• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

I found an amazing physics game engine demo from 2008 on YouTube. Will we finally achieve this on the 9th generation of video game consoles?

svbarnard

Member

It's only 2 minutes long so watch the whole thing through. This came out in 2008 but it was way ahead of its time, watch it you'll see what I'm talking about, I mean it's 2022 and we still don't have a full-fledged video game with lifelike destruction like that. But I do think the PS5 and Xbox Series X will finally have the power to have that kind of destruction in video games.

On a side note I'm still so pissed off at Microsoft for bringing out the weak Series S, because it's going to be the limiting factor for the entire 9th generation of video game consoles.

So what do you guys think, do you think we will finally eclipse a physics game demo that came out 14 years ago?
 
Last edited:

svbarnard

Member
An honorable mention would be this physics game demo that came out in 2013, do you think we will finally achieve this with the PS5 and Xbox Series X?

 

M1chl

Currently Gif and Meme Champion
Yeah it fucking sucks, that devs are not doing more in the physics front, I no longer care about how the game looks, game already looks good enough. The problem with many of these engines are that you need to make things specifically for it, like for example, you need to make wood, tag it as such, make all those materials when wood breaks, so it does not look like glass and so on. However there are technologies, which are sort of helping this and in UE5 there are a lot of them. Physics can be run, as probably one of few things, completely on compute (those ACE units), however even normal Zen 2 CPU is powerful enough to handle shitton of stuff on screen. So lets hope together.

This is why CONTROL is king from last gen.
 

LordOfChaos

Member
Cellfactor: Revolution?




It used the Ageia PPU:


Thats why it looked so good, it was a HW card handling the physics.

edit - oh yeah that video, I think we can just do that in software nowadays tbh, I've seen that kind of rigid bending in games since then I'm sure.

The PhysX card was absorbed into Nvidia GPUs and we have far more power to spare for this now, so really it's about development choices

There was a time you could even use a second weaker GPU for the PhysX card, but single GPUs are easily many many times more powerful than the Ageia PhysX card now alone

"10 years later" in 2016 haha we're all old as shit

 
Last edited:

JOEVIAL

Has a voluptuous plastic labia
When I saw 2008, I knew the physics engine you're referencing.

That would be the Euphoria engine (the demo in the video is a plug-in for Euphoria from what I remember). Euphoria soft body physics were used most notably in The Force Unleashed, GTA IV, and Red Dead Redemption. It's a crying shame that plug-in was never used during the generation, but it's because consoles at the time could not handle the real-time calculations required for such physics simulations.

Anywho... yes the AAA space has not really "advanced" since the 360/PS3 generation. Publishers have been riding the waves of consumer trends since then. For some reason it's easier to market better graphics than it is to market better physics.

I'm all for advancing the industry forward with better physics, but consumers will have to change how they like their brainwashing in order for that to happen.
 

mdkirby

Member
Yeah it fucking sucks, that devs are not doing more in the physics front, I no longer care about how the game looks, game already looks good enough. The problem with many of these engines are that you need to make things specifically for it, like for example, you need to make wood, tag it as such, make all those materials when wood breaks, so it does not look like glass and so on. However there are technologies, which are sort of helping this and in UE5 there are a lot of them. Physics can be run, as probably one of few things, completely on compute (those ACE units), however even normal Zen 2 CPU is powerful enough to handle shitton of stuff on screen. So lets hope together.

This is why CONTROL is king from last gen.
I think the big problem comes from how stuff is built, and that so much is faked/illusion. Which is both for performance limitations and time/resources.

Even basic things, like say a picture frame, you cant just make the outer SHAPE of the frame, you would need to make each piece of wood its made from for each of the 4 sides (define them as plywood or mahogany in the engine so it knows how those pieces of that thickness should break/behave contextually), how those pieces are fastened together too would need considering, dowel, stapled?, glued?. What its back is made of, the piece of string it hangs from being a piece of string, and also having physical properties, how that is attached to the back. A separate model for the sheet of glass in the front etc. It just vastly increases the workload for EVERYTHING.

