A Generational Leap in Graphics [2020] *Spawn*

[Dampf]

Comparison between cutscene model and ground truth offline rendered model.

It's amazing to think we are nearly there. We have that technology to achieve similar detail to the right image in real time right now.

Real Time Raytracing for light, eye reflections and shading. Sampler Feedback for much better skin texture detail. Much improved hair rendering using the much more powerful CPUs or machine learning based algorithm on GPU. Mesh shaders for higher geometry and micro detail. And I'm sure the peach fuzz is possible too.

All we have to do now is to wait for devs to go crazy with these new technologies and leave last gen behind!

 

[techuse]

Sure about that? Uncharted 3 did look very nice graphically. Larger gap then DS to CP2077 pc perhaps, but dont forget that leaps have gotten smaller on pc too, but to a much lesser degree (hardware wise). Blame power usage id say.

Uncharted 3 was a great looking game. ND did amazing things with that hardware. It didn't come anywhere near Crysis however.

The gap is enormous. This is why I just shake my head when people call Cyberpunk the new Crysis,

 

[PSman1700]

The gap is enormous. This is why I just shake my head when people call Cyberpunk the new Crysis,

When people say its the new Crysis, they mean the new Crysis to todays/modern standards. I think leaps where bigger backthen, even on pc. I can also think that, if they really wanted, and pushed for a setting like Crysis instead of a city with lots of going on, they ofcourse could have achieved even more. They are abit different in setting. The gap was bigger indeed. Still, i think UC3 looked really good, in special the desert environments (somehow they held up much better then the greenery/city ones). A clear leap over what the PS2 could ever do or even dream of.

 

[chris1515]

Yes. The same for G70 to GCN1, you cant compare them TF for TF either. Fact remains that in TF metrics alone, the improvement was 10 times the increase, as opposed to 5 what we got now. Thats pure metrics. Counting in going from G70 arch to GCN1.1 arch and the improvement is even larger then what we see from GCN to RDNA. If the PS5 GPU is 7 times more powerfull then the PS4s, the PS4s gpu was close to 20 times more powerfull. Architectural changes account for both.
And no, basing performance increases purely based on resolution is faulty to begin with. In special considering last gen games.



Maybe use common sense.



Still doesnt change on what we got as the best visuals possible on PS5 at launch vs the PS4 at launch. No matter what, if the leap was as big as Chris1515 promises, we sure should have seen it in a game natively designed around the PS5 (bluepoint claims so). In special if the development time was three years.

When you ave no idea of what you are talking, shut up!!! When I talk about something I have an idea of what I am talking not like you.

https://fr.slideshare.net/Slide_N/playstation-cutting-edge-techniques

A bit of EDGE library

Visibility and Geometry Culling • Vertex and lighting processing • Post processing Offloading the GPU; 41. Slide 41 • 'Edge' SPU library to offload vertex work from RSX™ – Trade SPU time for GPU performance • Animation, ... Network Processing on an SPE Core in Cell Broadband EngineTM.

Some of the rendering workload done on CELL SPU. DICE was very good with SPU/SPE for example.

And I don't promise any gap, I said it will be very visible at least in gameplay.

It's amazing to think we are nearly there. We have that technology to achieve similar detail to the right image in real time right now.

Real Time Raytracing for light, eye reflections and shading. Sampler Feedback for much better skin texture detail. Much improved hair rendering using the much more powerful CPUs or machine learning based algorithm on GPU. Mesh shaders for higher geometry and micro detail. And I'm sure the peach fuzz is possible too.

All we have to do now is to wait for devs to go crazy with these new technologies and leave last gen behind!

Horizon 2 trailer Aloy has some peach fuzz, better geometry end of visible polygon on ears , better skin rendering and better hair.

EDIT:
https://www.diva-portal.org/smash/get/diva2:388137/FULLTEXT01.pdf

Migrating Mesh Skinning Deformation Functionality from RSX to SPUs on the PlayStation 3

 

[PSman1700]

When you ave no idea of what you are talking, shut up!!! When I talk about something I have an idea of what I am talking not like you.

It seems its the opposite. I am stating that in hardware terms, the leap is smaller this time around. Even architectural wise this is the case, as provided by another poster above.

Some of the rendering workload done on CELL SPU.

