Water effect (Flooding/Rivers)

[jackofthetrade32]

Water effect (Flooding/Rivers)
I think they need to add a better water effect.
If it is raining, it should flood if there aren't any gutters.
If some water is running on a sink, it should fill up correctly and should drain onto the floor.

Better yet. There should be storms that cause mud and...
However, storms and stuff should wait until HL3
It wouldn't take too many more millions of lines of code to do it.
I'm a programmer and I know I could do it if I had there code with their authorization.

 

[Shuzer]

Currently, realtime fluid dynamics/"physics" take up too much processing power and can't be achieved at playable flamerates with modern, mainstream hardware.

 

[Neo_Kuja]

Yup, needs too much processing power to do it effectivily.

 

[Mechagodzilla]

Fluid dynamics won't be possible for a long time. And, if you think about it, they aren't really necessary in most scenarios. They'd just be a nice addition, like Red Faction's breakable walls.

 

[Mr-Fusion]

There'd suddenly be all these liquid related traps integrated into the gameplay.

Like a bucket of water sitting on a door slightly open, the combine walks in and GET SOAKED!!! COMPLETELY SOAKED!! AHahahahah1111

 

[63]

No fluid water dynamics yet! Check back in 5 years!

 

[Polykarbon]

I'm not trying to question your ability as a programmer, but you should already understand that it's not possible with current hardware. This if of course assuming that you realize how to render fluid dynamics with the right algorithims. You should know that alg properties like continuity and Navier-Stokes are solutions bound by computational power. Actually, there are so many problems with creating dynamic fluid that a full blown solution is still being brought together. There's flow control, vortex motion, and forces within the fluid itself. The exact programmable equation is tough, although it of course isn't impossible.

Not denying you, just letting you know that you won't see it until atleast Unreal Engine 3.0+.

Cauchy's equation of motion is a pretty guiding example of how the basic properties of motion for it will be brought together.

 

[redundant]

I'm a programmer and I know I could do it if I had there code with their authorization

then why aren't you working for valve?

 

[OoGoff]

If u wanted real time fluid dynamic and you where really lazy, you could make them 3d balls that would roll and tumble with the normal physics calulations.

Or make them entities like a blob. So if X amount of little circles come together it will mesh it into a blob that just does a blur effect through it. To get the motion right you would have to adjust the friction against the blob. Imagine a blob of jelllo and how it acts. Well thats how u could do it. If it comes to a ledge as soon as a certain amount of the blob (a percent depending on how large/ so the larger the blob the smaller the %) and would revert into a circle form until it hit the ground or whatever. I think toy get the point.

 

[craig]

If u wanted real time fluid dynamic and you where really lazy, you could make them 3d balls that would roll and tumble with the normal physics calulations.

actually... at present this is the only possible way to simulate 'fluid dynamics'.

Pixars Renderman software, which was used on Finding Nemo basically did what you're describing, because calculating the movement of minute particles just isnt possible (even with their renderfarm).

They had '3d balls' to simulate movement of water (small, but a helluva lot bigger than a water particle) and when they found something they liked (movement wise) they basically had the software have just one mesh cover the balls to make it appear like fluid.

I think I read like 4/6 different things were done to create realistic looking water. And this was in the best water animation I've ever seen.

 

[Kschreck]

It's possible to simulate certain water conditions including flooding. Heck Half-Life 1 had some scenes where the water would flood a chamber. It's just a matter of realism. It looks as if Half-Life 2 has quite a good water system compared to other games. The water is rendered nicely. Say you wanted a room to flood like in Half-Life 1, you could have the water pooring in and just make the water in the room rise making it look like it's actually flooding and to add even more to the realism, have the water waving back and fourth and a few other small effects around the source of the flooding. I believe this stuff is possible in Half-Life 2 if Valve wanted it to be, however I never saw a specific scene of a water source flooding into another water source.

