Support NeoGAF

[Meccanical]

That model wasn't even in game as far as you know, what's your point?

 

[A More Normal Bird]

I used Aegies' "66% more efficient than 360's" statement as an attempt to calculate that difference out. While I'm aware that a solid answer may not be possible, is that number that off-the-mark when it comes to comparing 360's shader architecture to the GCN architecture?

We don't know how close to its 240GFLOPS rating Xenos got during real world operation, so we don't know what 66% better is (nor do we know much about the 66% improvement either). Massively simplified and hypothetical example: Xenos operates at 40% efficiency (of 240GFLOPS) = 96GFLOPS. Latte operates at 55% efficiency (of 176GFLOPS) = 96.8GFLOPS. XB1 operates at 66.4% efficiency (66%more than 40%) = 810GFLOPS (then apparently we take only 90% of that, so 729GFLOPS).

All this gets us is a bunch of potential estimates of a hypothetical performance rating. Even if you got the right efficiency figures, they would be context sensitive and wouldn't factor in other hardware and software differences. I don't think it's worth the bother.

That said, if you wanted to compare the relative worth of FLOPS ratings, here's what I would do. Go to Anandtech's GPU comparison site, select an HD4xxx or HD5xxx card, compare it to a 7 series part that has everything else roughly equal and then compare their overall performance to their FLOPS rating.

 

[Ryoku]

My whole question hinges on the claim the model is ingame.

Even so, we don't know if the game will be doing more work than "99% of other games out there". I don't think anyone is claiming such a thing. How can you when the only information we have is the polygon count of one model?

Wasn't it stated that the Kasumi model was 120k polygons or something? Either way, I'm still skeptical that the Bayonetta model in-game will contain 130k polygons.

 

[Mihael Mello Keehl]

Even if the model is in game so what? There will be other games to nag on about I bet.

 

[JordanN]

Even so, we don't know if the game will be doing more work than "99% of other games out there". I don't think anyone is claiming such a thing. How can you when the only information we have is the polygon count of one model?

Then you weren't around for when this actually happened. I've said everything you just said a few pages back.

Wasn't it stated that the Kasumi model was 120k polygons or something?

I said 99% of video games. I'm already aware of one.

 

[krizzx]

Even if the model is in game so what? There will be other games to nag on about I bet.

The entire purpose of me even bringing up the model has been lost. The fact that it was 130k alone is irrelevant out of context, but the topic has been shifted to that.

 

[Mihael Mello Keehl]

The entire purpose of me even bringing up the model has been lost. The fact that it was 130k alone is irrelevant out of context, but the topic has been shifted to that.

Exactly thats why I said that. Something NFSMU that show clear visual advantages dont really matter because resident evil revelations.

 

[Ryoku]

Then you weren't around for when this actually happened. I've said everything you just said a few pages back.

What is your argument, exactly?

I've been keeping up with this thread since it started. I just recently slowed down my posting since... well, it can get pretty crazy and reckless in here.

 

[Easy_D]

Wii U = Dreamcast

Xbox One = Xbox

PS4 = Wii

I thought the 3DS was Dreamcast and Vita Xbox.

 

[JordanN]

What is your argument, exactly?

This,

I'm more interested in how a game is doing more work than 99% of other games out there while still being considered "impressive" on substantially weaker hardware unless again, Bayo2 operates within certain perimeters (i.e bad or sectioned off environments) or simply, the model was never a game to begin with.

My whole question hinges on the claim the model is ingame.

 

[PdotMichael]

Why is it that crazy to think that the 140k Bayonetta model is a LOD0 model for cut-scenes or it's the base for normal/displacement mapping etc and the final in-game model will have way less polygons?

Or you guys can just wait until the first real in-game scenes.

 

[Meccanical]

Why is it that crazy to think that the 140k Bayonetta model is a LOD0 model for cut-scenes or it's the base for normal/displacement mapping etc and the final in-game model will have way less polygons?

Or you guys can just wait until the first real in-game scenes.

Or we can keep arguing!

 

[M°°nblade]

Strawman argument. I made no such claims in the way that you state them.

Not in the same way but they do boil to the same.