Which means the Cell is providing its power to some GPU resources instead of CPU logic. It doesnt change the story all that much.

And I don't promise any gap, I said it will be very visible at leasy in gameplay.

There will be a gap, i didnt say there wouldnt be. PS5 games are going to be better looking then PS4 games (much so), in special 60fps seems to be getting the norm (and higher resolutions ofcourse). What i do think is that the leap from PS3 to PS4 was bigger. Its evident in the first party games designed around each system at launch. Aside from looking at metrics, which people always have done on this forum.

 

[chris1515]

It seems its the opposite. I am stating that in hardware terms, the leap is smaller this time around. Even architectural wise this is the case, as provided by another poster above.



Which means the Cell is providing its power to some GPU resources instead of CPU logic. It doesnt change the story all that much.



There will be a gap, i didnt say there wouldnt be. PS5 games are going to be better looking then PS4 games (much so), in special 60fps seems to be getting the norm (and higher resolutions ofcourse). What i do think is that the leap from PS3 to PS4 was bigger. Its evident in the first party games designed around each system at launch. Aside from looking at metrics, which people always have done on this forum.

Again CELL SPU were very powerful for the time, it was sharing some of the ressource from graphics processing but doing other things at the same time, with animation, physics, sound processing. It does not change the story at all.

The Jaguar CPU is the reason AC Unity or Just Cause 4 framerate trouble. And it is the reason any i3 CPU was good enough to have 60 fps on PS4/Xbox 1 games. The weakness of the CPU was good for PC player who can keep the same CPU for years and easily reach 60 or 100 fps or more without having the last CPU.

There is a good reason, the PS5 Tempest engine use SPU DMA memory model...

The other poster did not give any proof like I said this is good to talk about the G70 but this is not the full picture. The full picture is CELL SPU + G70.

On this forum where you are since a few months, people looks at real world performance not paper specs if everyone could define a level of performance benchmark would not be useful. All face off website would not be useful.

In Tflops paper specs 1080Ti is superior to 2070 Super and Vega 7 is superior to 1080Ti, 2080, 2080 Super and as powerful as a 2080 Ti. But if we add paper memory bandwidth Vega 7 have much more than 2080 Ti in rasterization it is a superior GPU to 2080 Ti.

<ModEdit> Tlops paper specs means absolutely nothing. I show you use a non useful metric, something not important.

I love this spec so good.
https://www.amd.com/en/products/graphics/amd-radeon-vii

THE WORLD’S FIRST 7nm GAMING GPU

People don't buy a GPU reading AMD website...

Again compare what is comparable Killzone Shadow Fall. Demon's souls is a remake of a 2009 PS3 game it means the level design and construction is not build for the PS5 or the PS4. It limit what they can do. If you listen well to the Digitalfoundry interview, they don't know what to do with the CPU power. And in Killzone shadow Fall you have some scene like the city one, you can't do on Demon's souls remake because they are limited by the original game. They packed the scene as much as they could but this is only a remake of a 2009 game.

Like I said wait a little, games fully designed around PS5 not remake, maybe God of War Ragnarok will be only next gen. I joke, this will be a fucking cross gen game.

EDIT:
https://www.extremetech.com/gaming/...hat-killzone-shadow-fall-was-originally-290gb

The statement comes directly from Guerrilla Games technical director Michiel van der Leeuw. He posited that the ludicrous original 290GB size of Killzone Shadow Fall is mainly because the game doesn’t have any assets that have been made for consoles with a lower spec. He did mention cross-generation games as a main reason for this, as the upcoming Killzone is PS4-only, rather than releasing on both the PS3 and PS4 like the upcoming Assassin’s Creed IV or NBA 2K14. He also noted that Shadow Fall’s larger, open areas contribute to that ridiculous 290GB size, as the surface area of the zones are about five-to-10 times larger than those in Killzone 3.

The area is Killzone Shadow Fall are larger than on Demon's souls because this is a PS3 games and a remake.

 

[PSman1700]

The Jaguar CPU is the reason AC Unity or Just Cause 4 framerate trouble.

And still, the jaguar was a massive upgrade over the Cell in general gaming tasks. The Cell was only very good at certain tasks. Efficiency was out the window too.

The other poster did not give any proof like I said this is good to talk about the G70 but this is not the full picture. The full picture is CELL SPU + G70.

