Are these first set of games for PS4 and XboxOne not really a good indication of what the powa they are capable of translating into visuals?
Yup. It'll be the same as it has been every gen since the beginning of time. We're still a long way off from hitting a plateau where adding more power and the development community as a whole being more experienced with a system doesn't result in better visuals.
Yeah, there will be a few games that will be a huge jump, like games that have not been rushed and are designed just for next gen tech and not being made with the 360 and ps3 in mind.
Devs always optimize and get better with hardware the more experience they have. I expect great things. I honestly think the games we'll see this fall will be far and away superior to the launch games. Just looking at Sunset Over drive and Destiny says that to me.
I'm sure it will increase, but not that huge a visual jump
How? Already, PS4 and XB1 exclusives (that are not held back by last gen consoles), are showing huge leaps over even later games in last generation. Remember, developers are still yet to write engines specifically for next gen platforms to make use of unified memory pools and GPGPU functions etc. This will result in stuff that the PS3 and even equivelent PC hardware will not be able to do.
Isn't this thread about visual jumps on the same system within a gen? I don't know what the graphics and capabilities of last gen platforms have to do with anything
I'm sure it will increase, but not that huge a visual jump
How? Already, PS4 and XB1 exclusives (that are not held back by last gen consoles), are showing huge leaps over even later games in last generation. Remember, developers are still yet to write engines specifically for next gen platforms to make use of unified memory pools and GPGPU functions etc. This will result in stuff that the PS3 and even equivelent PC hardware will not be able to do.
Isn't this thread about visual jumps on the same system within a gen? I don't know what the graphics and capabilities of last gen platforms have to do with anything
No, what i'm saying is that there will infact be a visual jump. And certianly a graphical asset quality jump. Like character models having 10 times the polygons as last gen etc. As well as better lighting and textures. Problem is, it will start to reach a diminishing returns point and it will hard for the laymen to see the difference as much, regardless of the developer putting those extra hours in making detail an order of magnitude better.
...okay, I have no idea what this has to do with what I said
...okay, I have no idea what this has to do with what I said
You said it will increase, but not by much. I said, it will actually increase a lot. But it will get harder and harder for the average person to see the difference. Even though it still has about 10x better graphics assets as the PS3 on paper. (Diminishing returns)
Which is what I said: "not that huge a visual jump"
I don't see much of a leap this gen, they started with old hardware out the gate, unlike last gen which started with top end. They finally got the RAM they needed, but threw out everything else
Hello. If you look at Oblivion and then you look at Skyrim on the same platform, it seems to be a big visual jump.
Are these first set of games for PS4 and XboxOne not really a good indication of what the powa they are capable of translating into visuals?
At the moment, all multiplats are cross gen ports, meaning that there the PS4 version is similar to the PC version in most cases, where the last gen consoles is running in low settings equivelent. However, already the PS4 exclusives such as Killzone SF, and Infamous SS have shown significant increases in their geometry over pervious gen titles of the same genre. We will see massive jumps.
This is an Uncharted 4 model leaked. They are said to have over 200,000 polygons. To think, Uncharted 3 had 30,000 polygons for character models. So now they have enough polygons to render each individual hair on his beard and head pretty much. And not just character models, but over all geometry of the levels will increase dramatically.
If the face alone has over 200,000 poly's the full bodies must be over a million, then add all the overall geometry of the levels! I shudder to think just what kind of performance a PS4 would push it out at, if one even could.
I'm not dissing the PS4, it's a decent amount of power for a low price, but first party developers with no interest in any other systems can't get their games to run at 1920 x 1080 and 60 fps.
According to developers the learning curve this gen, with its x86 architecture is not as steep as previous gen. Therefore its logical to assume that this gen won't see as huge leaps as previous gen did. Many of the ps3 launch games only ran on the PPE in cell and the RSX, leaving all 6 usable SPEs idle.
I think there'll be a considerable jump. Even exclusive games released up to this point where probably rushed to meet the consoles' launch windows.
Not only that, but there are many API optimizations bound to occur, adoption of multithreading and parallel processing as standard, and not only that, but I'm certain that, even with that considered, the hardware is still being way underutilized (heck, Crysis in 2007 looked as good as some current games, and it would run a lot better with a similar hardware as those consoles have). Even on the PC market... I think that a hardware that's supposed to be about 5x more powerful than a PS4 (in the case of a high-end rig) should be able to render gfx that are a lot, lot better, than what we're seeing today... But I'm not a developer, and not a programmer of any sort, so it's only an assumption.
The state of 3D game engines and 3D asset development at the start of the last gen is a completely different ball of wax compared to where it is now.
Back then you had brand new API's no one had worked with, completely new hardware no one had worked with, completely new way of asset development that no one was familiar with.