Because according to you, anything that indicates equivalence or inferiority of Wii U hardware (you mentioned performance, stutter and missing graphical features) says absolutely nothing about 'power'.
You selectively narrow the measurements that can be used to measure the 'power' of consoles to 'how the game looks' and you do that because it fits with your opinion rather than because they are invalid methods of measuring power.

On the other side, the evidence you propose for the Wii U superiority you believe in, lacks any critical view. It's based on a vague poly count number from a dev trailer and an interpretation from an article rather than objective in-game measurements. That makes the stark contrast. Basically you subjectify the whole discussion.

 

[OG_Original Gamer]

We're talking two different sets of hardware though where the far weaker one is supposedly pushing 3x the detail than what the more powerful one is doing. They [KillZone] should be able to have both higher poly characters and a detailed environment in theory. Unless there's something about Bayo2 that I'm missing (environments and characters are super weak? But others said it was also "impressive").

It may be true, but the environment, amount of rendered NPC'S on screen(they're polygon budget), physics. I don't think Japanese devs use normal mapping, which could explain the higher polygon numbers.

In some cases, the entire model will not be rendered, depending on position of the camera.

 

[JordanN]

Or we can keep arguing!

Can someone please read my post? I feel like a world of information was just ignored (like you think I don't know about the above?).

 

[Ryoku]

 

[JordanN]

I'm dying here too man (even more so than you). Hopefully a mod reading this understands I'm trying to make legit posts and not derail.

It may be true, but the environment, amount of rendered NPC'S on screen(they're polygon budget), physics. I don't think Japanese devs use normal mapping, which could explain the higher polygon numbers.

That would be odd. Platinum had a blog for the original Bayonetta about making normals. So did Squenix.

 

[Ryoku]

They [KillZone] should be able to have both higher poly characters and a detailed environment in theory.

This is an assumption on your part, nothing more. It's completely possible that the environment, post-processing, resolution, texture resolution, tessellation, lighting engine, wind engine, number of NPCs, physics, etc. are more resource-intensive together that effectively limit the total number of polygons a character model has.

 

[JordanN]

This is an assumption on your part, nothing more. It's completely possible that the environment, post-processing, resolution, texture resolution, tessellation, lighting engine, wind engine, number of NPCs, physics, etc. are more resource-intensive together that effectively limit the total number of polygons a character model has.

But we know Wii U isn't on par with PS4. If Wii U can handle 130k character in addition to having more detail (overall) compared to PS3/360 at its current power level, than PS4/XBO should have no problem easily surpassing that.

Now, how much post processing, # of NPC's, textures etc eats into this? I don't know. But this is why I called it theory (it should still be able to handle some of those things).

 

[OG_Original Gamer]

I'm dying here too man (even more so than you). Hopefully a mod reading this understands I'm trying to make legit posts and not derail.



That would be odd. Platinum had a blog for the original Bayonetta about making normals. So did Squenix.

I stand corrected. Is the series known for having large detailed environments, high on screen enemies. YouTube search initiated.

 

[pottuvoi]

So far one game (Killzone Shadow Fall) is managing 40k characters. That's basically One and a half Nathan Drakes per character.

Drake was around ~37k polygons and Chloe ~45k polygons in U2, so Killzone has quite reasonable polycount.

Byo2 polycount is certainly plausible as well, especially if there is LoDs for guns etc.

Rebel Strike on the Gamecube pushed 20 million polygon at 60 FPS with dynamics shadows, advanced lighting, bump mapping, 100s of individual A.I. enemies on screen at once, destructible environments.

You do know that when they talk about 20 million polygons they are talking about polygons per second, this makes any talk about fps meaningless.
Simply put framerate doesn't change the polygon throughput.

 

[Ryoku]

But we know Wii U isn't on par with PS4. If Wii U can handle 130k character in addition to having more detail (overall) compared to PS3/360 at its current power level, than PS4/XBO should have no problem easily surpassing that.

Now, how much post processing, # of NPC's, textures etc eats into this? I don't know. But this is why I called it theory (it should still be able to handle some of those things).

Sure, PS4/XBO should have no problem easily surpassing that assuming they left out all the other intensive features they use. I think you're underestimating how intensive some of these features are that are present in KZ:SF.

It isn't a viable theory if there is no foundation. This is all based on assumptions. KZ:SF is doing things that may be really intensive. If we remove these things, then sure, it opens up different possibilities, such as higher poly-count models.