He provided insight that for the GPUs, going from the G70 RSX to the AMD GCN1.1 GPU, that architectural improvement alone accounts for the TF difference the PS4 to PS5 gives. No matter how you put it, Cell helping out abit here and there, the GPU leap was much larger back then, as opposed what we get this time around.

On this forum where you are since a few months, people looks at real world performance not paper specs if everyone could define a level of performance benchmark would not be useful. All face off website would not be useful.

For the first, i clearly havent been here 'for a few months'. Second, registration date doesnt equal technical expertise. Paper specs are one thing, thats true. But they cant be ignored either. On the other hand, just going by a resolution increase and claim by soley going after that to say GPU x is 7times stronger vs GPU y isnt the whole story either. In special considering what were looking at. Going from PS3 to 4 wasnt an as easy transition as going from PS4 to 5. Also, scaling has improved alot, and the obvious focus on cross-gen transition we never saw on a level like this before. Mid gen refreshes are a part of last generation too, easing things up aswell. Scaling has never been as important now as back then.

In Tflops paper specs 1080Ti is superior to 2070 Super and Vega 7 is superior to 1080Ti, 2080, 2080 Super and as powerful as a 2080 Ti. But if we add paper memory bandwidth Vega 7 have much more than 2080 Ti in rasterization it is a superior GPU to 2080 Ti.

Again, i fail to see what a 1080Ti vs 2070S has to do with the whole discussion. I never claimed that GCN flops equal RDNA flops. What i am saying is that G70 flops dont equal GCN flops. Going from G70 to GCN provided a massive architectural improvement.

A clown is a clown or maybe you need to smoke less weed or drink less alcohol, Tlops paper specs means absolutely nothing.

I guess that this kind of commentary has got you banned over at GAF before. I dont mind it, its just the internets. But for your own good, keep things civil, a mod doing his/her duty might actually take action sometime, somewhere. Not now, but sometime.

TFlops do mean something, but they are not everything and cant be compared like-for-like between architectures. We know that, did i ever say somewhere differently? If so, quote me on it.

Again compare what is comparable Killzone Shadow Fall. Demon's souls is a remake of a 2009 games it means the level design and construction is not build for the PS5 or the PS4. It limit what they can do. If you listen well to the Digitalfoundry interview, they don't know what to do with the CPU power. And in Killzone shadow Fall you have some scen like the city one, you can't do on Demon's souls remake because they are limited by the original game.

Well, ofcourse they are different games. So where UC3 (best looking PS3 title) for PS3 and SF for PS4. Yet, the leap was large enough to even have DF as of 2020 mention how large a jump it was. A larger leap then what DS does over the best looking PS4 titles. Same for order 1886, yes the gameplay was a dud (although i think it had a place), the game totally blew away everything backthen, a larger leap then what we have now.

They dont know what to do with the CPU power? I mean comon.... What about higher framerates, finally a baseline of 60fps for starters?
Shadowfall's city scenes seem to me harder to do something amazing with then the smaller environments DS is playing around with. Anyway, if the leap would be as large as the last one, we should have seen it by now, since we do have a PS5 native exclusive from a AAA studio that worked on the game for three years. Same goes for rift apart, except for the faster loadings, ofcourse.

Like I said wait a little, games fully designed around PS5 not remake, maybe God of War Ragnarok will be only next gen. I joke, this will be a fucking cross gen game.

Yeah, we'l be waiting i guess. And again il be hammering it, looking back at say 1886, going to UC4 and HZD, and finally GoT etc, i do not think the differences where generations apart either. I still think HZD/DS take the cake. So looking at demon souls as great as it is (and rift apart), i doubt we will see such a leap from that either.
its not like the PS2 days anymore, where we really saw huge differences between launch day and final year games. Thankyou X86/pc parts instead of exotic designs.

 

[chris1515]

And still, the jaguar was a massive upgrade over the Cell in general gaming tasks. The Cell was only very good at certain tasks. Efficiency was out the window too.



He provided insight that for the GPUs, going from the G70 RSX to the AMD GCN1.1 GPU, that architectural improvement alone accounts for the TF difference the PS4 to PS5 gives. No matter how you put it, Cell helping out abit here and there, the GPU leap was much larger back then, as opposed what we get this time around.