For the same reason that most games dont do advanced gore. It means to not just build a 3d model shape of person, bound to an invisible skeleton rig. But actually building their insides, bones, flesh, and having those all play nice within said rig. and then behave correctly when damaged/blown up. Its suddenly a much much bigger task and takes a ton of time and money.

All this just means very few companies have the time or resources to actually do it, and those that do often spend those resources elsewhere. Some of Sony's studios do take that extra time to add that sort of detail, such as the last of us 2 example above.

Common libraries could vastly help tho for some stuff. If something like speed tree for instance added proper destruction/material physics to all their tree libraries, so trees could realistically bend, snap, crack, chop, carve, have leaves pulled or brushed off. Then anyone who implements speed tree into their games (which is most companies) technically be able to instantly have some pretty nice stuff. But its also the sort of thing that could tank performance and may be one of the first things to be cut, particularly if targeting cross gen.
 
As Mdkirby said, this is just "hardware power" outstripping our ability to take advantage of them.

Originally the wish is that you can have the computer brute-force physics into everything. But in the end the only way to make it work is by custom designing everything to the level of detail you could afford. But at some point you realized you can't build a universe atom by atom, you have no time for that. Game studios work hard to create AI that can speed up the grunt work, but in the end you just had to custom make as much as you can.

It's like, we can already have photorealisim; but the money you need to spend to make it is not proportional to what you get back from gamers. Hiring artists is as expensive as ever and computers couldn't replace any of them.
 

CuNi

Member
As Mdkirby said, this is just "hardware power" outstripping our ability to take advantage of them.

Originally the wish is that you can have the computer brute-force physics into everything. But in the end the only way to make it work is by custom designing everything to the level of detail you could afford. But at some point you realized you can't build a universe atom by atom, you have no time for that. Game studios work hard to create AI that can speed up the grunt work, but in the end you just had to custom make as much as you can.

It's like, we can already have photorealisim; but the money you need to spend to make it is not proportional to what you get back from gamers. Hiring artists is as expensive as ever and computers couldn't replace any of them.

I feel that's also the reason why tripple a dev cost and time has been rising so dramatically.

But also, with photogrammetry being used more and more, studios should slowly be able to build up a good sized database of assets over some time and with advences in AI I wouldn't be surprised if the major features of the next generation of engines in like 5 or 10 years includes even more tools to generate content on the fly rather than by artists. Like being able to let AI generate meshes that are a morph between 2 selected, more support for animation, design etc.

Could be also wrong tho. But I feel if we don't solve the problem of modern content taking a shit load of time to produce be it a single model or character for a game, we won't see a big improvement in looks or scope of games anytime soon.
 

Kuranghi

Gold Member
The PhysX card was absorbed into Nvidia GPUs and we have far more power to spare for this now, so really it's about development choices

There was a time you could even use a second weaker GPU for the PhysX card, but single GPUs are easily many many times more powerful than the Ageia PhysX card now alone

"10 years later" in 2016 haha we're all old as shit


Will be good to see that taken advantage of it a bit more in the coming years hopefully
 

Drew1440

Member
The PhysX card was absorbed into Nvidia GPUs and we have far more power to spare for this now, so really it's about development choices

There was a time you could even use a second weaker GPU for the PhysX card, but single GPUs are easily many many times more powerful than the Ageia PhysX card now alone

"10 years later" in 2016 haha we're all old as shit

They could, but that could cut into the graphics, plus Nvidia made it very difficult to enable physx with an AMD video card.
It's a shame CPUs like the cell didn't take off, that could have done wonders for physics.
 

LordOfChaos

Member
They could, but that could cut into the graphics, plus Nvidia made it very difficult to enable physx with an AMD video card.
It's a shame CPUs like the cell didn't take off, that could have done wonders for physics.

Not really a lot with asynchronous compute and generally finding ways to fill in inevitable gaps in execution hardware, it would take a tiny fraction of the power of the GPU to do what PhysX was doing back in the day, see the LTT video which was already 8 years ago where the dedicated GPU card for PhysX doesn't add that many frames most of the time

It's really almost all developer choice, there's also non-PhysX physics libraries that run cross platform. GPGPU now trivially does most of what Cell was better than 2005 CPUs at.
 

