View Single Post
  #8  
Old June 15th 11, 09:46 PM posted to alt.comp.periphs.videocards.nvidia,comp.games.development.design,comp.graphics.algorithms,comp.graphics.api.opengl,microsoft.public.directx.graphics
Wolfgang Draxinger
external usenet poster
 
Posts: 13
Default Under water for Battlefield 3.

On Wed, 15 Jun 2011 21:36:51 +0200
"Skybuck Flying" wrote:

That's where the challenge is...

Try to make it look good, try to make it look interesting and full of
detail.

Like under water plants, fish, water bubbles, big water bubbles of
explosions.


This is not a algorithmic problem, at least not on the graphics front.
This is a artistic challange, and maybe a problem of proper physics
simulation.

And explosions under water, in reality they don't look like you think
they do: The key characteristic of an explosion is, that it creates a
shock wave, shock waves travel with the speed of sound (in water about
1000m/s). A shock wave consists of a high pressure front (shock front)
and a low pressure rear. If the pressure difference is large enough, on
the low pressure side the vapour temperature of water drops below 0°C
(this is called cavitation) and high pressure means a local rise of
temperature. So at the shock wave boundary the water is literally
boiling like in a kettle. But only at the shock wave boundary, which
radially expands with 1000m/s; behind it, it condensates immediately.
Due to the expansion of the shock wave, the pressure drops with 1/r²,
which means that in a few meters from the explosion the pressure
difference is too low to cause cavitation, which defines the limiting
radius for the bubble sphere shell an underwater explosion creates.

Fishes are just a challenge in NPC design, and texturing. Scales would
probably rendered using some tangent shading model. Plants are, well
plants, like on land, only they move differently; slower and more fluid.

There are probably no games out there that really do it well, so it
might be an under developed graphical thingie.


Well, it's not a matter of graphics algorithms; optics work quite the
same way under water, than they do outside. I mean, we (humans)
literally live in an ocean of an optical dense medium called air.

Perhaps sea-creature-shells can be added to beach for extra beach
details, perhaps ocean/sea/beach bottom as well with more detail.


Again an artistic challenge. And if you raise the amount of detail also
probably hitting the resource limits of current GPUs.

Soldiers could even get hurt by swimming to close to sea correl which
might cut them up.


That's game logic, totally off topic in all but one of the NGs you're
crossposting.

There could also be maps with big builders in the beach sea or
air/sea tunnels which can be swim through.


Well, if you're patient an are willing to wait for the game I'm
developing with some friends: There'll be a cave diving map.

In the far future say 15 years from now, devices might be so
powerfull that lightning and glass/water effect/reflection could even
be calculated... and water would be dripping between soldier fingers
and over their heads when they come out of the sea...


glass/water effects can be rendered already, they're just not very
accurate at the moment, because they work on an approximation of the
environment. Anything beyond that requires the use of realtime
raytracing. Technically our GPUs are already performant enough to do it
at low resolution (640x480 or so). So give it maybe 4 years only.

But the realtime refraction effects we have today are already good
enough, you'd have to study a single frame very carefully to make out
the errors of approximation.

So there is still a lot to be done at that front !


At the moment I see only one open front in game graphics development
and that is good looking dynamic indirect lighting.

Even rainbow effects of water splashing against the sides of a harbor
might be possible or so


Those are easy. The principle of diffraction of light in water mist is
well understood and so easy to emulate.

What modern game development really needs is a new challenge on
the gameplay frontier. FPSs, RPGs, racing games, combat games haven't
changed in years, they follow the same principle ever since. Invent
some new kind of gameplay, that's where we need to move on. I mean,
just look how popular Minecraft is, and it has ****ty graphics if you
compare it with modern standards.

Graphics, duh, a game doesn't get good only because it has kick ass
graphics. I still prefer good old DeusEx or System Shock over umm, I
don't even remember the names of the countless, soulless games I've
played in the last years. But create some innovative gameplay, like
Portal or Mirror's Edge, and you'll have a place in my heart forever.


Wolfgang