For the first, i clearly havent been here 'for a few months'. Second, registration date doesnt equal technical expertise. Paper specs are one thing, thats true. But they cant be ignored either. On the other hand, just going by a resolution increase and claim by soley going after that to say GPU x is 7times stronger vs GPU y isnt the whole story either. In special considering what were looking at. Going from PS3 to 4 wasnt an as easy transition as going from PS4 to 5. Also, scaling has improved alot, and the obvious focus on cross-gen transition we never saw on a level like this before. Mid gen refreshes are a part of last generation too, easing things up aswell. Scaling has never been as important now as back then.



Again, i fail to see what a 1080Ti vs 2070S has to do with the whole discussion. I never claimed that GCN flops equal RDNA flops. What i am saying is that G70 flops dont equal GCN flops. Going from G70 to GCN provided a massive architectural improvement.



I guess that this kind of commentary has got you banned over at GAF before. I dont mind it, its just the internets. But for your own good, keep things civil, a mod doing his/her duty might actually take action sometime, somewhere. Not now, but sometime.

TFlops do mean something, but they are not everything and cant be compared like-for-like between architectures. We know that, did i ever say somewhere differently? If so, quote me on it.



Well, ofcourse they are different games. So where UC3 (best looking PS3 title) for PS3 and SF for PS4. Yet, the leap was large enough to even have DF as of 2020 mention how large a jump it was. A larger leap then what DS does over the best looking PS4 titles. Same for order 1886, yes the gameplay was a dud (although i think it had a place), the game totally blew away everything backthen, a larger leap then what we have now.

They dont know what to do with the CPU power? I mean comon.... What about higher framerates, finally a baseline of 60fps for starters?
Shadowfall's city scenes seem to me harder to do something amazing with then the smaller environments DS is playing around with. Anyway, if the leap would be as large as the last one, we should have seen it by now, since we do have a PS5 native exclusive from a AAA studio that worked on the game for three years. Same goes for rift apart, except for the faster loadings, ofcourse.



Yeah, we'l be waiting i guess. And again il be hammering it, looking back at say 1886, going to UC4 and HZD, and finally GoT etc, i do not think the differences where generations apart either. I still think HZD/DS take the cake. So looking at demon souls as great as it is (and rift apart), i doubt we will see such a leap from that either.
its not like the PS2 days anymore, where we really saw huge differences between launch day and final year games. Thankyou X86/pc parts instead of exotic designs.

I said GCN 1.1 and RDNA 2 GPU aren't comparable too, this is the same thing and visible with backward compatibility games. Realworld performance.

Battlefield 4 PS3 versus PS4

PS3 and 360 version
https://www.eurogamer.net/articles/digitalfoundry-battlefield-4-current-gen-face-off

PS4 and Xbox One version
https://www.eurogamer.net/articles/digitalfoundry-battlefield-4-next-gen-face-off

BattleField 4 is 1280 x 704 30 fps on PS3 and 1600*900 60 fps on PS4. The PS4 push only 3.19 times more pixel than the PS4 but with better settings. And Battledield 4 runs very well on PS3 better than 360 but CELL SPU were really well exploited at the end of generation.

Battlefield 4 PS4 suffers a lot with destruction going down to 30 fps in some mode and maps and with performance from 40 to 50 fps on PS4 because Jaguar is shit... At least this time we have very stable 60 fps .

https://www.eurogamer.net/articles/digitalfoundry-battlefield-4-next-gen-face-off

I was never in GAF, wrong... And I speak like this because you always have the same argument and nonsense repeat ad vitam eternam when no paper spec means nothing not on consoles or PC. Saying RDNA2 Tflops is equivalent to GCN 1.1 Tflops is exactly the same than saying g70 Tflops and GC1.1 Tflops is equivalent. In this case, I don't use an imaginary 20 times more powerful GPU but realworld performance and they are comparable.

The settings of Destiny 2 are only high on PS5 like PS4 Pro but some effect are in very high in BF4 PS4. I think this is comparable.

EDIT: TO1886, UC4 and HZD runs on the same machine and are very different game. TO1886 is a fully linear game, UC4 is wide linear and HZD is open world. Streaming is a bigger limitation on UC4 and even more HZD.

TO1886, UC4 use baked ligthing. HZD is dynamic lighting.