GymWolf

Gold Member
I feel like this is already a thing.
It's not even close.

Wood breaks but not with that level of realism.

The first star wars game with this tech had an area where you can bend some trees with the force push but it wasn't exactly on this level of complexity.

Horizon has some trees that bend and break but again, not even close.
 
Last edited:

Clear

Member
The main issue is real physics is kinda chaotic. So the more liberally you apply it, the more likely the chance is you are going to run into ugly corner-cases that waste dev-time patching up.

At the end of the day, a game is just a type of show. So if a solution comes along that allows you to do specific complex, realistic physical interactions in a somewhat predictable manner then that'll nearly always get the nod over a more generalist but potentially finicky solution.

Its always more about effect than anything else.
 

mdkirby

Member
As Mdkirby said, this is just "hardware power" outstripping our ability to take advantage of them.

Originally the wish is that you can have the computer brute-force physics into everything. But in the end the only way to make it work is by custom designing everything to the level of detail you could afford. But at some point you realized you can't build a universe atom by atom, you have no time for that. Game studios work hard to create AI that can speed up the grunt work, but in the end you just had to custom make as much as you can.

It's like, we can already have photorealisim; but the money you need to spend to make it is not proportional to what you get back from gamers. Hiring artists is as expensive as ever and computers couldn't replace any of them.
Yeah, Ai I think will be the thing that can eventually be a huge boon to this and will eventually be able to do a lot of the heavy lifting for asset creation. There's an ai artist whose work I really like, and makes some truly amazing stuff, some of the things he creates by edging an ai the right way puts the monsters of FROMSOFT to shame:

https://billy-hoffman.pixels.com/?f...o8tnyPutx3-b_6t5ETVmcWIzrekZ8DftZf8a7RWZGIzOg

In the same way, currently depending on resources, everytime you walk past 2 NPC's in some games, they will be having the exact same short conversation with each other, every single time. Games that spend more, they may have some variety in there, but you'll eventually hear the same thing again. Those that spend even more, those conversations may start including new chatter based on the story beats/progression. But they will always repeat. The technology already exists for those ai to be given lives/histories/interests and knowledge about both the world and progress of narrative. They could then have very convincing conversations with each other

https://www.theguardian.com/technology/2022/jun/12/google-engineer-ai-bot-sentient-blake-lemoine

There's also extremely convincing ai voice replication via companies like descript https://www.descript.com/overdub , that could take any voice actors voice and have it read out endless conversations written on the fly by a chatbot. Suddenly you could have a world that actually feels alive with living breathing people.

Asset creation will follow the same path IMO. Once all these enormous time sinks can be much more automated a lot of the barriers are removed to nice material physics being in way more games. In the same way that a LOT of indie games in past years have gone for super stylised looks, and/or empty 'dead worlds' or avoided humans. Because making good animated people is hard. With UE5 out now, I recon we will start seeing a lot more indie games going for more realistic graphics, and less narratives that are variants of 'everyone magically vanished, solve the mystery why'.
 

svbarnard

Member
It's not even close.

Wood breaks but not with that level of realism.

The first star wars game with this tech had an area where you can bend some trees with the force push but it wasn't exactly on this level of complexity.

Horizon has some trees that bend when and break but again, not even close.
Thank you so much. Yes I have not seen a game with destruction that looks as lifelike as that 2008 demo. Though I think that's going to change during the 9th generation of consoles.
 

theHFIC

Member
On a side note I'm still so pissed off at Microsoft for bringing out the weak Series S, because it's going to be the limiting factor for the entire 9th generation of video game consoles.
*for multiplat releases. Rival platform console exclusives and PC exclusives will have other limiting factors.
 

th4tguy

Member
I think a lot of it boils down to, for the work needed to implement it, what is the payoff? How does it improve the gameplay overall? How does that element of the game translate to better sales?

I think for most devs, the work did not equal a good payoff. It wasn't going to improve sales of the game by any noticeable amount. Why focus so much hard work and time on crazy physics when you could spend half that amount of time on better gameplay/ performance/ graphics which would equate to more sales.

