Merry Xmas Pottsey

[Pottsey]

“That ran very well actually considering it was recorded by some random with FRAPS or similar which causes a performance hit anyway. ”

It dropped down to about 1FSP and without FRAPS drops down to under 5fps. I don’t call that very well. I am still waiting for your evidence showing Crysis doing it without slow down.

As for the liquid forgot about the way it looks it’s a very old video. The video was to demonstrate how liquids could be used. Surly you can see how liquid flowing out a fuel tank like that could be useful in games? Clearly you would want better looking liquid then that but it’s not about the looks, but the way its acts. It’s the idea of the liquid flowing out the tank based on pressure that good.

No I am not saying you need a PPU x10 more powerful then we have now to do that. I was saying imagine if we all had a PPU x10 more powerful then today we can do things far in advance of the video or any current game.

You say its looks like crap well its x20 better then the CPU can do.

 

[mrk]

If Crysis had fully realised water in it then i think it would look far better - namely waves lapping on to the shore and waterfalls cascading down the rocks.

That sort of thing would be fantastic.

There are small waves that do lap, you won't find bigger waves because there's no gusty winds - it would be abnormal to have big waves and no wind!

There also are several waterfalls in Crysis.

 

[Pottsey]

“Pottsey, I've told you a few times now. I don't know how to use Sandbox2 but you only have to download the demo to prove me wrong. The Sandbox2editor is included with it.”
Why do I need to download the demo? I just posted a video proving you wrong. There is also a link back a few pages with a thread with loads of people reporting the same thing about performance and how the game is CPU limited when it comes to high end physics.

Your the one saying the game can do high end physics even after the videos and other people who tried it say it cannot. You need to post evedince, not me.

 

[Darkwave]

It dropped down to about 1FSP and without FRAPS drops down to under 5fps. I don’t call that very well. I am still waiting for your evidence showing Crysis doing it without slow down.

How many times do I have to tell you? I tried yesterday, I took a good 45min out of listening to you whinge about Crysis to try getting to grips with CryEngine2 Sandbox2 and I just couldn't get my head around it.

Like I said: All it takes for you to blow my argument right out of the water is to download the demo, which includes CryEngine2 Sandbox2. If you want something done right then do it yourself, because I took an honest whack at it.

As for the liquid forgot about the way it looks it’s a very old video. The video was to demonstrate how liquids could be used. Surly you can see how liquid flowing out a fuel tank like that could be useful in games? Clearly you would want better looking liquid then that but it’s not about the looks, but the way its acts. It’s the idea of the liquid flowing out the tank based on pressure that good.

Show me some Ageia-generated physics in action, in a game (a good one) and I will buy a PPU.

No I am not saying you need a PPU x10 more powerful then we have now to do that. I was saying imagine if we all had a PPU x10 more powerful then today we can do things far in advance of the video or any current game.

I agree with this. If the PPU was 10x more powerful then it would probably generate some seriously mind-blowing physics affects, the only problem there is it would need 10x as much developer support as well.

You say its looks like crap well its x20 better then the CPU can do.

That's because the majority of gamers are obviously happy with the way physics are progressing without the PPU. Crysis physics are pretty incredible, and blow me away way more than any Ageia-generated physics I've seen so far.

It's the same as graphics. Games like Quake IV and Doom 3 went over the top with bump mapping on their textures, I thought it looked hideous but technologically speaking it was a fantastic step forwards. Now replace those game titles with PPU titles, and "bump mapping" with "PPU physics", and you have your answer.

Why do I need to download the demo? I just posted a video proving you wrong. There is also a link back a few pages with a thread with loads of people reporting the same thing about performance and how the game is CPU limited when it comes to high end physics.

Your the one saying the game can do high end physics even after the videos and other people who tried it say it cannot. You need to post evedince, not me.

I never said anything about "high physics" though, in fact I didn't have a clue what it was and asked you to define it for me more than a few times. I also never stated, as a fact, that the CPU can do a gigantic nuke that easily, which is why I tried being your string-puppet and wasting the best part of an hour in CryEngine2 Sandbox2 trying to find out. If your video is anything to go by though then it looks like the CPU can not only do it, but do it whilst simultaneously recording and encoding the video in real time.