 

[lherre]

This thread is hurting my head..lol

Yep ...

I'm astonished about some claims ... multipliers ... etc

 

[JordanN]

Drake was around ~37k polygons and Chloe ~45k polygons in U2, so Killzone has quite reasonable polycount.

Byo2 polycount is certainly plausible as well, especially if there is LoDs for guns etc.

I actually went back and forth on this. Turns out I misinterpreted the source (read it as "30,000" but not the "over" part).

 

[krizzx]

Drake was around ~37k polygons and Chloe ~45k polygons in U2, so Killzone has quite reasonable polycount.

You do know that when they talk about 20 million polygons they are talking about polygons per second, this makes any talk about fps meaningless.
Simply put framerate doesn't change the polygon throughput.

Where did you here that? The polygon throughput is directly proportionate to the frame rate.

The count you can achieve at 120 FPS is half what you achieve 60 FPS and is half what you achieve at 30. FPS means frame per second, so that means 20 million polygons were drawn 60 times every second. Everytime the frame rate doubles, the tax on the system doubles.

If the polygon count was frame rate independent, then pretty much every game would be using its consoles max.

 

[pottuvoi]

Where did you here that? The polgyon count is directly proportionate to the frame rate.

The count you can achieve at 120 FPS is half what you achieve 60 FPS and is half what you achieve at 30. FPS means frame per second, so that means 20 million polygons were drawn 60 times every second.

When talking about polygons per second, the per second part makes the frame count in a FPS meaningless.
Also if you take 20Million polygons per frame * 60 and you get 1200million polygons per second, which is just silly number for current or previous generation.

 

[lightchris]

Where did you here that? The polgyon count is directly proportionate to the frame rate.

True, but you misunderstood the point. For example, 20 million polygons per second at 60 fps mean 333k polygons per frame. If another game draws 20 million polygons/s at 30 fps, that means 666k polys/frame. The overall polygon throughput remains the same.

 

[lherre]

So a technical demo is a game now? wow !!!

 

[krizzx]

When talking about polygons per second, the per second part makes the frame count in a FPS meaningless.
Also if you take 20Million polygons per frame * 60 and you get 1200million polygons per second, which is just silly number for current or previous generation.

I don't know how you came about that math. Polygons throughput have been calculated relative to frame rate as long as I can remember. The math does not work that way. Its not 1200million just as a 2x4 wooden board is 8.

Its 20million at 60FPS. This is twice as taxing on the hardware as it would be at 30 FPS.

 

[JordanN]

So a technical demo is a game now? wow !!!

You referring to Killzone?

 

[lightchris]

Polygons throughput have been calculated relative to frame rate as long as I can remember.

Only if you talk about polygons per frame, not per second.

 

[lherre]

You referring to Killzone?

No.

 

[Mihael Mello Keehl]

No.

lol

 

[Hermii]

No.

I just wonder something, serious question. Do you have the definite answers ?

 

[krizzx]

Only if you talk about polygons per frame, not per second.

What? They are relative to one another.

FPS = Frames per second
It is the same in this instance.

Factor 5 indicated they could get 20 million polygons/second per second with all effects.

 

[lherre]

Killzone SF is a game, the garden demo is only that, a tech demo. You can't compare both.

 

[JordanN]

lol

Just playing it safe.

"Court's kinda fun when it's not my ass on the line! Nachos?"

 

[krizzx]

Killzone SF is a game, the garden demo is only that, a tech demo. You can't compare both.

It was an interactive tech demo that ran live on the hardware. It was the same as any other game on the hardware except it wasn't sold.

 

[ArchangelWest]

It was an interactive tech demo that ran live on the hardware. It was the same as any other game on the hardware except it wasn't sold.

Not to mention that pretty much the only parts worth bragging about were on rails/extremely limited interaction.

*Meant that about KZ SF, but the same can be said for the Garden demo since it had no game play though. The only actual game play portions of KZ SF were confined to the rooftop where none of the awe and splendor of the city were visible.

 

[M°°nblade]

It was an interactive tech demo that ran live on the hardware. It was the same as any other game on the hardware except it wasn't sold.

And except it wasn't a playable entity.

