Likely that the next consoles will have 1GB or more of RAM?

[Phil]

I remember reading this article at Gamasutra, where one author proposed the idea that perhaps in the future, textures won't be used anymore and that polygons themselves will be painted. He went on to say that textures are a horrid nuisances with their caching, compressing, transfering them etc. and that once characters and game-worlds become more detailed, the need of space for those textures will arise.

What are the chances next generation systems might incorporate such an approach? Is this fairly realistic, can it be achieved?

I'll see if I can find the link to that article - if I do, I'll make sure to post it.

 

[Vince]

Back from the mini-holiday - hope everyone enjoyed it.

I remember reading this article at Gamasutra, where one author proposed the idea that perhaps in the future, textures won't be used anymore and that polygons themselves will be painted.

This was briefly talked about in the article from Gamasutra that I referred to - it was a GDC presentation/paper from a guy named David Brickhill who assumed a baseline increase from this generation to the next as the last to this. He saw the bottlenecks (as already seen on PS2) moving even further away from processing power to bandwith and RAM/storage space. So, the paper was on incredibally dense meshes and he talked of ways of storing them effeciently - wavelet and subdivision I remmeber. If I wasn't such a lazy turd who forgot his password, I'd link you.

 

[Tagrineth]

I remember reading this article at Gamasutra, where one author proposed the idea that perhaps in the future, textures won't be used anymore and that polygons themselves will be painted. He went on to say that textures are a horrid nuisances with their caching, compressing, transfering them etc. and that once characters and game-worlds become more detailed, the need of space for those textures will arise.

What are the chances next generation systems might incorporate such an approach? Is this fairly realistic, can it be achieved?

I'll see if I can find the link to that article - if I do, I'll make sure to post it.

Well, the thing is, once you get polycounts THAT high you have to worry about Z-fighting in the distance (already a big problem in some games)... and in general, that many vertices will take a LOT of space.

 

[Fafalada]

A large (one could say the bulk) portion of our studio was populated with content creators and the process of creating it was what pretty much consumed most of the development time...

Well, artist/programmer ratio in our office is something like 4:1 (it used to be even higher at some point).

Well, the thing is, once you get polycounts THAT high you have to worry about Z-fighting in the distance (already a big problem in some games)...

Eh, what kind of distances do you want to draw? 100km?
Anyway, In my experience ZFighting anywhere in the scenes is mostly result of lazy modelling, and rarely, if ever, ZBuffer precision (16bit Z would be problematic more often, but if you choose to go with that you Know you need to pay more care to it).

Space is another issue though. I'd think displacement mapping with something akin to high resolution normal maps of the model would be simpler and more straightforward way of compressing all that vertex data then what that article suggested. And all you need for it is a VU-style vertex processor with a larger block of low latency memory. (without resorting to a full hw implementation that is).

 

[Phil]

This was briefly talked about in the article from Gamasutra that I referred to - it was a GDC presentation/paper from a guy named David Brickhill who assumed a baseline increase from this generation to the next as the last to this. He saw the bottlenecks (as already seen on PS2) moving even further away from processing power to bandwith and RAM/storage space. So, the paper was on incredibally dense meshes and he talked of ways of storing them effeciently - wavelet and subdivision I remmeber. If I wasn't such a lazy turd who forgot his password, I'd link you.

Yeah, that's the one. I actually found it before, here's the link if anyone's interested.

Definately an interesting read... I remember being quite fascinated by the idea he proposed...