Just to expand on Dan's points a bit, there's also a bunch of other things you need/want to make sure the players see when they boot the game, such that the "Press Start" screen is a way of saying "The player has acknowledged the game": Stuff like selecting save data location/destination, "This game saves progress automatically, look at this icon for that", online pass prompts (e.g. Need For Speed Hot Pursuit), and a few other things the game might need. If you're particularly curious about seeing this in practice, create a fresh, online-enabled, Xbox Live profile sometime, load up something like Hot Pursuit, and write down every screen you see until you get to the game's main menu.
It could also be used as a way to reduce any initial disengagement on the user's part if the initial load time is long. Two 15 second load times will feel shorter than one 30 second load time if you need to press a button in the middle of them: If you can move some bookkeeping/loading after the user has pressed a button, instead of before, then that's a nice way of going about things.
" Well the animator could have done contract work for Bizarre Creations. Studios within the same publisher don't exist in a vacuum. But I do hope Activision isn't making a Bond game for the sake of making a Bond. I fear for Raven's future D: "
I saw the reel before it got taken down (there was maybe a 3 minute window between Joystiq putting it up and, apparently, the animator going "OH FUCKSHIT" and setting it to private and removing it) and it was NOT the stuff that Blood Stone was, not by a long shot.
There was actually a solid 2-3 minutes of footage (some greybox/maya animations, and some of actual gameplay) in that reel...it was quite startling to see.
Every with releases every five years, the amount of RAM in, say, the Playstation has increased by about a factor of 8 each time. Given that we probably won't see successors to the Xbox 360 and PS3 for a couple of years, I'd be shocked if the amount of RAM in each console was less than 4 GB total.
@buckybit: " About implementing 3D Support in your games: If all your objects in the game are already 3D models, you can use some sort of tags or triggers to define the 'depth' effects through a graphic card driver? "
Yeah, there are ways to take just the depth data of the scene and use that to fake the 3D, but it will look very iffy, and have a bunch of problems due to missing data (note that it'll look like a film that's been upconverted to 3D). Keep in mind that for true 3D support, what you want to know is, "Are you doing stereoscopic rendering?", since THAT is what provides for the best 3D experience. For most games, this isn't too difficult to add, if you don't care about performance/memory, since you can "simply" render the scene twice: once where the camera normally is, and again about 3 inches to the side, and feed one frame to each eye. The thing is, a GPU driver can't do that magically, and the program needs to add specific support for that functionality.