FPS with photo realistic graphics?
Daniel
Aren't they just textures, or are the current computers not fast enough
to handle this?
Michael Frayn
Mike
Greg wrote in message
<16A46573A0D09476.FB7D524B...@lp.airnews.net>...
>On Fri, 03 Mar 2000 17:21:10 -0500, Daniel <dan...@amazingmedia.com>
>wrote:
>>Why can't they make something like Unreal with "photo quality" graphics?
>>Aren't they just textures, or are the current computers not fast enough
>>to handle this?
>
>Higher rez textures, higher poly counts. All of that and more will be
>necessary for true photo realism. It's probably not but a few years
>off, though. I think Tim Sweeney over at Epic has said as much.
Slash
scribbled:
>That will be the day. I don't really care about alot about the
>current FPS, but give me photo realism and we will all become addicted
>all over again. It will be like the first time I played Doom.
Duke Nukem Forever seems to have more of a 'realistic' look than every
game I've seen recently, at least from the screenshots. Should be an
interesting game if it gets released before I die. :)
-Slash
"In the beginning, there was nothing, which exploded."
- Terry Pratchett
Destroy
> I don't really care about alot about the
> current FPS, but give me photo realism
I guess thats what every gamer and game maker shoot for but what happens when
we actually get to photo realism. Then what? What will we have to look forward
to after that?
rob
I don't think we're anywhere near that point. Our 'virtual worlds' are
simplified, polygonal little playgrounds, only impressive to those
who know how bad video games _used_ to look.
In most areas there's no progress at all. Even with the fastest cpu and
video
card, you will still be trapped behind a little two-dimensional
porthole,
wagging a stick to interact with the world, with no sense of smell
or touch.
I'm not too worried about perfection.
Destroy
> or maybe there will be some kind of revolution in level design that is
> beyond my ability to foresee. :)
Yup. Game editors will build the scenes for them via use of 3D cameras
capturing say for example a busy office room and rendering/inputing it
into the game. The game editor will be told to seperate the different
objects like people, desks and plants then store similar data in huge data
banks for later game scenes or use. Thats the way I see it.
Now for make believe scenes like Star Trek, I haven't clue as to how
they'll do it.
Rikk
current FPS, but give me photo realism and we will all become addicted
all over again. It will be like the first time I played Doom.
On Fri, 3 Mar 2000 15:24:31 -0500, "Michael Frayn" <mfr...@wlbc.net>
wrote:
Antos Dragon
The problem is not so much as the required 32-bit color, but rather
the # of polygons that can be pushed by current hardware. The first
issue is, of course, curved surfaces. This was tried in Unreal on
models, and in one area, it could bring down the game to a whopping 2
FPS. (Naliboat.UNR), so it was removed from the game (even though the
option remained). Quake3 has a better implementation of this on
architecture, but it still isn't what you would call 'perfect'. Now
imagine EVERY item in a game that is 'supposed' to have curves having
near-perfect curves, then you start to see the problem. The amount of
Polys would be insane, and no card could render that fast enough, and
of course, I don't think any customer level CPU on PCs could calculate
all of that mess.
And that is just curves...not to mention the other details...
I'm talking about HUGE texture sizes (2048x2048 and higher) and
more...
Of course, this is where compression comes into play...
I know nothing about rendering techniques, but I believe we need an
entirely different way of rendering to get that much detail, much like
Intel is redesigning its chips (x86 'emulation'?) to break the limits
of current mhz capability, and the old drawbacks to the tired x86
architecture.
Just take a human arm...you would have to model all the curves, the
blood vessels the hairs ,scars, the ridges...
Of course, the goal is to have photorealistic scenery in a 3D engine,
photorealistic models, architecture, textures, and so forth but we
aren't there yet. Hell, even each grass strand could be made of
polys, instead of just being a texture. But doing that would give you
about 1 FPS at best.
Look a Shen Mue..even THAT isn't photorealistic, even though the
amount of detail visible is astonishing.
Just relax man...we are getting there slowly...we have made VERY large
strides, but don't ask for a leap before the walk...