Fast forward to today, and everyone knows everythign there is to know about modenr PC hardware architecture, everyone knows how asset creation and 3D rendering works with modern API's.
That doesn't mean there aren't going to be improvements, however. But, they won't be anywhere near as drastic as last gen and they will also benefit PC - since most wil be about threading CPU workloads and utilizing GPU compute and all dat RAM.
Hello. If you look at Oblivion and then you look at Skyrim on the same platform, it seems to be a big visual jump.
Are these first set of games for PS4 and XboxOne not really a good indication of what the powa they are capable of translating into visuals?
Major point people tend to forget or ignore, Game engines were only single threaded and did not make use of the hardware that was available for the 360 or PS3 more then a year after their launches. Those consoles were spouting top end, proprietary and new hardware that the software was not up the standards to make use of. The 360 was using a triple core cpu and the first unified shader based gpu in 2005 when software and games were still being created for 1 core and gpu's with the older style dedicated lanes and processors for graphics work that did not work together. PS3 was using a gimped G70 "geforce 7800" gpu and later on had the assistance from the general purpose processor the "Cell" using its SPE's to handle other graphics work helping off load the RSX allowing more to be done.
Because of this "new" shift in hardware standards and the time learning and making use of this very customized hardware those console's abilities got better as the software and game engines matured finally hitting a plateau in graphics ability roughly 4 years after release mainly do the lack of memory, this forced developers to compress and shrink OS foot prints, learn and create new ways of streaming compressed data along with creating using new less taxing forms of effects and filters etc like FXAA.
Now lets fast forward to 2013 with the new generation of consoles. their performance ratio is nowhere near of their previous predecessors. Their design shifted from using top tier hardware and losing money per unit to using a different method of using what was available and keeping the costs within of actually making money from the start hence using lower performing hardware to fit the bill. Both went to AMD to design what they wanted within the budgets. And what came out were consoles very similar to each other and to the architecture of Pc's. Because of this, coding and knowing the hardware in general would be easy for the developers. The learning curve will not be as long or steep as previous console generations. Most of the tools being used and the understanding of the hardware's abilities have already been laid down by Pc since 2009. With the shift and the design to go with weaker hardware but with more memory available, it will be the reverse from last gen where they ran out of memory before taxing the hardware to taxing the hardware before using all the memory available.
While developers will learn how to maximize the resources on both consoles and create new methods of doing things as time goes on. We will not see the massive progression in graphics ability from the start to the end like we have seen in the past excluding last gen ports and cross generational games.
Hello. If you look at Oblivion and then you look at Skyrim on the same platform, it seems to be a big visual jump.
Are these first set of games for PS4 and XboxOne not really a good indication of what the powa they are capable of translating into visuals?
At the moment, all multiplats are cross gen ports, meaning that there the PS4 version is similar to the PC version in most cases, where the last gen consoles is running in low settings equivelent. However, already the PS4 exclusives such as Killzone SF, and Infamous SS have shown significant increases in their geometry over pervious gen titles of the same genre. We will see massive jumps.
This is an Uncharted 4 model leaked. They are said to have over 200,000 polygons. To think, Uncharted 3 had 30,000 polygons for character models. So now they have enough polygons to render each individual hair on his beard and head pretty much. And not just character models, but over all geometry of the levels will increase dramatically.
If the face alone has over 200,000 poly's the full bodies must be over a million, then add all the overall geometry of the levels! I shudder to think just what kind of performance a PS4 would push it out at, if one even could.
I'm not dissing the PS4, it's a decent amount of power for a low price, but first party developers with no interest in any other systems can't get their games to run at 1920 x 1080 and 60 fps.
I was talking full character models. You are probably reffering to how a 7850 gpu performs on PC with current games. Fact is, KZSF already has more geometry than any other FPS. Take Crysis 3 on PC for example, while it has much better effects and textures than the console version, and arguably is some of the best graphics in games today, it still has the same character models, animations, geometry, and game code as the PS3/360 versions. This is an example of consoles holding PC's back. Killzone SF has noticably better geometry than Crysis 3, and character models, and animations etc. This is because it is not worrying about the PS3/360.
As for multiplats running at less than 1080p 60fps, these are unoptimized compared to a console exclusive which has all the development time in one system. They can focus more on the fundamentals of the game because of this.
KZSF usage of polygons for npc models is depending on distance, amount of NPC's on screen. While they claim 40k model/s its with only one model on screen within a few feet of you on screen. The developers created a complex method of adjusting polygons based on # of and distance. Having more polygons does not mean better results in quality overall when its normally is not rendered.
Hello. If you look at Oblivion and then you look at Skyrim on the same platform, it seems to be a big visual jump.
Are these first set of games for PS4 and XboxOne not really a good indication of what the powa they are capable of translating into visuals?