 

[krizzx]

Not to mention that pretty much the only parts worth bragging about were on rails/extremely limited interaction.

The only times that game being on rails made a different in performance were when the games were only drawing what was in front of you and nothing else. That is not the case here, so it does not discount the performance.

In the Wii U Garden demo, the person with the controller had full control of the camera. You could view the entirety of the environment. In this instance being on rails was simply a choice by design as the demon was meant to be guide to the Wii U's graphical capabilities. It wasn't done to cloak graphical inability.

And except it wasn't a playable entity.

Except it was a playable entity in the same way as the Zelda tech demo. The player had full radial control of the camera and various environmental interactions.

 

[M°°nblade]

Exept it was a playable entity in the same way as the Zelda tech demo. The player had full 360 degree controller of the camera and various environments environmental interactions.

So you say navigating a camera in a realtime demo with scripted animations is the same as playing a game then?

 

[krizzx]

So you say navigating a camera in a realtime demo is the same as playing a game then?

Strawman no.3.

I said what I wrote.

 

[M°°nblade]

Strawman no.3.

I don't think you understand what strawman means.

Do you deny saying 'It was the same as any other game on the hardware except it wasn't sold.' then?

 

[Goodlife]

Killzone SF is a game, the garden demo is only that, a tech demo. You can't compare both.

Was Killzone any more of a game then the Killzone 2 reveal for PS3 was?

 

[krizzx]

I don't think you understand what strawman means.

Do you deny saying 'It was the same as any other game on the hardware except it wasn't sold.' then?

A strawman is an argument a person creates by augmenting a statement made by another person then making an argument against there own augmented version as opposed to what the person actually said while implying it is the same statement. It is akin to a loaded question in which you ask a question that presumes something was stated when it was not.

When your statement begins with something along the line of "so you are basically saying" or what not then there is a good chance that you just made a strawman.


Now games being the same as as any other I did state.

There are plenty of games that of the same nature. Take Flower on the PS3 for example

 

[lherre]

Was Killzone any more of a game then the Killzone 2 reveal for PS3 was?

Killzone 2 was a CG ... killzone 4 trailer was basically gameplay recorded from the demo showed later.

 

[M°°nblade]

A strawman is an argument a person creates by augmenting a statement made by another person then making an argument against there own augmented version as opposed to what the person actually said.

Yet you literally did say that the bird tech demo was the same as any other game on the hardware except it wasn't sold.
So I just asked you again, apart from not being sold, navigating a camera is the same as playing a game?
That's not really a strawman is it? I'm just asking you to clarify.

Now that is something did state. There are plenty of games that of the same nature. Take Flower on the PS3 for example

But in Flower you actually control/move an object which interacts with the environment. Can you control or move the bird in Nintendo's tech demo?

 

[Mihael Mello Keehl]

Just playing it safe.

"Court's kinda fun when it's not my ass on the line! Nachos?"

ok..... Ill leave it at that.

 

[D-e-f-]

Yet you literally did say that the bird tech demo was the same as any other game on the hardware except it wasn't sold.
So I just asked you again, apart from not being sold, navigating a camera is the same as playing a game?
That's not really a strawman is it? I'm just asking you to clarify.


But in Flower you actually control/move an object which interacts with the environment. Can you control or move the bird in Nintendo's tech demo?

In the garden demo you had 360° camera control on top of the guided view on the TV.

edit: after checking some vids, it seems to have been 180° camera control, at least during the flight parts. not sure if there's a video of someone actually trying everything with the thing.

Don't expect much from Watch Dogs, because I've think it has been unofficially confirmed to just be a port of the 360 version and runs worse. I'd say look to Arkham Origins instead. The team who are making it are the ones who ported Arkham City to the Wii U, so they have more experience with making Batman on the Wii U than on the 360/PS3. Though, don't expect any multi-platform game to show significant difference from one version to another at this scale. Most devs aren't going to drop that kind of money on their own.

Just one more thing to note about this:

The Wii U version of Watch Dogs is being developed by Ubi Bucharest. They did the ZombiU multiplayer so they know their Wii U tech as well. And it was reportedly in development alongside the other next gen versions while the last gen versions hadn't even started development. Just a thing to consider. *I* hadn't heard anything else about the different versions since.