Rumble Roses developer on PS3 & X360

I will pick on a couple of points.

blakjedi said:
...Sony fans with no idea what the RSX is loock at cutscene video and say "see that clearly proves PS3 will be better, and its running on a single 7800GTX which is weaker than the RSX." Those same people will then turn around and say "we have no idea what is in RSX but it will be better than Xenos."

...You cant use a movie which has a different budget and constraints technically to compare to gameplay.

...Why is this so hard for people to grasp then just as improtantly let alone?

Has it been that hard for you to find equivalent cutscenes from X360 to compare?
 
blakjedi said:
What is really crazy about these constant bickerings and back and forths is that nearly all comparison arguments on these boards lately have false premises.

Sony fans with no idea what the RSX is loock at cutscene video and say "see that clearly proves PS3 will be better, and its running on a single 7800GTX which is weaker than the RSX." Those same people will then turn around and say "we have no idea what is in RSX but it will be better than Xenos."

...and was it really "Sony fans" at large or maybe one person you happen to have a heated argument with one day? That's a big difference of taking one person's views and then generalizing it to be the view of any and all people who aren't reflexively critical of every Sony news event. These kinds of generalizations serve no good to anyone.
 
Sonic took care of him. Didn't you notice how quiet it became?

That said, Rumble Rose isn't that bad of a game... the localization make me wanna bleed from my ears, but that's the Konami standard, so just import it... the engine has it's issues, but it's a decent wrestling engine -- yukes made it after all. Not surprisingly, it reminds me a lot of DOA, pretty girls average engine, but personally I think RR has more style. Again, somewhat of a Konami trait, IMO. Anyway, not the greatest not the worst, but there's fun to be had, and the different personas are a neat touch I think.
 
Sonic took care of him. Didn't you notice how quiet it became?

OMG I just mentioned that in another thread. I noticed, but I didn't notice. Whew. Glad to hear that news too because interviews should be coming out for the next couple of weeks from Japan.
 
mckmas8808 said:
Me and a couple of other people are disappointed in you Jawed. Tell me where your hidden information is at about the RSX. We would like to see it.
No information about the architecture of RSX is hidden. It's G70 with half the fill-rate and PCI Express replaced with Flex IO, running on 90nm.

NVidia have made the direct comparison themselves, saying "they're the same", only RSX is faster. Only the fanbois are still holding out for the "missing piece".

Jawed
 
What a noisey and pointless discussion. Embarassingly I feel it's somewhat my fault, having asked one person their reasons for feeling both machines at peak would perform similarly. Still, now I'm awake...
dukmahsik said:
why is xenon a shortfalling? im trying to understand all this.
In 'shortfalling' I meant XeCPU falling short in peak attainable performance to that of Cell, not that the XeCPU has a particular shortfalling in design.

Now if everyone takes a deep breath and calms down, I'll ask this question. It's not a loaded question and I'm not trying to prove anything one way or another; just asking people's opinions.

The consensus is Cell>XeCPU, Xenos>RSX. I think we can generally agree on that. The question is, which is more important for next-gen?

My opinion is the CPU. Looking at the visuals we're already getting, they are pretty amazing. We have loads of enemies all lit and shaded and whatnot. It looks good. I can't see what difference more graphics power would make, other than some IQ improvements. There won't be more enemies on screen than we already have (screenfull amounts) and higher polycount models then what's currently being seen won't make much difference in a moving game. If you were to take GOW as it is now and increase polycounts and everything by 3x, how much difference will you actually see? The only still graphics area that'll show in screenshots I can think of needing more oomph is AA and effects like flames which are traditionally weak. Actually Flames are more an animation thing.

What I feel does need improving is moving graphics. It'd be much more realistic to lose the mocap animations and replace them with physics based IK skeleton animations. It'd be much better to have faces modelled and animated on a muscle based system rather than using keyframed models. It'd be great to have smoke that follows the motion of air around objects passing through it. To me, the area that really needs improving is the area of animated, moving images, and this is the domain of the CPU. I would rather have MGS quality visuals and skeleton based IK anims and all these other new methods then have 3x the visual quality of MGS (though I don't know what that would look like!) and have limited animations and feet that don't match steps and the same animations techniques used for years. In the area of vertex and pixel shaders improving on RSX would I feel have less discernable impact then working on the animation systems instead.