What I have been saying all along, had you been reading it, is that Crysis is purely GPU limited and has absolutely no problem producing any of the physics effects in the game on the CPU as demonstrated by the task manager graphs I provided that were recorded as I ran a loop of Crysis' own built-in CPU benchmark.

 

[Chronictank]

“you can argue semantics all you like but its the final effect that matters at the end of the day in which case warmonger is nothing special”
You do like dogging questions don’t you. I keep asking for evidence and you never post any. I am still waiting to hear how the PPU is a negligible diffrance UT3. You keep saying warmonger is nothing special and older game do the same physics. If the physics are nothing special then show me older games as good.

I failed to see this "hit in framerate" you saw in the videos i posted, i also showed you a second video where someone with a PPU tried it in UT and said there was little difference between the 2 in his summary. Just because you choose to ignore it doesnt mean i didnt post it

Seeing as you have a PPU you could test it yourself and proove me wrong, although i would have to question the validity of your results as you are an avid supported of the PPU

Warmonger is an online only game. Online only games are not about the best graphics possible they always have worse graphics then single player games. This is about physics not graphics. And Warmongesr physics are something special. Online only games will never have better graphics then a single player game of the same generation.

Excuses, Crysis also is a multiplayer game yet still looks far better than warhawk, as does COD4, and UT with or without physx which by everyone who has tested it say there is no real noticable difference

As title, running with it on or off makes no noticible difference to game play on all the normal levels.

http://forums.overclockers.co.uk/showthread.php?t=17808007

”Things like that make a big difference to actual gameplay, and you don't need a PPU for it.”
You do need a PPU if you want to increase the scale. Yes the physics are good in Crysis but they are all on a small scale. Imagine the same physics you love but instead of one Jeep it’s a large fleet of jeeps. Instead of a 5 metre blast wave you now have a 1 mile blast wave. Instead of 1 tree faling down you have a large section of trees falling down. All things current CPU's cannot do but a good PPU could.

You keep saying this, but there are no games using physx which come close to what Crysis achieved in terms of visual quality as a whole. I am beginning to think that is a dead herring or are really all game developers working on the technology not using it properly? i am getting pretty doubtful

“I expected textured bricks,”
The bricks are textured. Those are old run down buildings in a war zone. The bricks where meant to be dark. There are lighter colour bricks on the none warszone areas of the map. Some of the under ground bricks are more like silver tiles.

Yet they still look poor, you beggining to see the problem?
If Physx is THAT good, then post 1 game which will wow us visually
Warhawk is not one of them

 

[Richdog]

Guys I don't know why you're bothering with all this.. Pottseys sig tells you all you need to know...

 

[Pottsey]

“Crysis physics are pretty incredible, and blow me away way more than any Ageia-generated physics I've seen so far.”
That doesn’t make any sense as all the Crysis physics are in PPU games. Yet in the PPU games you and others say the very same physics even 2 years ago are not incredible. All the leave, tress, wind from explosions e.c.t is not incredible but when Crysis does the very same physics its incredible?




“I never said anything about "high physics" though, in fact I didn't have a clue what it was and asked you to define it for me more than a few times. I also never stated, as a fact, that the CPU can do a gigantic nuke that easily,”
You said “There is no slowdown in Crysis when you mow down half a forest, had you played it you would know.” I then showed you a video of mowing down half the forest and being CPU limited at 1fps.




“If your video is anything to go by though then it looks like the CPU can not only do it, but do it whilst simultaneously recording and encoding the video in real time.” & ”I failed to see this "hit in framerate" you saw in the videos i posted,”
The screen locked up and everything stopped moving, the gun on the person stopped wavering, the FPS clearly went down to 1 or less. It wasn’t for long 4 or 6 seconds.





“What I have been saying all along, had you been reading it, is that Crysis is purely GPU limited and has absolutely no problem producing any of the physics effects in the game on the CPU as demonstrated by the task manager graphs I provided that were recorded as I ran a loop of Crysis' own built-in CPU benchmark.“
Not this again. I proved to you it was CPU limited when lots of physics are going on, which is why there are not lots of physics in the default game. Being CPU limited is why the physics are low by default. Crysis is not purely GPU limited like in the video. They keep the physics low so it wouldn’t be CPU limited during normal gameplay. If they where not CPU limited they would have put far better physics in.






