scooby_dooby said:
And that argument is false. Period.
Some fuzzy logic in this post. Just saying this gen, most large games have been rubbish, doesn't change the fact that with more space there's the definite possibility of using it to the games advantage, if nothing else to add more variety. eg. A 1GB level on DVD that spans 40 screens-worth of scenery by tiling the earlier graphics at the end fo the level, could be upped to 4 Gb on BRD and provide more variation in the graphics without repeating the earlier tiles etc.
As to whether this will happen or not, that's all down to speculation, and you can point to that not happening this time around as evidence that it doesn't matter. That doesn't change the fact that more space = more options though; a very valid argument.
Having now read the article, it's not a conclusive debate (shocker!). 50 ish percent increase isn't bad, but I think they underestimate how much resources can be gobbled up. The idea of better compression techs coming along to save the day sound iffy to me too. I can't see we're going to get any better compression. You've got limited lossless compression, or something like JPEG. If you want more compression it means losing more fidelity. I don't know that's there's any new ways to encode data so you don't suffer those quality losses.
If we take another consideration, rather than comparing launch title growth, compare XB360 to XB games in terms of estimated content size. If an XB game takes up 3 GB, XB360 at on average 4x the data size (more complex models, higher-res textures, yadayada) would be 12 GB. And that's not allowing for more diversity. Consider a game like Baldur's Gate Dark Alliance 2. There's a few levels in that with really dull repeating critters. Only like 2 monster types with a couple of variations. An XB360 version of that sort of game would take up several times the data with more detailed models and textures. Then if you add in more monster varieties you're taking up even more space. If the monsters have 4x the poly counts and 4x the texture size, and there's 3x as many monster varieties, that's a good 12x the storage requirements.
If you're only doubling the data sizes, aren't you limiting the increase of quality to 2-4x? Aren't we expecting an order of magnitude improvement into poly counts (model detail) and the like?