What do the rest of yuo think on this matter, and why? If you feel more graphics rendering power is needed, please explain.
 
After reading through this entire thread (actually more skipping through it), I really think the console-discussion forums should incorporate a method of strictly reducing the amount of posting privileges by members with less than 200 posts. Even better, make them have a certain reputation level to be allowed to post more than one reply per topic, make a new thread etc. This is becoming far too annoying - even if you ignore them, you still can't avoid new members coming into the thread and responding to the trolls. :devilish:
 
Phil said:
After reading through this entire thread (actually more skipping through it), I really think the console-discussion forums should incorporate a method of strictly reducing the amount of posting privileges by members with less than 200 posts. Even better, make them have a certain reputation level to be allowed to post more than one reply per topic, make a new thread etc. This is becoming far too annoying - even if you ignore them, you still can't avoid new members coming into the thread and responding to the trolls. :devilish:

Too late now, that onetimeposter has got more than 200th post.
But seriously i think this guy is deadmeat junior.
 
shifty

It'd be much more realistic to lose the mocap animations and replace them with physics based IK skeleton animations.
I think the problem is you are assuming this is better than what we have, but realistically, you don't know if this soloution is even a viable one for game. What makes you certain physical based animation woudl look more natural than hand animation or mo cap of a real human being?

It'd be much better to have faces modelled and animated on a muscle based system rather than using keyframed models.
Again, this is an assumption based on what data? btw, most games don't use a morph targets, they use bones to animate a face. Still the best method is morph targets. It's not viable to set up a muscles system underneath the skin in todays games. Not when you could achive a similar effect more cheaply with morph targets.

It'd be great to have smoke that follows the motion of air around objects passing through it.
This already happens in some of the games today. In fact there was demos released by epic of this occuring in unreal 2.5 (aka wepon fire causing smoke to seperate. There's other games that have this effect as well on current generation PC and game consoles.

i find it really odd that many peopel on the net are saying the same thing, they want better things without knowing if they are even pluasible or infact better then the tech more widely used in games. even worse they somehow think that PS3 is going to allow this. Is this really based on comments from KK and things he said in interviews? because if it is, I think everyone is falling for the same hype/shit they did last gen.
 
Last edited by a moderator:
Also anothe rhting I noticed is tha tmany people are assuming there's a whole lot of phsycis occuring in the MGS4 video. Imo, everything was scripted. The pieces of the wall exploding was scripted. the dust and particles were scripted, (possibly withthe exception of blowing dust) but that's something very easy to simulate. Anyway it may have been scripted as well.

I see a lot of people saying htings like the part where the wall is shot and the chucks of dereis go flying tha tthey thought it was awesoem phsysics. Ifs it just me or are people forgetting this is a game engine and everything in the cutscene that are in engine are 98% of the time scripted?
 
Qroach said:
I think the problem is you are assuming this is better than what we have, but realistically, you don't know if this soloution is even a viable one for game. What makes you certain physical based animation woudl look more natural than hand animation or mo cap of a real human being?
Because I've seen it in action, in the NaturalMotion demos and in offline renders, plus it's obviously a technique with more potential. Consider a model with an animation to walk up stairs. You could map the motion exactly to a stairwell but then what when you want to step over a piece of rubble slightly taller than the steps? Or walk over a body and the motion's stride isn't long enough? Or while running hit a falling brick that should cause your character to trip, and then get up again? Motion capture is limited to a few set-pieces and interpolating them. There's no way it can achieve the realism of a fully simulated body.

...most games don't use a morph targets, they use bones to animate a face. Still the best method is morph targets...This already happens in some of the games today. In fact there was demos released by epic of this occuring in unreal 2.5 (aka wepon fire causing smoke to seperate. There's other games that have this effect as well on current generation PC and game consoles...they want better things without knowing if they are even pluasible or infact better then the tech more widely used in games.
Okay, ignore my specifically named examples which were just for illustration purposes. The idea is more CPU power = more animation options, potentially using new tehcniques not even though of before now, or using more techniques than currently capable now. Do you disagree with this? Do you think better animations are less important than better than say G70 class graphics?

