Polyphony Digital Should use this for GT6.

[regoon]

http://www.youtube.com/watch?v=00gAbgBu8R4

This will definitely help the improvement of trees and other scenery problems. Cars would benefit too.

 

[another_jakhole]

I only watched the first minute or so, and he says it takes up a lot of processing power. So that being said, if GT6 is released for the PS4. Then, I think it can be possible.

Another thing though, is that the premium car models in GT5 take forever to create, and Kaz has said that they're fit to be on the PS4. Would there be a problem transferring those models that use polygons to point cloud data? I'm sorry if they discussed this in the rest of the video, but this is all I want/need to know. Otherwise, very good stuff.

 

[Da-Bomber]

That's incredible if PD got their hands on that GT6 won't have to have 2d trees.

 

[Aloha62_2]

This has been popping up every 6 months for the past 4/5 years and nothing has come of it so far it also doesnt show what is feasible in the way of shading, textures, shadows and animations.

The same doubts now are as valid as when the company released its other videos several years ago.

 

[PAWEL17]

The ps4 would have to be a huge advancement in technology. But i like the idea of scanning the objects from real life, PD would then have to scan all the car models, and tracks, and then just take all there time on the physics engine. I see this kind of technology happening in a decades time or so. But great find.

This has been popping up every 6 months for the past 4/5 years and nothing has come of it so far it also doesnt show what is feasible in the way of shading, textures, shadows and animations.

The same doubts now are as valid as when the company released its other videos several years ago.

Its a technology company, they are working on the tech, its just up to the industry to use it or not.
But if you say its been in development for so long, some things take time, and if its feasible, i have the same doubts aswell, but the gaming industry has to advance, and this seems the way they will eventually go....hopefully.

 

[Machate-man]

This will definitely help the improvement of trees and other scenery problems. Cars would benefit too.

Unless you can run it on today's Ps3, XboX or Wii, then it's worth it.
I'm not investing 600+ on a new console for some 'atoms'

 

[dhandeh]

A Gran turismo game with no 2d trees just wouldn't be right.

 

[another_jakhole]

Oh, nice. Thank you PAWEL17. And interesting Aloha62_2. Obviously, I had no idea.

A Gran turismo game with no 2d trees just wouldn't be right.

Oh how right you are.

 

[CSLACR]

Has anyone asked what TV can actually display their claimed graphics worth crap?

It's easy to talk about being limited by pixels, but TV's only display so many, IF this works as they say it does, it'll be like playing a great PC game on a 240i monitor.

I also noticed all they showed was scenery, and zero gameplay at the point I got tired of watching.

 

[GfxJG]

I won't be fully convinced until they create something alive with that. Sure, scenery is good and all, but I want to see characters, or at least things moving.

 

[SigmaViper11]

Has anyone asked what TV can actually display their claimed graphics worth crap?

It's easy to talk about being limited by pixels, but TV's only display so many, IF this works as they say it does, it'll be like playing a great PC game on a 240i monitor.

I also noticed all they showed was scenery, and zero gameplay at the point I got tired of watching.

I didn't hear anything about being limited by pixels. i heard no limits on polygons which would be a lot different.


I have hope for them. Like any tech it will take quite a while to work out the bugs. But it looks promising. I can't wait to see what it can do. Imagine how everything else in the game could run. Particle effect, physics, etc. . . If it can be optimized to work with todays equipment it would be like bringing a new world order to gaming.

 

[CSLACR]

I didn't hear anything about being limited by pixels. i heard no limits on polygons which would be a lot different.


I have hope for them. Like any tech it will take quite a while to work out the bugs. But it looks promising. I can't wait to see what it can do. Imagine how everything else in the game could run. Particle effect, physics, etc. . . If it can be optimized to work with todays equipment it would be like bringing a new world order to gaming.

But it can't. Today's equipment will have to be modified to match the system, they even say in the video something along the lines of nothing can run it.

i heard no limits on polygons which would be a lot different

Ahem - So this system will magically make my TV resolution change?
I have a PS3 on 480i, and guess what? It displays at 480i, not what the software can max out at, 480i.
So take the highest resolution the general gaming population can afford (1080) and that's the point of which there is no point in currently improving graphics beyond.

 

[PHCharls]

That's great idea

 

[Griffith500]

This is very interesting. It's quite clear that this stuff will almost certainly not appear in this generation, and probably not the next - probably not for some time, in fact. Hardware vendors are notoriously slow to change, and this new technique needs new hardware to make the most of it, much as polygons did back in the mid-nineties. Note also that polygon graphics has existed since at least as early as the eighties. Then there's developer inertia. ATI and nVidia introduced (separate) tesselation "solutions" years before DX10 / 11 was strutting around with it draped over its shoulder. It's only really Microsoft and the (ATI) Xenos chip in the XBox 360, coupled with Microsoft's developer support strategies, that has brought tessellation to the fore.