 

[jet jaguar]

You could do it, but then the game would have to be in text.

 

[Pi Mu Rho]

Or you'd have to have each frame posted to you every week, after it's rendered.

 

[Iced_Eagle]

A cool effect they have in Splinter Cell 3 is that rain dynamically changes the specularity of objects to make them look wet. Maybe HL2 will use a similar thing. There are also spots where water is told to go, and it makes a puddle and the color changes making it appear deeper.

Some sort of trickery like that could be done to make everything appear really wet.

Raging rivers in real-time? Let's all hook up our computers just so we can run that in real-time, eh?

 

[Polykarbon]

That isn't simulating true dynamic fluid. That is simply creating a solution to overly static water for a situation that benefits the creator.

The calculated process of dynamic fluid has NOT been entirely solved. I would agree that the method you described is a good way to duplicate the effect, but it is not even close to a real solution.

Using sphere formations under a dynamic mesh of triangles, with a vertex a particle being initialized is great for an effectual application, but that would never work well in a game.

True dynamics are still years off. Anyone saying otherwise is wrong. I expect it so come soon, but not as soon as some of you might hope for.

 

[ferd]

lol
He's a programmer, and he he want's real-time dynamic fluids.

 

[redundant]

i somehow doubt his credibility as a programmer if he thinks he can implement real-time fluid dynamics into half life 2 on today's hardware.

not supposed to sound like a flame, but keep dreaming.

 

[figge]

As some people said... you can always 'fake' it... (real time rendered water is faked to... but a diffrent fake... )

 

[PvtRyan]

You're a programmer right? How should I interpret that? Oh wait, you made a Geocities site about Dragonball Z, right?

 

[DEATH eVADER]

I reckon when Valve has released Halflife2, they will start upgrading Source. Starting with adding fluid dynamics

 

[Impulse147]

ok so fluid dynamics aren't possible, would it be possible to allow it to rain and the give everything a kinda wet look, so dirt changes to mud, and also when u step in it it sounds different then it did b4 it rained......i'm gonna e-mail this too valve.

 

[Rupertvdb]

You could use the dynamic terrain to create these water effetcs. In worldcraft you have to have an sctual physical block rise, if the dynamic terrain is as malleable as it is in the E3 2003 vids then it shall simply be a case of having a small bit and just 'shisting' it. That shall give as close to dynamic fluid as is possible, in my mind.

 

[The Dark Elf]

"allegedly" textures in Source can be changed in game. Which "might" allow for the wet look during rain. Someone will have to ask Valve about that

-

The water effects, im pretty sure this was mentioned in the Valve info thread, are calculated during compile. Which means they likely aren't going to work on deformed surfaces. I think this was why the water is now flat rather than having the wavey look HL1 had.

This isn't so bad though, water moving that much isn't going to be reflecting an awful lot anyway. So just create an animated rushing water effect to apply to a moving deformed brush of sorts. That possibly could do the trick.

 

[Impulse147]

yea rushing water is quite possble, dod has it.

 

[Rupertvdb]

the only problem with the brush system in that insteance is that you need amazing quality textures and really high face brushes or else it just doesn't look that good. Hopefully decent sprites can disguise the look of the thing.

Water is always going to be a bitch in games until fluid dynamics are feasible.

 

[Vegeta897]

You're a programmer right? How should I interpret that? Oh wait, you made a Geocities site about Dragonball Z, right?

ROFL

Even I have made one of those! :dork:

Yeh, do't expect fluid dynamics in games for a while.

The best they can do is those cool splashes that I have seen around, the water height map changes dynamically in response to objects interacting with it.
But even that wouldn't cut it, and would probably look bad. What Valve has now is perfect.

 

[The Dark Elf]

the only problem with the brush system in that insteance is that you need amazing quality textures and really high face brushes or else it just doesn't look that good. Hopefully decent sprites can disguise the look of the thing.

Water is always going to be a bitch in games until fluid dynamics are feasible.