At the moment, all multiplats are cross gen ports, meaning that there the PS4 version is similar to the PC version in most cases, where the last gen consoles is running in low settings equivelent. However, already the PS4 exclusives such as Killzone SF, and Infamous SS have shown significant increases in their geometry over pervious gen titles of the same genre. We will see massive jumps.
This is an Uncharted 4 model leaked. They are said to have over 200,000 polygons. To think, Uncharted 3 had 30,000 polygons for character models. So now they have enough polygons to render each individual hair on his beard and head pretty much. And not just character models, but over all geometry of the levels will increase dramatically.
If the face alone has over 200,000 poly's the full bodies must be over a million, then add all the overall geometry of the levels! I shudder to think just what kind of performance a PS4 would push it out at, if one even could.
I'm not dissing the PS4, it's a decent amount of power for a low price, but first party developers with no interest in any other systems can't get their games to run at 1920 x 1080 and 60 fps.
I was talking full character models. You are probably reffering to how a 7850 gpu performs on PC with current games. Fact is, KZSF already has more geometry than any other FPS. Take Crysis 3 on PC for example, while it has much better effects and textures than the console version, and arguably is some of the best graphics in games today, it still has the same character models, animations, geometry, and game code as the PS3/360 versions. This is an example of consoles holding PC's back. Killzone SF has noticably better geometry than Crysis 3, and character models, and animations etc. This is because it is not worrying about the PS3/360.
As for multiplats running at less than 1080p 60fps, these are unoptimized compared to a console exclusive which has all the development time in one system. They can focus more on the fundamentals of the game because of this.
KZSF usage of polygons for npc models is depending on distance, amount of NPC's on screen. While they claim 40k model/s its with only one model on screen within a few feet of you on screen. The developers created a complex method of adjusting polygons based on # of and distance. Having more polygons does not mean better results in quality overall when its normally is not rendered.
Look, I'm not looking for another argument with another 'know it all' hermit. Fact is, on all fully 8th gen games (including PC of course), geometry will increase over 7th gen games (even PC ports like Crysis 3). As for Killzone, source that all this 'cut down methods' went on please. And please do because no one ever likes to link to proof of this apart from their own 'speculation'. Basically, all i'm saying, is that for the same amount of distance rendered, Killzone SF beats out last gen FPS games in terms of polygon count. Really is it worth trying to correct people while not giving evidence? Also, you really think 40,000 poly models is the limit? Infamous SS has 120,000 polygon models in an open world game. There is arount 11 million polygons being rendered on average in the game (obviously not if you look up in the sky or something). Not saying PC games can't do this, which lets face it, is the only reason you were itching to post on me, because you thought I was getting 'too carried away' about PS4's capabilities. Why do you people feel the need to keep people in check all the time?. Fact is, as it stands, PC has no 8th gen games out there. I am simply saying that 8th Gen games across all platforms will increase in geometry (fundamental assets), by a long way over 7th Gen. Simple as.
Its not an argument about the increases of polygon counts from last gen to this gen. But once you hit a point having more polygons on models does not mean more detail. Its called diminishing returns, These consoles also cant run Tessellation to the degree as higher end gpu pc's too. With ISS its the same story as Killzone etc, Characters/environmental details usually have many levels LODs, which will vary in number of polygons/triangles according to the distance and locations they are viewed. which means you hardly ever see the full count all the time unless your in a cut scene or in an area where not much is there.
KIllzone ratios, and you can see, further the distance and more npcs the poly counts drop.
We're no where near diminishing returns on poly count hermits will stop posting that bs when they get an 8th gen game.
,lol so much denial, about polygon counts and diminishing returns , Diminishing returns doesn't mean "a wall" it means that throwing more polygons at things does not magically make something 10x or 100x better.Its is much more important for its all about the textures and bump mapping and normal mapping which is much more important to the detail of the model than making your models ten's of thousands or even hundreds of thousand of polygons. where is only reached in specific scenarios.
We're no where near diminishing returns on poly count hermits will stop posting that bs when they get an 8th gen game.
lol so much denial, about polygon counts , its all about the textures and bump mapping and normal mapping which his much more important to the detail of the model than making your models ten's of thousands or even hundreds of thousand of polygons. where is only reached in specific scenarios.
But we're not just talking about models though my dude. Inb4 you post that meme.
The hardware will allow higher standards and the abilities in graphics with things like lighting and physics and more complex and detailed worlds then last gen that is true. But the jumps that some are suggesting just wont be like the same ratio as last gen because of the hardware limitations and the ease of knowing the hardware and coding from the start.
We're no where near diminishing returns on poly count hermits will stop posting that bs when they get an 8th gen game.
lol so much denial, about polygon counts , its all about the textures and bump mapping and normal mapping which his much more important to the detail of the model than making your models ten's of thousands or even hundreds of thousand of polygons. where is only reached in specific scenarios.