What's more interesting is they say this "atom" stuff runs at 20 fps in software. Right. But that software is running on some hardware or other, so I'm curious as to what that is. Presumably they mean it's not some bespoke graphics hardware, rather some sort of traditional "CPU" architecture. Although, that would potentially include vast arrays of multiple units of said CPUs, like traditional render farms.

As for 1080p being the "highest resolution", that's neither here nor there. Not everyone plays games on a TV, and the future is always but a blink-of-an-eye away.

 

[deep_sky]

Polyphony should use Naughty Dog and nothing else.

 

[Griffith500]

Polyphony should use Naughty Dog and nothing else.

Sure, if they decide to make another third person action game, maybe. I'm sure they can handle that themselves, though. The focus and range of compromises are different for the different kinds of games the two developers are known for. Both companies have quite solidly stuck to what they know, but have made both third person action games and racing games.

 

[another_jakhole]

Sure, if they decide to make another third person action game, maybe. I'm sure they can handle that themselves, though. The focus and range of compromises are different for the different kinds of games the two developers are known for. Both companies have quite solidly stuck to what they know, but have made both third person action games and racing games.

CTR!! One of my all time favorites.

And good heavens with your last post, great information. Really informative.

 

[Griffith500]

CTR!! One of my all time favorites.

And good heavens with your last post, great information. Really informative.

Apparently, this stuff is very similar (if not identical) to Sparse Octree Voxel rendering. The stuff that's (provisionally) slated to be in idTech 6 (Rage is idTech 5).



Also this:

Re Euclideon, no chance of a game on current gen systems, but maybe several years from now. Production issues will be challenging.

From Twitter.

Just relayin' the info.

 

[another_jakhole]

Just did a couple quick searches because I only barely know who John Carmack is (side note - my brother definitely knows about this more than I do) and I watched that video, but all I have to say is, We Need People Like Carmack. From the little bit I know of him, his tweet saying,

Re Euclideon, no chance of a game on current gen systems, but maybe several years from now. Production issues will be challenging.

, just goes to prove if PD keeps the car models they've created for next gen, we really are just seeing the beginning of the future of GT (example - http://www.gtpla.net/wp-content/uploads/2010/12/gran-turismo5-interior-photomode19.jpg).

If that makes any sense. Basically, the limits the PS3 has on GT5 really are a problem for Kaz's perfectionist attitude.

I hope I'm on the right page because the future of GT is more important than the present. If you want, take out "GT" and put in "gaming", because that's what matters most. I've been ready for the future of gaming, and here I was thinking we were already there.

 

[VBR]

That is absolutely mind blowing, & I had multiple technogasms whilst watching it, I'm off now to take a cold shower!

 

[JamboGT]

The problem with this is the memory size of models which would be huge and then the animation. Every frame would require the movement of each molecule. Basically the demo had no moving parts for a reason.

Bowing to the legend John Carmack it is possible that a few generations away this will start to happen but as of now it is unrealistic to expect it soon.

 

[PAWEL17]

Has anyone asked what TV can actually display their claimed graphics worth crap?

It's easy to talk about being limited by pixels, but TV's only display so many, IF this works as they say it does, it'll be like playing a great PC game on a 240i monitor.

I also noticed all they showed was scenery, and zero gameplay at the point I got tired of watching.

Im sure, by the time, this tech is used, the Tv's or monitors will be able to support it.

 

[JTT]

omg !!!
amazing !!!

 

[Griffith500]

The problem with this is the memory size of models which would be huge and then the animation. Every frame would require the movement of each molecule. Basically the demo had no moving parts for a reason.

Bowing to the legend John Carmack it is possible that a few generations away this will start to happen but as of now it is unrealistic to expect it soon.

We already have polygon grouping for animations via "bones" etc., so I expect a similar thing here. Still doesn't change the fact that a vast number of points need transforming fairly frequently, but this uses some sort of "detail-scaling" algorithm, so that it's not pushing all the points all the time.

This stuff is definitely possible, and has been the subject of much study over the years ("point-based graphics"), and John Carmack, whilst perhaps not necessarily The Last Word™, definitely thinks voxels are the (eventual) future, at least for landscapes (as we've already seen in the likes of Delta Force, Outcast, Crysis, Minecraft () etc.).


However, there is a general impression (on the wider intertoobs) that these particular developers are being less-than-sincere / honest with their claims here for supposedly fraudulent reasons. An impression, mind you.

That's not to say the tech is impossible, more that it's possible these guys haven't actually done exactly what they say they've done...