Source has particle generators of sorts. So using those to help give the impression of rushing water and mist to go with it might help hide the lack of detail. And I suppose could always add props made of polygon layers and animated to further add to the effect. Be far from perfect but its possible to an extent.

 

[Sprafa]

MOH: Pacific Assault "pretends" to simulate water in a very convincing way, IMO.

the Mecqon physics engine also has a "pretend" liquid simulation taht doesn't consumes THAT much but gives a good "pretention" of liquid simulation.

 

[Rupertvdb]

Source has particle generators of sorts. So using those to help give the impression of rushing water and mist to go with it might help hide the lack of detail. And I suppose could always add props made of polygon layers and animated to further add to the effect. Be far from perfect but its possible to an extent.

So particles could just make a decent disguise over the joints in the water... i'm sure there shall be ways.

the Mecqon physics engine also has a "pretend" liquid simulation taht doesn't consumes THAT much but gives a good "pretention" of liquid simulation.

Nah, i think the fluid is that demo kind of sucked, very buggy...

 

[Sideswiped]

crap.. ignore post

 

[Flyingdebris]

i know unreal 2003 had the ability to do interactive water. however it looked a tad weird, like you were interacting with an indestructable jello mold

 

[jet jaguar]

I could see a solution in the near future(5-10 years) for small bodies of water(no bigger than a swimming pool) where water is defined as particles the size of golf balls or marbles since it'd be nearly impossible with current technology to render water in a game real-time at a molecular level. So basically you'd have a pool full of 'water marbles' that collectively react like realistic water would(waves, splash, form puddles through interaction of gravity and slopes, evaporate and disappear...).

 

[jackofthetrade32]

Well, to tell you the truth. I knew that it will be another 5+ years before anything like this would happen on the PC side.
I really posted this to see what type of response I would get.

I know I can programming all of these effects, also I know that with the current technology that it really isn't possible to do all that I want done.
Programmers have always been in front of technology. We just wait for the day that some new technology will come out so we can put our real skills in place.
Also, we would move A WHOLE LOT fastest if we got rid people that shouldn’t be in the business to begin with. India! (They need to fix their own country first)

When technology is ready, I will have working code that should work in a hologram set if it is ever invented in my lifetime. But, we need to start using more biological based technologies to compete that task at hand.
Thank you
Live long and free

 

[sonkev]

SPLINTER CELL 3 HAS FLOODING WATER EfFECTS, WITH ITS NEW ENGINE

 

[Rupertvdb]

but is that fluid dynamics because if it isn't you just pulled a stupid.

 

[Raxxman]

I think one of the major problems is just how complicated fluid dynamics are. I study fluid dynamics for my masters, and I can tell you that prediciton of turbulent flow (IE a molecule of water flowing down a river) is going to be impossible

however the ball idea is a nice plasuble workaround but I can see it falling down around stuff like if water splashes and hits the ground.

 

[Daiceman9]

but is that fluid dynamics because if it isn't you just pulled a stupid.

His caps lock key tripped him into the puddle of stupid.

 

[Rupertvdb]

i read a book on chaos teory and it described why water flow is so awkward, in a uniform flow the precition is fairly simple, once you have anything at all introduced you lose the simplicity is a split second, it changes froma uniform thing to a totally complex and beyond predictable model, a sinlge molecule can still be travelling further down the stream (or whatever it is in) and still be influencing the flow. This can obviously be simplified to create a fairly good dnamic system but the cut off point is ( i imagine) hard to find and even harder to get consistent.

 

[Vegeta897]

but is that fluid dynamics because if it isn't you just pulled a stupid.

He pulled a Charley Gordon!
(Cheers if you understood that)

I gotta look into that...

 

[EVIL]

dammit.. so I cant take a 50 feet high water tank down so everything gets under water? :'(

ahhwell.. I still have reality for those things *attaches explosives to water tank*