hi everyone,
I am 31, I m belgian, so excuse my english,
I ll try my best.
3D games tries to recreate the reality on a flat screen.
and on this point everyone could help the programmers,
even me ! or you !
at least, I think …
I will discuss this point.
You certainly remember the ancient egyptian paintings and sculptures ;
no perspectives at all : a man by profile represented with 2 legs and 2 arms ,
we know that in reality we only see 1 of each.
This realy changed with the European rebirth around 1500 and leonardo da vinci,since then,
we judge the perspective as the “good” 2D representation of 3D reality.
From wolfenstein , through quake , to RTCW-ET ,this is the perspective that programmers use to
figure 3D reality.
But I am not satisfied of it !!!
When I look with my own eyes a scene of the real life,
when I look straight forward I almost see a 170 degres angle,
if I turn my eyes on the left and on the right, the angle is about 260 degres.
And that’s very important, vital .
In RTCW-ET and others , the best is 90 ,because there is no deformation of the scene,
but try to modify this : /cg_fov 160 (here is 2 captures),
http://yann59.free.fr/1fov90.jpg
http://yann59.free.fr/2fov160.jpg
play 5 minutes and go to the bathroom to throw up .
The brain have too much work to recreate the world we are in, because there is too much deformations.
Why is that ?
How many times did you go through a door and didn t even see
an oponent that obviously you couldn t have missed in real life ?
Sometimes it s funny , I shoot an oponent in his eyes but the player doesn t see me !!!
Am I the only one to be annoyed by that ?
I assume that most of the time ,our eyes , I mean hardcore gamer’s eyes, don t even MOVE !!!
Just as the eyes of movies spectators, film directors knows that , they never place an important
scene at the extreme right or left.
We spend almost all the playing time looking at the crossair,
when we need to look to the right we move the mouse to the right,
but our eyes are still looking the crossair !!!
OK , sometimes we look around with our eyes, lets say in an ellipse around the crossair ,
but rarely at the extremes up , down , right or left , and never the 4 corners of the screen.
In RTCW-ET , when we look with the binoculars ,
we see that :
http://yann59.free.fr/3binocularsellipse.jpg
an ellipse , and black around it.
the real essential area of the screen, the area that we look 95 per cent of the time
is an ELLIPSE centered on the crosshair.
So here is my idea :
within the ellipse we see with a cg_fov 90;
and
outside of it, the thin area between the ellipse and the limits of the screen,
we condense a cg_fov 160,
in order to figure the rest a our real field of vue.
here is a possible result:
http://yann59.free.fr/4mix.jpg
in this exemple the ellipse is too small , the contraction I tried with photoshop didn t work as well
as I espected, it s like we were diving with a mask , but it s pretty new , you have to
imagine this in movement , with a bigger ellipse, with a thinner area outside the ellipse ,
and a better method of contraction , or a different form , like one between ellipse and rectangle
with rounded corners, perhaps a smoother transition between the 2 fov.
Of course the thin space is completely compacted, condensed, but the essential is here, comprehensible,
a doorway on the right , the lamp above it, the tank on the left, the wires on top,
and this area is not so important, so we doesn t need to SEE it without deformations,
we just need to GUESS it , like in the real life .If we detect something threatening , we turn .
Is my idea new ?
What do you think of it ?
May be, someone else already had this idea, tested it , without success,
let me know.
nay