 

[Raitziger]

No they should not. I think PD is the company that needs the least help from other making good graphics. Content side and workforce could use help though.

 

[Griffith500]

Is this war?

Notch (he of Mojang, they who make Minecraft) calls the whole thing a scam!
Although he's not claiming the tech is rubbish, just these developers.

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.

If you've got the stomach for it, here's a nice article outlining the potential advantages and disadvantages of using voxels, and what it may mean for the future.

EDIT: there may be a better title and home for this thread now...?
EDIT: Some fellow making a voxel-featuring 3D engine of his own also shares his thoughts, with some cool videos.

 

[Bobalob]

I'll wade in with my limited knowledge on this subject, but I was particuarly amazed with the original delta force coming out with actual MILES of distant landscape working perfectly on my piece of **** (pentium 166mhz maybe at the time lawl), so impressed I shot them an email asking why I couldn't use my super doopers voodoo 3dfx card with the game and I got an email back explaining (without detail) it was impossible with the voxel engine. Since other deltaforce games have been released with 3daccel but I was led to believe it wasn't full acceleration.

I also love John Carmack together with ID software they have released some proper next gen games, their follow ups used whole new engines and every sequel was a breakthrough, now we've been sold call of duty 15 times :/

 

[Griffith500]

I'll wade in with my limited knowledge on this subject, but I was particuarly amazed with the original delta force coming out with actual MILES of distant landscape working perfectly on my piece of **** (pentium 166mhz maybe at the time lawl), so impressed I shot them an email asking why I couldn't use my super doopers voodoo 3dfx card with the game and I got an email back explaining (without detail) it was impossible with the voxel engine. Since other deltaforce games have been released with 3daccel but I was led to believe it wasn't full acceleration.

I also love John Carmack together with ID software they have released some proper next gen games, their follow ups used whole new engines and every sequel was a breakthrough, now we've been sold call of duty 15 times :/

Later Delta Force games, i.e. Land Warrior and on, don't use the voxel terrain, which was annoying because you couldn't hide in the grass anymore. The terrain appeared to go on for miles, but I'm sure you noticed the tiling / repeating? It was still perhaps at least a kilometre square of unique terrain per map, but the players and placed scenery objects didn't stop or wrap at the borders (like Ace of Spades, thoroughly recommended, by the way!), so sniping was always fairly hardcore.

Being a 3DFX owner at that time, Delta Force 2 crippled my computer too! It was impossible simply because the graphics cards wouldn't know what to do with the data, and couldn't be tricked using fancy tricks with the drivers / APIs because it was all fixed-function. Of course, newer cards and APIs allow for clever tricks that make real voxels possible in hardware, but people found a way to "cheat" and turn voxels into polygons, too (Crysis).

 

[Zenmervolt]

Ahem - So this system will magically make my TV resolution change?
I have a PS3 on 480i, and guess what? It displays at 480i, not what the software can max out at, 480i.
So take the highest resolution the general gaming population can afford (1080) and that's the point of which there is no point in currently improving graphics beyond.

I'm not sure that you fully understand how graphics rendering works.

While pixel count (and therefore resolution) is important for certain things like textures and anti-aliasing, the number of base polygons doesn't change from 480i to 1080p. While a 1080p game capture will have more pixels it will not have more polygons.

The graphics engine doesn't render each pixel on the screen individually, but keeps track of many levels of polygons. In fact, it renders many times more pixels (and polygons) than the screen displays, but since parts of the 3D image cover other parts, it's not displaying every polygon (or pixel) that it actually renders. For example, when you're watching two cars pass each other, both cars are being rendered, even when you can't see most of one car because the other is in front of it.

The more polygons that can be rendered, the better the overall screen image will be, regardless of the resolution. More polygons means more detail in areas that are seldom seen (e.g. the underside of cars), more detail in background images (e.g. track buildings and spectators), and even things like better collision detection because of more thorough modeling. There would be better damage as well since damage could do more than merely deform an existing polygon.

While you're right that at a certain point there are no more noticeable returns, we're not anywhere close to that point yet, even at 480i.

 

[SigmaViper11]

But it can't. Today's equipment will have to be modified to match the system, they even say in the video something along the lines of nothing can run it.

Ahem - So this system will magically make my TV resolution change?
I have a PS3 on 480i, and guess what? It displays at 480i, not what the software can max out at, 480i.
So take the highest resolution the general gaming population can afford (1080) and that's the point of which there is no point in currently improving graphics beyond.

Indeed you are correct sir, as i was notified a little bit ago, as well as Griffith500 posting Notch's statemen. Its not practically and in some forms not at all yet, usable.

So here's to hoping it may have some form of selective usage in the relatiely near future.