What advances in realtime 3d graphics would you like to see.

Can you please let me keep some of my precious hair??
Everything at equal distance from what you look at, that is in peripheral view, is also in focus. If I can’t read the post while looking at the file menu it’s because (not scientific answer ahead) the brain is only capable of reading text at the very center of our field of view. WHile I look at the file menu the thread is still sharp and clear, I just can’t read it. If it was out of focus it would be blurred.
You are confusing width and height of field with depth of field
The narrower the focal length of the viewpoint, the more blurred the out-of-focus stuff will be. Narrow depth of field (especially on macro shots) makes phots look impressive. It would look impressive in a 3d envirnoment, but as many have pointed out it would make playing the game difficult as the player would have to aim at everything he/she wants to look at.

An experiment follows…
First hold up two fingers. One close to your face and one at maximum arms reach. Place both fingers so that vertically they are next to each other. Now look at the near finger. Without changing your focus on the near finger notice the far finger. It’s blurred. It’s out of focus.
Now hold both fingers up, both at arms maximum stretch, both in front of you, so that they are about 30cm (a ruler’s length) away from eachother. focus on one finger. Notice the other - it’s sharp and clear, because it’s in focus. That is, it is it is on the edge of the same imaginary sphere as the other finger.

But the most important question: Is it faster? If it doesn’t offer a performance improvement, then it’s not superior.

And, really, you don’t need to pay someone to develop several models… just make the hi-res model, rip the textures out of that, and then downsize the model to lower poly count. Just what they did in Doom 3.

I don’t know about you, but ANYTHING I’m not looking directly at is most certainly not in focus, whether it be a finger or a part of the computer screen. It may be blurry for a different REASON, but it’s still blurry.

And besides, I still don’t know why you want to replicated the effects of a camera lens when trying to simulate what the human eye does.

But where do you get the lower poly count model? Either you have an artist make it, or you have an offline algorithm to make it (and some number of other versions, taking up precious RAM), or you have a realtime algorithm to make it.

Anyhow, I don’t see that as an area of CGI that is particularly in need of Massive Revolutionary Advances…

We obviously have a different idea of what “in focus” means. When I do your experiment the other finger is most definitely NOT sharp and clear, It may technically be in focus because all the light rays are being correctly focused onto your retina but it is not sharp, the details are blurred I can’t make out the lines clearly etc (maybe my eyes just don’t work as well as yours, who knows). As I said to the previous poster, I accept that everything is actually in focus on your TV screen, but the effect is as if the parts you are not looking at are slightly out of focus.

My point is that I don’t see anything useful about making what you are not looking at, even less clear than it already is. It would sabotage your ability to react to unexpected events by picking up movement.

Another point. Your eyes have a much larger depth of field than what you seem to want to simulate. The effect that a camera gets with a narrow depth of field is actually quite unnatural. If you go outside and have a look around you don’t have the huge disparity between things that are in focus and those that aren’t.

You seem to be interested in making a game look cinematic, i.e., you’d like to have some of those neat camera effects that you see in film, like when the frame is on two people, one close and one a little farther away. The farther away dude is talking and in focus, he stops talking and the camera refocusses on the close dude for his reaction.

I am more interested in making a game realistic, as if I was there, not as if I was watching a movie. In real life, everything I want to look at is in focus, because I’m looking at it. I don’t have to think about looking at it or focussing on it, I just do it and it happens.

I don’t believe it would be practical to achieve this. And if it was, I still don’t think it would be useful. A neat idea perhaps, but not ultimately useful.

Use the plugin that says “reduce polys”. It may need some tweaking, but if you have an artist/modeller already there making your full poly version, asking him to make a decent reduced-poly version takes just a few minutes.

On this note, I just started playing through Resident Evil 4 again, and sure enough, they do use this effect, in real-time, in the game. However, its used only during the cinemas, obviously for cinematic effect. So it’s definitely doable, but just as you, myself and othesr have mentioned, impractical for actual gameplay scenarios.