From the sounds of it you seem to think these high powered CPU's aren't going to achieve much progress over what we already have. Games will use the same techniques as running on a PC now and there won't be any major improvements. Where is that number-crunching power going to be used then? :???:

Imo, everything was scripted. The pieces of the wall exploding was scripted. the dust and particles were scripted, (possibly withthe exception of blowing dust) but that's something very easy to simulate. Anyway it may have been scripted as well
But if we want these things to be physics based in game, rather than scripted, we'll need more CPU power to accomplish that, no? Not looking at current games but potential of the systems, when fully maxxed out, do you not believe there is more capacity for these effects and animation techniques?
 
I agree, we should see more than "canned" animation in the next gen. Physics-based modelling/animation should be the goal of next-gen (and no, please, no more Havok style rag dolls, which look like garbage to me).

Jawed
 
i dont know, there can never be enough graphics power and there can never be enough cpu processing power... for each generation both graphics and cpu processing power becomes more and more important.

in the end, IF cell > xenon and xenos > rsx, im sure that xenon and rsx will be more than good ENOUGH to have great physics, ai, and graphics. if cell > xenon it'll be by a tad and if xenos > rsx it'll be by a tad.

im sure both will come within 90% of each other.
 
dukmahsik said:
i dont know, there can never be enough graphics power and there can never be enough cpu processing power... for each generation both graphics and cpu processing power becomes more and more important.
What will more graphics power add though? Look at the trailers we're seeing. MGS and Kameo and whatever. Imagine the GPU's are 2x as powerful as those producing these graphics (on first gen titles). What would you actually SEE that was different? It's the old story of diminshing returns. A ball rendered in 16 vertices isn't gonna look like a ball. A ball with 100 vertices looks better, 1000 very smooth, 1,000,000 looks no better but at considerably extra rendering cost. Though I'm NOT saying we've peaked at graphics potential (not by a long chalk) I can't see where a doubling of polys and shaders is going to produce much of a noticeable difference. Obviously you think otherwise so you must be seeing areas that can be improved. What are they?
 
dukmahsik said:
i dont know, there can never be enough graphics power and there can never be enough cpu processing power... for each generation both graphics and cpu processing power becomes more and more important.

in the end, IF cell > xenon and xenos > rsx, im sure that xenon and rsx will be more than good ENOUGH to have great physics, ai, and graphics. if cell > xenon it'll be by a tad and if xenos > rsx it'll be by a tad.

im sure both will come within 90% of each other.

Huh?! Is that even english?

DeanoC said they won't be within 90% of each other.
 
Shifty Geezer said:
What will more graphics power add though? Look at the trailers we're seeing. MGS and Kameo and whatever. Imagine the GPU's are 2x as powerful as those producing these graphics (on first gen titles). What would you actually SEE that was different? It's the old story of diminshing returns. A ball rendered in 16 vertices isn't gonna look like a ball. A ball with 100 vertices looks better, 1000 very smooth, 1,000,000 looks no better but at considerably extra rendering cost. Though I'm NOT saying we've peaked at graphics potential (not by a long chalk) I can't see where a doubling of polys and shaders is going to produce much of a noticeable difference. Obviously you think otherwise so you must be seeing areas that can be improved. What are they?

im looking at the real games shown so far, i still see flat sided models on things such as wheel fenders, spectators, etc. the whole point of graphics is to push realism. you can't tell me you will be seriously happy next gen graphics 5 years from now? look at pixar movies, they are somewhat close to realistic graphics but not quite there. and furthering graphics doesn't have to be doubling the polygon processing power or double the shading power. with each new gen, there are better and different ways of generating a ball. who says we'll have to use polygons 2 gens from now? im not directing these questions at you directly per se but I wont be happy until graphics can produce life quality realism.
 
"GI: What do you think you could pull off on one system that you could not on another?

Uchida: I think there are some things you can do with the PS3 than you cannot with the 360. The processing speed is much quicker. The rendering variation that the PS3 offers is wider. But the online of the 360 is amazing and is probably better than the PS3."

I wonder why he(and others) don't talk about the Xenos being more powerfull compared to the stock G70 in the current PS3 devkit. You would say it's more powerfull(on paper) but then again I haven't seen any 360 game that given me proof of this yet.
 
Back
Top