Just compare a game such as Wheel of Time or Quake3 or UT (or even,
egads... Ultima IX), to Doom or Quake1 or Duke3D. Then you can
appreciate exactly how much we've accomplished.
On Fri, 03 Mar 2000 17:21:10 -0500, Daniel <dan...@amazingmedia.com>
enlightened us by scribbling this gem of wisdom:
>Why can't they make something like Unreal with "photo quality" graphics?
>Aren't they just textures, or are the current computers not fast enough
>to handle this?
Remove the part of my email that doesn't belong there, to email me..
We are the Borg...
You will be assimilated...
Resistance is futile...
Andrew R. Gillett
> Just think how long it's going to take for game companies to actually
> build photorealistic games. I wouldn't be surprised if they have to
> go to some kind of "standardized scenery," or something. Either that,
> or maybe there will be some kind of revolution in level design that is
> beyond my ability to foresee. :)
One thing which will help a lot is writing code which creates much of the
graphical detail itself, without an artist having to exactly specify it.
For example, in Republic: The Revolution, pavements/sidewalks, roads and
bricks and things like that will have fractally generated cracks. So even
if there was only one pavement texture, all the pavements in the game
would not look identical.
--
Andrew Gillett http://argnet.fatal-design.com/ ICQ: See homepage
"Aaaaaaaaaaaaaaaahh." - Richard Herring
Worker Working
>
> One thing which will help a lot is writing code which creates much of the
> graphical detail itself, without an artist having to exactly specify it.
> For example, in Republic: The Revolution, pavements/sidewalks, roads and
> bricks and things like that will have fractally generated cracks. So even
> if there was only one pavement texture, all the pavements in the game
> would not look identical.
>
Exactly. We are at the point where it seems ludicrous to do these
things by hand. There should be programs that do layouts for standard
items and then let the artists go in and add touches to the scenes.
Mark Stevens
wrote:
>Wow, that's gonna be damn cool :)
Cool to look at, perhaps, but will it be as good to play? Gaming
history has shown that it's always innovative game design that pushes
things forward.
--
Mark Stevens
Eep²
Daniel wrote:
> Why can't they make something like Unreal with "photo quality" graphics?
> Aren't they just textures, or are the current computers not fast enough
> to handle this?
--
http://tnlc.com/eep/ - Active Worlds, Tomb Raider, 3D game comparison
Enable line/word wrap if text not wrapping!
Eep²
> > Either that,
> > or maybe there will be some kind of revolution in level design that is
> > beyond my ability to foresee. :)
>
> Yup. Game editors will build the scenes for them via use of 3D cameras
> capturing say for example a busy office room and rendering/inputing it
> into the game. The game editor will be told to seperate the different
> objects like people, desks and plants then store similar data in huge data
> banks for later game scenes or use. Thats the way I see it.
>
> Now for make believe scenes like Star Trek, I haven't clue as to how
> they'll do it.
The same way they've BEEN doing it ever since staged theater began: sets, props, and costumes--oh my! Duh. And throw in some computer animation for special effects.
Banjo Bagel
>That will be the day. I don't really care about alot about the
>current FPS, but give me photo realism and we will all become
>addicted all over again. It will be like the first time I played
>Doom.
Also say hello to 3 year release delays and multimillion dollar
budgets. Unless gaming becomes extremely mainstream, I imagine
there will be a limit to how photorealistic games can get.
--
Banjo Bagel | banjobagel_...@hotmail.com
IE5 Whore Website: http://members.home.com/ajyi
"He who thinks he knows, doesn't know. He who knows
that he doesn't know, knows. For in this context,
to know is not to know. And not to know is to know."
- Tao-te Ching
Destroy
> things by hand. There should be programs that do layouts for standard
> items and then let the artists go in and add touches to the scenes.