“Seeing as you have a PPU you could test it yourself and proove me wrong, although i would have to question the validity of your results as you are an avid supported of the PPU”
As I said before I have not found a way to run a fair test as you cannot record a timedemo. As I cannot run a fair test I know the video you posted is unfair. The video you posted wasn’t a benchmark it had no CPU fps number next to PPU FPS numbers. For all you know the PPU was getting 20% more FPS over the CPU. All you did was show a video without FPS numbers then go and say the PPU does nothing.






“Excuses, Crysis also is a multiplayer game yet still looks far better than warhawk, as does COD4, and UT with or without physx which by everyone who has tested it say there is no real noticable difference”
Crysis is not an online only multiplayer game. Surly you know the difference between an online only multiplayer game and a single player game with a small multiple component. There have been threads on this forum and others saying UT is gets a FPS boost with the PPU.





“http://forums.overclockers.co.uk/sho...php?t=17808007”
He posted no FPS score difference what if the PPU was 30fps faster? He never posted the CPU score what if it was 10x slower on the PPU map?






“You keep saying this, but there are no games using physx which come close to what Crysis achieved in terms of visual quality as a whole.”
As I am not talking about visual quality as a whole. With Crysis I wasn’t even talking about Ageia’s PPU most of the time, as said more then once. I was talking about how it is CPU limited when you increase the physics and that if we all had PPU’s the physics could be pushed forward far past what it has now. You’re the one who keeps brining up the visual quality as a whole. Go back and reread start of the thead. This started as I bought up Crysis is an example of how we need PPU’s with exmples of when Crysis is CPU limited. I said right near the start and more then once afterwards I was talking about PPU’s in general and not Ageia PPU with Crysis.





“Yet they still look poor, you beggining to see the problem?
If Physx is THAT good, then post 1 game which will wow us visually”
So UT3 isn’t visually a wow game?





“Yet they still look poor, you beggining to see the problem?”
No as they look better then CPU games as proven by the screenshots.

 

[Dave10]

“Yet they still look poor, you beggining to see the problem?
If Physx is THAT good, then post 1 game which will wow us visually”
So UT3 isn’t visually a wow game?

yes, but the clumsy physx levels are not

 

[ihatelag]

LOL my local computer exchange shop has 4 PhysX cards to buy 2nd hand...sort of speaks volumes about the physics card!

 

[Marc Fraser]

“If your video is anything to go by though then it looks like the CPU can not only do it, but do it whilst simultaneously recording and encoding the video in real time.” & ”I failed to see this "hit in framerate" you saw in the videos i posted,”
The screen locked up and everything stopped moving, the gun on the person stopped wavering, the FPS clearly went down to 1 or less. It wasn’t for long 4 or 6 seconds.

I don't know how anyone can say that video proves the CPU can handle high level physics effects. The frame rate drops to well below 1 FPS which cannot be blamed purely on FRAPS.

Have you all read the website I linked to earlier where the people who recorded the video are discussing the effect on the game. The person who recorded it says it has a massive hit on performance.

The problem with assessing the effect physics is having on the cpu is that the system will not put the same load on the CPU each frame. Take the video as an example. When the nuke initially goes off the framerate stalls for a couple of seconds and then stutters due to the high level of calculations that are taking place. As the explosion dies down and the the physical forces in the scene decrease, less objects are being effected and the framerate picks up.

 

[Chronictank]

“Yet they still look poor, you beggining to see the problem?
If Physx is THAT good, then post 1 game which will wow us visually”
So UT3 isn’t visually a wow game?

