[MUD-Dev] Re: FOVs and rendering tricks to vastly improve immersiveness
ceo at grexengine.com
ceo at grexengine.com
Thu Jan 2 20:45:52 New Zealand Daylight Time 2003
on Tue, Dec 17, 2002 at 06:59:53PM -0500, Amanda Walker wrote:
> On 12/17/02 4:33 PM, Brian Hook <brianhook at pyrogon.com> wrote:
>> People rarely noticed, and what it did was give a sense of
>> massiveness to everything. It also gave you a sense of speed,
>> which was important. As speeds have decreased to become more
>> realistic, games have felt slower and players have complained.
>> For a hoot, look at the size of a doorway in a typical FPS. It's
>> about the size of a garage door in real life. If you were to put
>> in a real sized doorway (30-36" wide), the amount of precision to
>> navigate through there at full speed (30mph) is tough to acquire.
>> It looks like a slot. If the door was sized normally in height,
>> you'd constantly feel that you were about to slam your head into
>> top of it.
> Well, some of that is a result of the wide-angle field of view
> (which also increases the sense of speed, I realize). In FPSs I
> often narrow down my fov a bit to make things more "realistic".
> It also makes it easier to aim, which is pleasant.
...I go the other way, making it easier to see people sneaking up
I've never gone this far, except in non-competitive gaming
(I.e. just mucking about), but Panoramic Quake is quite cool...
For serious competitiveness, I can heartily recommend Fisheye Quake:
(I think the link has been posted here before, but just in case
anyone missed it).
I also once designed and implemented a prototype of an FPS engine
with "peripheral vision". Basically, where a traditional 3D engine
projects the transformed polygons from 3D onto a flat rectangle to
get a 2D image, I projected onto three rectangles, which were
adjacent sides of a cuboid. By keeping the left and right rectangles
much narrower than the centre one, you ended up with an easy-to-use
traditional main projection, squarer than normal, with two flaps on
the side, where everything appeared to be rather strange.
I was aiming to simulate the *effectivness* of real peripheral
vision - e.g. you can't see well far away from the centre of your
vision, but you CAN see very effectively anythnig moving right at
the VERY edge of your vision.
In the projections outlined above, things near to the seam between
adjacent view-planes "jump" across the border, and are hard to track
visually for e.g. aiming. But anything right at the extreme outer
edges is actually easy to track and look at. And the strange seam is
a reminder that the left and right edges are beside you, rather than
in front of you.
Crummy diagram :)...
\ vmain /
vl I\ /I vr
I \ / I
o = apex of viewing frustrum / focal point of viewer / inside the
vmain = traditional planar slice through the view frustrum (viewed
from on top)
vl = my additional plnar slice, at 90 degrees to vmain, with a
projection of everything to the left of the player, but "thinner"
than vmain - so that you can't see much behind you as well as to
vr = just like vl.
The slashes are the view frustrum, viewed from the on top.
The relative strangeness of all this is because I ultimately wanted
to create a game where different players / races (or suffering
various temmporary powerups etc) had considerably different views of
the world. Later versions of Moraff's World had a nice feature where
if you played as a Giant, the 3D FPS was rendered with an extremely
high view origin, so that you actually felt in danger of scraping
your head along the ceiling as you played.
The triple projection above would easily allow "predatory" races,
with little or no vl and vr, and "prey" races, with small vmain
compared to vl and vr - making it much easier for them to see
creatures sneaking up on them.
Predators would perhaps get other bonuses, like limited Fisheye, to
make aiming MUCH easier, within a traditional 90 degree (or
MUD-Dev mailing list
MUD-Dev at kanga.nu
More information about the MUD-Dev