So basicly we need autoeditor engines for the game editors.
bottled water
look at Q3 & UT. The graphics are superb but it is far from realism. It
looks like all the colors are exaggerated (sp?). The buildings looks like
they were just built the day before the tournament, all shiny and clean, the
paint were probably still wet. The edges are sharp and there's no thrash on
the floor. Take a look around your room (right now) and you'll notice that
color usage in real life are in the light side. I thought System Shock 2
was more realistic looking though, other than those aweful polygon people.
Daniel wrote in message <38C03AD5...@amazingmedia.com>...
Destroy
> history has shown that it's always innovative game design that pushes
> things forward.
Tell that to id. Seems they are a huge force in our gaming society and I
see little innovation of design from their products. Now if you wanna
talk Valve, then we got something. :)
Michael Frayn
>Cool to look at, perhaps, but will it be as good to play? Gaming
>history has shown that it's always innovative game design that >pushes
things forward.
Very true, but having those graphics will certainly help the realism
factor, which is very important to me. Innovative game design doesn't
necessarily take advanced technology.
Mike
Andrew R. Gillett
> The main problem is that the programmers go overboard with colors. Take a
> look at Q3 & UT. The graphics are superb but it is far from realism. It
> looks like all the colors are exaggerated (sp?). The buildings looks like
> they were just built the day before the tournament, all shiny and clean, the
> paint were probably still wet.
Better than the bland graphics of games such as Half-Life.
--
Andrew Gillett http://argnet.fatal-design.com/ ICQ: See homepage
Gandalf takes the small key from you. Gandalf says "What's this?"
Thorin takes the small key from Gandalf.
Thorin says "This was Thrain's key."
Trent
>
> Why can't they make something like Unreal with "photo quality" graphics?
> Aren't they just textures, or are the current computers not fast enough
> to handle this?
Photorealistic graphics didn't help adventure games. I don't see why
they would help FPS's. I'd rather have characters generated by an
artist with actors just doing the voices. All the actors I see in
computer games seem like castaways from bad porno films.
--------> Trent
Thomas j. Evans
>Rikk wrote:
>
>> I don't really care about alot about the
>> current FPS, but give me photo realism
>
>I guess thats what every gamer and game maker shoot for but what happens when
>we actually get to photo realism. Then what? What will we have to look forward
>to after that?
Hmm... how about innovative gameplay... compelling stories...
interesting characters... you know, the stuff that game designers
don't give a hoot about right now.
-T
Harvester
<ghe...@airmail.netgetridofthis> wrote:
>I think it'll be the same as now. For every Half-Life or Duke Nukem
>Forever, there will be ten bargain bin bound FPS. Maybe as these
>things get more and more difficult (and expensive) to make, the less
>talented teams will be forced to find something else to do.
is Duke Nukem Forever already a certified FPS classic?
Mark Stevens
wrote:
>Very true, but having those graphics will certainly help the realism
>factor, which is very important to me. Innovative game design doesn't
>necessarily take advanced technology.
If there's any constant in videogame development, it's that technology
will evolve to allow more complex/realistic visuals. That's a given,
so it's not really worth daydreaming about it -- it'll happen
eventually! I'm just wondering when the next Elite, Dungeon Master,
Carrier Command, Populous, Civilization or Doom is going to arrive.
Mark Asher
wrote:
Heh -- funny, isn't it? DNF has been delayed almost as long as
Daikatana.
Mark Asher
Grantie Boy
games and then you'd be in a world of hurt and you'd lose your job,wife,
income the start lobbing grenades at your neighbours and end up in an
institute.
Daniel <dan...@amazingmedia.com> wrote in message
news:38C03AD5...@amazingmedia.com...
Andrew R. Gillett
> On Sat, 4 Mar 2000 19:43:34 -0500, "Michael Frayn" <mfr...@wlbc.net>
> wrote:
>
> >Very true, but having those graphics will certainly help the realism
> >factor, which is very important to me. Innovative game design doesn't
> >necessarily take advanced technology.
>
> If there's any constant in videogame development, it's that technology
> will evolve to allow more complex/realistic visuals. That's a given,
> so it's not really worth daydreaming about it -- it'll happen
> eventually! I'm just wondering when the next Elite, Dungeon Master,
> Carrier Command, Populous, Civilization or Doom is going to arrive.