But we're not just talking about models though my dude. Inb4 you post that meme.
Diminishing returns doesn't mean "a wall" it means that throwing more polygons at things does not magically make something 10x or 100x better.
when it comes to system wars, i've observed a kind of a polygon paradox. i was thinking about this last night.
really, you want a mesh with clean topology eg: less polys, you don't want unnecessary topology because you have to render that unnecessary topology and that will take 'work'. looking through 'showreel' equivalents of 3D modellers, many of them 'flex' their ability to do things like take 5M poly sculpts and drop them down to 10K poly characters ready for in-game deployment.
but i've seen people argue polygon count as a metric for graphical goodness. well, does it make sense to argue polygon count when they'll want to keep topology as simple as possible without compromising their vision for the visuals? i'm not sure it does. it's kind of circular because a high poly count mesh might actually be a very poor mesh, and a low poly count mesh could look better.
obviously a model can be highly detailed yet also have an efficient, clean topology. a scene with many "good" yet low poly count models will render much better than a scene with few "poor" yet high poly count models.
anyway. @_Matt_ i consider you an expert in the field. feel free to eviscerate my argument if I'm completely wrong, but it seems to me that poly count isn't a very useful metric when it comes to assessing visuals or system power?
^Your soo full of it..... the amount of polygons for the world environment is not the main concern, its all about the mapping and meshes of those objects and areas. and lol with the bias crap ignoring the facts
Doubt it, and I don't really care. Graphics are already beyond what I need for an immersive game session. I want better game experiences far more than better graphics.
2) Last gen was fucking weird. The tech was alien to a lot of devs and so it took a while for devs to unlock their full potential. This gen, consoles use the same x86 tech as PC and so, as quite a few devs say, today, we're already looking at games near the full potential of this gen like ISS
2) Last gen was fucking weird. The tech was alien to a lot of devs and so it took a while for devs to unlock their full potential. This gen, consoles use the same x86 tech as PC and so, as quite a few devs say, today, we're already looking at games near the full potential of this gen like ISS
Even though consoles can offer 2x the performance of the equivelent PC because of low level access to the hardware?
That was proven false, and to the fact that optimization includes downgrading. Of course no body wants to include that tid bit
Its not an argument about the increases of polygon counts from last gen to this gen. But once you hit a point having more polygons on models does not mean more detail. Its called diminishing returns, These consoles also cant run Tessellation to the degree as higher end gpu pc's too. With ISS its the same story as Killzone etc, Characters/environmental details usually have many levels LODs, which will vary in number of polygons/triangles according to the distance and locations they are viewed. which means you hardly ever see the full count all the time unless your in a cut scene or in an area where not much is there.
KIllzone ratios, and you can see, further the distance and more npcs the poly counts drop.
Well I know that LOD's effect the amount of polygons on screen. But listen. I am not speaking of PS4 here, I am speaking of the 8th Generation in general. LOD's exist on all games (obviously), same with killzone, Crysis 3, all the way back to the PS1 days. All games npc's and poly counts drop the further away they are. But, I am simply reffering to the amount of polygons per distance rendered. And individual objects in game. Lets say a vase on a desk in a 7th gen game was composed of 500 polygons, on an 8th gen game, it might have 1,600 polygons. It's that simple. That's the detail I speak of.
The primary LOD's will increase in quality over time. But it's down to the developer to work on transitioning those LOD's and making them as seemless as possible. For instance, ISS actually only has 3 LOD's for each 3D object. The first LOD is very detailed mostly, but the second one is quite half arsed and looks blocky even to the average viewer. If they had more time and budget, they would have created more in between LOD's because it's easy to be buzzkilled when looking at those shitty LOD's while enjoying those graphics. However, regardless of this fact, ISS regularly renders over 11 million polygons to screen. This is because within the 100 meter radius, All objects (that are very detailed for an open world game), and the main character model, is all being rendered fully. Further than 100 meters away from your character, the secondary LOD's kick in. These are much lower resolution than the primary ones. But regardless, this often adds up to millions of polygons on screen (which is far beyond what the PS3/360 can do). It all depends on how the developer manages the LOD's.
My point is, polygon count, character models, animations, AI, physics, all this good stuff is going to improve vastly in the 8th generation over the 7th. Has done over all console generations. Of course it is reaching the diminishing returns mark, because a 60,000 polygon character to a 120,000 polygon character won't be as noticable as last gen increases because the polygons are so small now.
Speaking of diminishing returns. How about the time and budget it would take to get these ultra high fidelity visuals at a strong resolution and framerate. The way things went last generation with so many cutbacks and studios closing, i'm not sure that is worth the effort. A game like Killzone looks fantastic already and graphics should improve, but as stated the learning curve isn't there. I think we can only expect some combination of slightly better visuals, better resolutions, or better performance. But not all 3 for a given game
Log in to comment