UT3 is, UT3 physics maps aren't

I am beginning to see how futile it is to try have a discussion on this, every time i post something you come up with an excuse that it isnt a fair test or shouldnt be used. So i'll put the onus on you
provide just one game using physx (as we are arguing the use of physx cards not ppu's in general), which looks visually better overall than its same gen counterparts

 

[Finguz]

Same old same old as usual. If the PPU takes off ( in one form or another ) I can guarantee all the naysayers will be first in the queue to buy it. As things stand at the moment though, it's an overpriced doorstop with no 'killer apps' to get it noticed. Seperate physics processing is an excellent idea and bound to show nice improvements in immersion and performance once people get to grips with it. Of course this may never happen with the current incarnation

 

[Darkwave]

“Crysis physics are pretty incredible, and blow me away way more than any Ageia-generated physics I've seen so far.”
That doesn’t make any sense as all the Crysis physics are in PPU games. Yet in the PPU games you and others say the very same physics even 2 years ago are not incredible. All the leave, tress, wind from explosions e.c.t is not incredible but when Crysis does the very same physics its incredible?

I've not seen it do any of that and you've not shown me any of it either. All I've seen so far is crappy tech demos, bits of rubbish looking debris in G.R.A.W. and some newspapers fluttering around in C.O.H/V.

“I never said anything about "high physics" though, in fact I didn't have a clue what it was and asked you to define it for me more than a few times. I also never stated, as a fact, that the CPU can do a gigantic nuke that easily,”
You said “There is no slowdown in Crysis when you mow down half a forest, had you played it you would know.” I then showed you a video of mowing down half the forest and being CPU limited at 1fps.

Show me the word "nuke" in the paragraph you just paraphrased me in.

And yes, you showed me a video that was recorded and encoded on the CPU simultaneously as the game was running. And yet you somehow accuse me simply running the Crysis benchmark being "unfair testing."

Like I said a million times now; Had you played the game, or even tried the demo (which is all you need to do to prove me wrong), you would know that mowing down a forest doesn't make the game slow down because the game is totally GPU limited.

“If your video is anything to go by though then it looks like the CPU can not only do it, but do it whilst simultaneously recording and encoding the video in real time.” & ”I failed to see this "hit in framerate" you saw in the videos i posted,”
The screen locked up and everything stopped moving, the gun on the person stopped wavering, the FPS clearly went down to 1 or less. It wasn’t for long 4 or 6 seconds.

Which isn't bad at all considering that Fraps or similar was recording and encoding that video in real time. You do know that this kills performance in general, right? Especially when a lot is going on.

“What I have been saying all along, had you been reading it, is that Crysis is purely GPU limited and has absolutely no problem producing any of the physics effects in the game on the CPU as demonstrated by the task manager graphs I provided that were recorded as I ran a loop of Crysis' own built-in CPU benchmark.“
Not this again. I proved to you it was CPU limited when lots of physics are going on, which is why there are not lots of physics in the default game. Being CPU limited is why the physics are low by default. Crysis is not purely GPU limited like in the video. They keep the physics low so it wouldn’t be CPU limited during normal gameplay. If they where not CPU limited they would have put far better physics in.

So you're saying Crysis is not CPU limited, but it is when it uses effects not in the game and with Fraps simultaneously recording and encoding? Doesn't this just prove my point that the game is not CPU limited?

If your next reply doesn't just repeat the same ones as this reply did, then I'll bother answering it. I don't like being pulled around in circles, partially because I am beginning to think your entire debate style is to simply go around in circles until your opponent can not be bothered anymore, but also since I already wasted nearly an hour trying to get to grips with Sandbox2 for you.

It's time for you to get your arse into Sandbox2 do the same if you want to get results.

Guys I don't know why you're bothering with all this.. Pottseys sig tells you all you need to know...

You're right actually. I don't know why I'm wasting my time and effort debating about Crysis with someone who's not even played it, and isn't willing to download the demo.

 

[Pottsey]

“So i'll put the onus on you
provide just one game using physx (as we are arguing the use of physx cards not ppu's in general), which looks visually better overall than its same gen counterparts”
As I said before Warmonger the physics are visually better then the physics in CPU games. Your Crysis screenshot was all blocky, the warmonger screenshot had round edges and sharp edges which look better.

PPU’s are not just about looking visually better. They are about offloading the physics from the CPU for a FPS boost. Take UT 3 it runs better with a PPU then without a PPU.

You seem to be arguing that the GPU can render better shadows, better textures, better face modules in game A over PPU game B. Because the GPU is doing a better job at none physics stuff in game A, then the PPU is not needed for physics. That’s how it sounds to me.