It's the same with hyper real AI. At a certain point, it just doesn't pay off to have it. Hyper real AI just isn't fun in a video game.
 
Last edited:
The Force Unleashed has some areas where you can do that. Quantum Break uses that physics engine but I think it's prebaked for time ripple effects.
The engine is used in many movies and TV shows. I believe the Hometree destruction in Avatar is the first time it was used outside of games.
 

CamHostage

Member
The main issue is real physics is kinda chaotic. So the more liberally you apply it, the more likely the chance is you are going to run into ugly corner-cases that waste dev-time patching up.

At the end of the day, a game is just a type of show. So if a solution comes along that allows you to do specific complex, realistic physical interactions in a somewhat predictable manner then that'll nearly always get the nod over a more generalist but potentially finicky solution.

And particularly when the pop trends are realistic scenarios and naturalistic (or hyperrealism) look to objects and surfaces, the over-the-top trends of making AAA games specifically as kooky physics playgrounds has fallen off. (Except in the places where it is popular, as actual physics playgrounds like TABS, Boneworks, Beseige, Teardown, and Instruments of Destruction, but those don't have to worry about story or realism or really even "balance" where a physics effect gone wrong accidentally breaks forward progression in the game.)

There are titles like Rainbow Six Siege which are trying to blend realism and play-realism with physics. (And I wish it did it better; adding breachable walls is great and R6S does have rules/cheats to maintain map design with not all breakables, but furniture and decorations should be flying... it is a multiplayer game, so that makes it harder and more susceptible to hurt the ping, but physically-activated objects could only be computed locally and clip like heck for all I care, I just want shit to fly off the shelf.) It's not so common though. Much more often, your prebaked physics are fine for selling an exciting moment or a type of big hit, and then everything else that the player wants to shoot with a shotgun (the pillars, the trees, the windows,) the game designer doesn't really want to be shot up with a shotgun because they placed most of that stuff in the scene to establish the map or to close off access beyond their façade.

BTW, DMM works great for what was shown... applying it beyond that may not as easy as you'd think. For instance, that breakable wood concept, that works great if your surface is plywood, a single sheet surface that can splinter along pre-set or algorithmically designed fracture points depending on impact point. You get tensile strength, you get gravity, you get fibers depending upon each other for strength, you get a remainder of an object which can be further acted upon, all good stuff. That's a 2D plane though (with some abstraction, that they then apply some color or texture inside.) Apply that to a round tree and it's not going to work, because trees do not splinter like plywood. What they actually ended up with in the game did some of this, but the DMM implementation in Force Unleashed was nowhere near the complexity of this single sheet (and in-game may have used a lot more traditional methods of physics simulation in cooperation with the DMM definitions.) People see, "Ah ha, we now have wood in videogames!" and imagine destructible forests and splinterable scaffolding for a building and all that... but what you're seeing is how this piece of wood works, not all wood.




Especially as objects get more complicated, you're lucky if you're running into such simplicity of material here. Somebody else came up with the example of a picture frame, which not 1 thing, it's wood, glass, and paper; each element would need to have different physical properties with their own physics simulation that somehow would also need to collaborate between each other in that one object. Even if your game engine could allow designers could assign "this is glass, it breaks like glass; this is wood, it breaks like wood" (and they kind of do, although wood seems to be a real blindspot for something like Unreal Chaos Physics when I'm reading up about it, not sure how well other physics solutions are doing in this respect?,) only so much of your scene is that simple, and only so much of even the simple stuff do you want to be activated physically in an unexpected way.
 
Last edited:

CamHostage

Member
This is an actual Pixelux demo of a physics playground (probably a sales pitch video, before LucasArts swapped characters with stormtroopers and R2D2s) and you can see the things the actual tech does and doesn't do in relation to physics you may be used to in games today:


Whereas to compare/contrast, here's some things that just the Chaos Physics system is capable of handling on a "normal" basis, with out-of-the-box parameters.

