Comment by hackuser | Hacker News Reader

Comment by hackuser | Hacker News Reader

Parent Full thread

They discuss trying to mimic the focus of real human vision, making the foreground in sharper focus than background and peripheral objects. Why is this necessary? My human vision will do that itself.

For example, if I look at the room I'm in, foreground objects are in sharper focus. If I look at the screen of the computer I'm typing on, if I stare at this letter X, maybe 1-4 letters to each side are in sharp enough focus to read.[1] Why would the computer need to blur the non-foreground images further?

The only guess I have is that objects on a screen are in a false perspective, based on the screen size and the typical user distance from it. Cramming the real world onto a screen that is tiny relative to the immersive real world requires some tricks. But that's just a guess - does anyone know?

----

[1] It depends on the font size and probably on resolution too, of course.

I think you have the idea backwards a bit. The system isn't interested in making the background blurry, it wants to make your exact eye point the section of the image with the highest fidelity. Then the renderer can just sort of half ass the rest of the image. It's a way to make the apparent image better than what the hardware would allow.

It's not just sharpness of focus. There's that, because of - I believe - how densely different areas of our retinas are populated with cones and rods. That's not the same thing that's going on when you look at a tree in front of you, and then the mountains miles away.

The screen is flat. If you were playing a game that occupies 3D space, the game could mimic real depth by changing how it looks depending on where you're looking. It would blur things as they'd be blurred in real life. Or maybe they'd double up things close to us when we look far away, like when you hold a finger in front of your face and then stare far away, or vice-versa. I think we probably rarely notice these things in real life. Maybe this would make it feel magically more real.

You're confusing a few ideas here. Focus has to do with the fact that the real world can only be projected on to our retina sharply at one distance. The letters in your peripheral vision are hard to make out not because they are out of focus but because they are falling on a part of the retina that has low resolution (also see "visual crowding"). The wikipedia page on optical focus has a good explanation of focus blur. Suffice it to say real scenes have dynamic focus blur and computer screens don't and we can subliminally tell the difference. This tech will make rendered scenes look more like real ones.

Reply on news.ycombinator.com