Bruce Dell being a marketing guy and John Carmack being an implementation guy? They talk to the press differently. When you read MSDN blogs they read differently than when you listen to an interview with Steve Balmer, because Balmer is marketing and MSDN blogs are explaining. If you want to see it for yourself go to GDC and walk around the show floor. You'll hear it from EVERY representative on the show floor; they sound just like Bruce Bell. You can even go back to the conference away from the show floor and hear people that work at the exact same companies on the exact same products talk to you the way John Carmack explains stuff.

In fact I'm fairly sure I've heard iDtech 5 pitched with the line, "unlimited texture detail."

Thanks for keeping it civil though.


It's hard to keep things "civil" with you because you continually make things up to back your "arguments". I've gone through it with you in multiple threads regarding multiple topics. You're the type of arguer who doesn't care about reaching the truth, you care about coming up with some argument that can counter what someone just said, ignoring how connected to reality it may be.

Bruce Dell is a marketing guy? What makes you say Bruce Dell is a marketing guy??? Bruce Dell is a programmer.who founded his own company. John Carmack is a programmer who founded his own company. Both give demos about their tech to the press. BRUCE DELL IS A PROGRAMMER! I suspect you know this, you just are so incapable of admitting you're wrong that you cannot admit it.

Here's how it went. You were arguing that Bruce Dell's description of his tech is feasible. Most people disagreed, and gave plenty of reasons and evidence to that point. Rather than say "Okay, you guys are right, he's a liar" your brain immediately starts searching for a way to shift the argument to avoid being wrong, so you start talking about how we shouldn't be so nitpicky because this is "marketing talk" and everyone does this. When i point out that not everyone does it, with the example of John Carmack who could lie about the "unlimited detail" of his product but chooses to instead give us the truth, your brain starts looking again for a way out, and you land on the lie that Bruce Dell is excused from it because he's a "marketing guy", when in reality you know that he's a programmer just like John Carmack is.

Your inability to argue points like a big boy is disgusting and is to the detriment of this whole forum.

It's hard to keep things "civil" with you because you continually make things up to back your "arguments". I've gone through it with you in multiple threads regarding multiple topics. You're the type of arguer who doesn't care about reaching the truth, you care about coming up with some argument that can counter what someone just said, ignoring how connected to reality it may be.

Bruce Dell is a marketing guy? What makes you say Bruce Dell is a marketing guy??? Bruce Dell is a programmer.who founded his own company. John Carmack is a programmer who founded his own company. Both give demos about their tech to the press. BRUCE DELL IS A PROGRAMMER! I suspect you know this, you just are so incapable of admitting you're wrong that you cannot admit it.

Bruce Dell is a CEO. John Carmack is a Technical Director. Bruce Dell released a marketing video to hype his product and company to everyone, and John Carmack was recorded talking about his product at a developer's conference to show it to developers.

Here's how it went. You were arguing that Bruce Dell's description of his tech is feasible. Most people disagreed, and gave plenty of reasons and evidence to that point.[/quote]
Clearly it's feasible.John Carmack even agrees it's feasible. Most of this thread I wasn't even arguing about his technology, and just because you argue loudest doesn't mean nobody else disagreed with you either.

Rather than say "Okay, you guys are right, he's a liar" your brain immediately starts searching for a way to shift the argument to avoid being wrong, so you start talking about how we shouldn't be so nitpicky because this is "marketing talk" and everyone does this. When i point out that not everyone does it, with the example of John Carmack who could lie about the "unlimited detail" of his product but chooses to instead give us the truth, your brain starts looking again for a way out, and you land on the lie that Bruce Dell is excused from it because he's a "marketing guy", when in reality you know that he's a programmer just like John Carmack is.
[/quote]

Obviously he was marketing from the start. When he used the word "infinite" everyone's brain should have said, "oh. infinite. That's impossible, so it must be marketing," but instead they jumped to, "infinite?! Clearly his product is totally fake and brings nothing to the table."

By all means keep turning to insults though. That's a much better argument strategy.
Okay, okay, I get it. The joke's on me. You're a troll.

Your inability to argue points like a big boy is disgusting and is to the detriment of this whole forum.


Okay, okay, I get it. The joke's on me. You're a troll.


Clearly I'm the troll...

edit: I just finished watching the 40 minute interview he did. I really don't see the issue with it. He backs up most of his claims and doesn't really attack anybody.

This topic is closed to new replies.

Advertisement