 
Last edited:

DavidGzz

Gold Member
On a side note I'm still so pissed off at Microsoft for bringing out the weak Series S, because it's going to be the limiting factor for the entire 9th generation of video game consoles.

1. Don't you want games to always be 60fps?
2. The S will just play those games at 30fps.
3. MS and now Sony have a team making games for PC, do you really think PCs will have to be as strong or stronger than an S to play them? No, you'll be able to play them on much weaker PCs, you'll just have to have reduced graphical settings....which they can also do to put a game on the Series S.

Shit's a non-issue. Hell most any game will also play on the Steam Deck which is a good bit weaker than the S.
 

svbarnard

Member
Here's another physics demo that was uploaded in 2009, there is not a single video game that has realistic destruction as this 13-year-old demo, not a single game has destruction that looks as convincing as this. But I do think that the PS5 and the XSX will finally be the console generation where we get destruction that looks as good as this 13 year old demo. Seriously take a look.
 

Miles708

Member
Here's another physics demo that was uploaded in 2009, there is not a single video game that has realistic destruction as this 13-year-old demo, not a single game has destruction that looks as convincing as this. But I do think that the PS5 and the XSX will finally be the console generation where we get destruction that looks as good as this 13 year old demo. Seriously take a look.
As people already said in this thread, this is not gonna happen because it would made asset creation take 10 times the time. Games already need 5+ years to make.

Maybe for indie or smaller scope games, sure.
 

svbarnard

Member
As people already said in this thread, this is not gonna happen because it would made asset creation take 10 times the time. Games already need 5+ years to make.

Maybe for indie or smaller scope games, sure.
No you're wrong. The reason we haven't seen destruction like this in video games is because the PS4 and Xbox One had the weak ass Jaguar CPU. But now here we are in the 9th generation of video game consoles and this time around the PS5 and Xbox Series Sex have beefy gaming CPUs that will finally be able to display realistic destruction like this 13-year-old physics demo. PCs have had the power to display realistic destruction for years but the PS4 and Xbox One were holding everything back. Those fucking Jaguar CPUs were holding back video games from having realistic destruction.
 

Miles708

Member
No you're wrong. The reason we haven't seen destruction like this in video games is because the PS4 and Xbox One had the weak ass Jaguar CPU. But now here we are in the 9th generation of video game consoles and this time around the PS5 and Xbox Series Sex have beefy gaming CPUs that will finally be able to display realistic destruction like this 13-year-old physics demo. PCs have had the power to display realistic destruction for years but the PS4 and Xbox One were holding everything back. Those fucking Jaguar CPUs were holding back video games from having realistic destruction.

Totally destructible environments were already possible in the Xbox 360 era (Red Faction Guerrilla) and even on the original Xbox era (Otogi and Otogi 2).
It's not, or at least not only, a technology limitation.

Even if you're a game dev, your time on this earth is limited you know.
 
Last edited:

svbarnard

Member
Totally destructible environments were already possible in the Xbox 360 era (Red Faction Guerrilla) and even on the original Xbox era (Otogi and Otogi 2).
It's not, or at least not only, a technology limitation.

Even if you're a game dev, your time on this earth is limited you know.
I'M TALKING ABOUT REALISTIC DESTRUCTION SUCH AS THIS 13-YEAR-OLD PHYSICS DEMO
I I'M TALKING ABOUT DESTRUCTION THAT LOOKS REALISTIC!!!! Such as that 13-year-old physics demo. Again I must emphasize I am talking about realistic looking destruction, destruction that looks real. I believe the PS5 and XSX will finally have the power to display realistic destruction (destruction that looks real).
 

Kataploom

Member
As people already said in this thread, this is not gonna happen because it would made asset creation take 10 times the time. Games already need 5+ years to make.

Maybe for indie or smaller scope games, sure.
Or they could make a dynamic fracture simulation system... Which I don't see widely used on games honestly, since then the objects count would be multiplied per vertexes on scene considering now each vertex acount for simulation and not just every mesh/composed object (group of meshes, and whatnot), IDK if would works like this but just quickly thinking about it, seems like so...
 