“every time i post something you come up with an excuse that it isnt a fair test or shouldnt be used.”
But it is unfair. You post a video without FPS numbers as evidence the PPU is worse then the CPU. You post a blocky physics Crysis screenshot as evidence its better then PPU games. Saying the PPU is all blocky even though it has round edges in the PPU screenshot making it better then yoru Crysis screenshot.

 

[Marc Fraser]

UT3 is, UT3 physics maps aren't

I am beginning to see how futile it is to try have a discussion on this, every time i post something you come up with an excuse that it isnt a fair test or shouldnt be used. So i'll put the onus on you
provide just one game using physx (as we are arguing the use of physx cards not ppu's in general), which looks visually better overall than its same gen counterparts

The Physx API does not produce the graphics. It is an API for calculating physics calculations that can run on either a CPU or PPU. If you want an example of a game that uses Physx that looks pretty good there is gears of War. The whole of UT3 uses Physx not just the PPU levels.

When you start using the PPU to add additional effects to the game whether it be a destructable terrain with additional debris of just some extra particle systems you need to keep in mind that you are also likely to be placing additional burden on the GPU by giving it more to render. It may therefore not be posible to keep as higher visual quality overall compared to if you didn't add all the additional effects. However if the additional effect greatly improve gameplay it is a choice the developer may decide to make. I know we are on the graphics card board but there is more to games than just graphics. A game can be just as immersive if you turn done the graphics one notch but have a much more realistic environment.

For example, judging by the video, I think the wind effects added to the island level in GRAW2 make for a more realistic looking environment.

 

[Chronictank]

The Physx API does not produce the graphics. It is an API for calculating physics calculations that can run on either a CPU or PPU. If you want an example of a game that uses Physx that looks pretty good there is gears of War. The whole of UT3 uses Physx not just the PPU levels.

When you start using the PPU to add additional effects to the game whether it be a destructable terrain with additional debris of just some extra particle systems you need to keep in mind that you are also likely to be placing additional burden on the GPU by giving it more to render. It may therefore not be posible to keep as higher visual quality overall compared to if you didn't add all the additional effects. However if the additional effect greatly improve gameplay it is a choice the developer may decide to make. I know we are on the graphics card board but there is more to games than just graphics. A game can be just as immersive if you turn done the graphics one notch but have a much more realistic environment.

For example, judging by the video, I think the wind effects added to the island level in GRAW2 make for a more realistic looking environment.

Physics is already at a level where it is perfectly acceptable for gameplay, tree's swaying in the wind do not affect gameplay in any manner so that point is mute.
So now its about graphical quality and immersiveness of the enviroment;
What good is swaying tree's if they looks horrible compared to the rest of the game? the answer, no good at all as it brings down the effect of the enviroment imo as proven in UT3. Amazing looking game, add a PPU and the extra effects just look poor.
The alternative is as you say, tone down the graphics and have a less pretty game like warhawk, which to be frank isnt much better than HL2 from that vid (abeit poor quality and beta so i am saying that with a pinch of salt)
I would rather have a pre-scripted/low physics tree that looks stunning than a high physics poor looking tree personally

And to clarify we are talking about Physx cards, not PPU's in general. I think a PPU is a great idea, but in practice today it is dead weight.
Ideally a ppu would be built onto a gfx card/cpu rather than as a seperate entity wasting a slot

 

[Gerard]

Christ almighty physics threads should be taboo on this forum, same bs every single time.

 

[Chronictank]

Christ almighty physics threads should be taboo on this forum, same bs every single time.

don't read it if it bothers you so much
It was pretty obvious from the first couple of posts where this was going

 

[Gerard]

don't read it if it bothers you so much
It was pretty obvious from the first couple of posts where this was going

Doesn't bother me its just the same thing every time, pottsey being the single person that likes the physx cards and everyone else telling him theyre crap. How many times is it possible to have the same argument?

 

[Chronictank]

Doesn't bother me its just the same thing every time, pottsey being the single person that likes the physx cards and everyone else telling him theyre crap. How many times is it possible to have the same argument?

passes the time