Elite 4 will probably be released in roughly two years time. There is a
new version of Carrier Command being planned as an open-source project by
a group of fans, with me as one of the designers. Civilization III will
probably be out next year. Some people are speculating that Doom 3 will
be id's next project. There is no sign that a new version of Dungeon
Master is going to happen anytime soon. And the forthcoming Black and
White is the true descendant (as opposed to Populous 3) of Populous.
--
Andrew Gillett http://argnet.fatal-design.com/ ICQ: See homepage
"You're right there, Murray." - Martin Brundle, F1 97
Mark Stevens
<arga...@fatal-design.com> wrote:
>Elite 4 will probably be released in roughly two years time. There is a
>new version of Carrier Command being planned as an open-source project by
>a group of fans, with me as one of the designers. Civilization III will
>probably be out next year. Some people are speculating that Doom 3 will
>be id's next project. There is no sign that a new version of Dungeon
>Master is going to happen anytime soon. And the forthcoming Black and
>White is the true descendant (as opposed to Populous 3) of Populous.
Well, I didn't literally mean sequels for all those games. I was just
wondering what the next truly innovative game is going to be. My
money's on Republic.
D. Booth
Antos Dragon <kyle...@NOSPAMmy-deja.com> writes:
> I could be wrong about this, but here's my weak explanation.
>
> The problem is not so much as the required 32-bit color, but rather
> the # of polygons that can be pushed by current hardware. The first
> issue is, of course, curved surfaces. This was tried in Unreal on
> models, and in one area, it could bring down the game to a whopping 2
> FPS. (Naliboat.UNR), so it was removed from the game (even though the
> option remained). Quake3 has a better implementation of this on
> architecture, but it still isn't what you would call 'perfect'. Now
> imagine EVERY item in a game that is 'supposed' to have curves having
> near-perfect curves, then you start to see the problem. The amount of
> Polys would be insane, and no card could render that fast enough, and
> of course, I don't think any customer level CPU on PCs could calculate
> all of that mess.
>
> And that is just curves...not to mention the other details...
> I'm talking about HUGE texture sizes (2048x2048 and higher) and
> more...
> Of course, this is where compression comes into play...
>
You're pretty much correct. If you've ever seen how long it takes
to render just one frame in a dedicated rendering package like 3D
Studio Max, Maya or Lightwave, then you'll know what the problems
are. It can literally take hours for *one* frame when you have
serious number of polys, raytracing etc.
I think we need to look at other technologies rather than just
using the brute force approach (exemplified by the Voodoo 5 spec),
where we get ever faster CPUs for drawing polys and ever more RAM
needed for textures. For instance, the PSX 2 uses some kind of
fractal algorithms for natural shapes like trees etc., and we also
need to look into voxels (with hardware acceleration), NURBS etc.
Basically, we've taken using standard polygons to the limit - all
we can do is pump out more and more, rather than actually thinking
of an alternative way to render scenerios. A new paradigm is needed....
--
Dan B. - http://public.connect.org.uk/~dan/
\\\///
o o
Lee H
Snip.
> I think we need to look at other technologies rather than just
> using the brute force approach (exemplified by the Voodoo 5 spec),
> where we get ever faster CPUs for drawing polys and ever more RAM
> needed for textures. For instance, the PSX 2 uses some kind of
> fractal algorithms for natural shapes like trees etc., and we also
> need to look into voxels (with hardware acceleration), NURBS etc.
>
> Basically, we've taken using standard polygons to the limit - all
> we can do is pump out more and more, rather than actually thinking
> of an alternative way to render scenerios. A new paradigm is needed....
>
>
> --
> Dan B. - http://public.connect.org.uk/~dan/
>
> \\\///
> o o
When the amount of polygons pumped out equals the amount of pixels on the
screen they become as flexible as voxels , per pixel rendering is what we'll
get using the same principle as today's 3D accelerators , hardly efficient
though.
Lee H