Miles708

Member
I'M TALKING ABOUT REALISTIC DESTRUCTION SUCH AS THIS 13-YEAR-OLD PHYSICS DEMO
I I'M TALKING ABOUT DESTRUCTION THAT LOOKS REALISTIC!!!! Such as that 13-year-old physics demo. Again I must emphasize I am talking about realistic looking destruction, destruction that looks real. I believe the PS5 and XSX will finally have the power to display realistic destruction (destruction that looks real).

This is from the remastered, but the Xbox 360 version was pretty much the same if remember correctly

That already looks pretty realistic to me, not at the level of your video, but that's a tech demo and this is a complete actual game on consumer hardware from 2005.

Going further back, albeit with more limitations, otogi 2 had total destruction during combat
I can't timestamp on mobile for some reason, but I'm referring from 25:05 onwards.

That's on hardware made in 2001.


Point is, you should at least take in consideration the possibility that this kind of destruction is possible already and it's not done because devs do not consider it a priority.
 
Last edited:

II_JumPeR_I

Member
We sadly have gone backwards when it comes to physics, interaction with the environment and destruction.
Euphoria was really used well in the Force Unleashed games
 

svbarnard

Member
This is from the remastered, but the Xbox 360 version was pretty much the same if remember correctly

That already looks pretty realistic to me, not at the level of your video, but that's a tech demo and this is a complete actual game on consumer hardware from 2005.

Going further back, albeit with more limitations, otogi 2 had total destruction during combat
I can't timestamp on mobile for some reason, but I'm referring from 25:05 onwards.

That's on hardware made in 2001.


Point is, you should at least take in consideration the possibility that this kind of destruction is possible already and it's not done because devs do not consider it a priority.
Again I'm talking about destruction that looks real. That physics demo from 2009 looks pretty dang real when you look at it.
 

svbarnard

Member
Yeah same here. I feel like devs have not even fully tapped the potential of the PS4 before moving on. This is going to be a trend going forward.
Yeah you're totally wrong. The PS4 and the Xbox One have been fully tapped at this point, aren't you aware of the weak Jaguar CPU in the PS4?
 

Miles708

Member
Again I'm talking about destruction that looks real. That physics demo from 2009 looks pretty dang real when you look at it.
Even if we want to agree, for the sake of discussion, about the fact that now we finally have the technology to achieve that, it would not solve the problem that you're making each game at least 5x more complex, for asset creation, gameplay implementation, bug fixing, ai reactions and so on.

I'm not sure devs are all on the edge of their seats waiting to implement destruction in every new game. Destructibility brings a mountain of work and problems and the actual physics is only one part of those.
 
Last edited:

CamHostage

Member
Here's another physics demo that was uploaded in 2009, there is not a single video game that has realistic destruction as this 13-year-old demo, not a single game has destruction that looks as convincing as this. But I do think that the PS5 and the XSX will finally be the console generation where we get destruction that looks as good as this 13 year old demo. Seriously take a look.

So, from the content decryption:

A masterful mesh of Softbodies fracturing wood, and all being obliterated by a crash. This is not Real Time.

Pixelux's DMM was made for both cinema and interactive use; the game version was massively paired back, as you can see in the actual interactive demos from the studio and LucasArts (and even those were demo environments, not truly in-game scenarios; also, by the time LucasGames actually started putting the technology to use in an actual game, they had a heck of a time getting it to work.) One of the reasons why DMM still exists and is used frequently in film is because it was good for that; one of the reasons it's not used in games anymore is because it didn't work out as well as hoped.

Fracturing and other physics simulation optimization for realtime use is always in development and making progress. However, when it comes to game performance, developers will still typically choose performance and reliability over "realtime" use unless absolutely called on otherwise or unless performance remains high, so if they can 'fake it' with a smartly produced Alembic baked-animation or set a fracture map that most closely looks like splintered wood (rather than conceivably breaking the object apart on actual mathematics of physical force/angle and bonded material strength... which is a lot to think about,) then that's the game they'll make. There aren't a lot of games where splintering wood is key to the game (and many places where that is applicable, other techniques work out fairly okay.)
 
Last edited:
Top Bottom