And again all this game are fully designed around PS4 I am sure we will see a gap when we will go out of this shitty cross gen period on PS5 and Xbox Series X.

 

[pjbliverpool]

I agree up to a point. I'd say the performance gap between a 3090+DLSS and a PS5 is clearly bigger than G80 and the PS3. You can look at any gen 7 console title throughout their lifetime and not a single one ever came close to matching Crysis. 7 years of developer improvement and nothing came close. Similar to DX12U being new, there were several new paradigms of the time. Learning them improved things dramatically, but not nearly enough to close the gap. GPU wise Crysis performance was justified.

I'm not sure the comparison is entirely fair though as we're much further down the road of diminishing returns today than we were in 2007. Back then, it was relatively easy to turn superior GPU power into superior visual results. Today it's much less trivial, especially when you look at how resource intensive RT is.

I do agree with you though that CB2077 isn't as revolutionary as Crysis. That was never really my point. Only that it's a clear graphical step above everything that's come before - at least in it's genre.

Comparing the best looking games on a base PS4 to Cyberpunk on a 3090 and there is just no way the visuals aren't punching well below their weight with the 35-40x more GPU capability available.

I agree on this, but like I said in my last post, if we're talking about how well CB2077 utilises the available power then the comparison is much more favourable if you discount RT which is obviously a massive resource hog when used heavily. Removing RT, equalising resolution to the PS4 norm, and using more optimised standard settings still gives you a game that looks superb but runs fine (by console standards) on hardware that is only twice as fast as the base PS4. You may still argue that the graphics at those settings don't warrant double the PS4's horsepower but that's a much more ambiguous argument that the 35-40x argument above. So yes, while the game can look much better at Ultra settings with Ultra RT at 4K than at 1080p medium without RT, you're definitely well into the point of diminishing returns at that stage.

 

[PSman1700]

I said GCN 1.1 and RDNA 2 GPU aren't comparable too, this is the same thing and visible with backward compatibility games. Realworld performance.

Yes, and? I never said they where. TF for TF, RDNA2 is doing much more. But so does GCN over G70, obviously. Backwards compatible, yes, not something that did exist on PS3 to PS4 exactly. The comparison is kinda flawed there i think.

Now you take on BF4 again, have you actually played the game on both PS3 and 4? Its a world of a difference, try multiplayer. On PS3, the settings equal to lower then low (yes, i have tested that backthen), the FPS is capped to 30 opposed to 60 on PS4. Further then that, map sizes and player slots are limited to 24 players on smaller maps. Thats a fckn huge difference to me. Aside from the rather laggy and stuttery experience the PS3 version was.

Its not just 'but with better settings' either, settings matter alot, going from low all the way to Ultra (what BF4 pS4 is running, about), is a huge difference to performance. Going from low to Ultra, while doubling the frame rate, and upping the resolution considerably.... and then also enabling 64 player matches as opposed to 24 on the PS3, all the while opening up the maps to the larger sizes. Sounds like a generation difference to me. Forget about the PS3s performance when things get hasty.

Battlefield 4 suffers a lot with destruction going down to 30 fps and with performance from 40 to 50 fps on PS4 because Jaguar is shit...

See what it does to the PS3.

I was never in GAF, wrong... And I speak like this because you always have the same argument and nonsense repeat ad vitam eternam when no paper spec means nothing not on consoles or PC. Saying RDNA2 Tfops is equivalent to GCN 1.1 Tflosp is exactly the same than saying g70 Tflops and GC1.1 Tflops. In this case, I don't use and imaginary 20 times more powerful GPU but realworld performance and they are comparable.

Well, wherever forum that was you got banned for attacking members. Anyway, on to the technical discussion.
Again, i am not claiming that paper specs mean everything, they clearly do not. Im aligning those with what we see. Where did i state that RDNA flops equal GCN flops? Quote me on it.

Il try to explain again, we both are no english native speakers so it could be that. What i mean is, going from G70 to GCN, there was a huge architectural improvement, which means a better IPC for GCN. Its the same for GCN to RDNA2. Some argue though that the architectural improvements seen going from G70 to GCN1.1 where bigger then going from GCN to RDNA. But i cant prove it. Another member agreed on me on that one though.

TO1886, UC4 and HZD runs on the same machine and are very different game. TO1886 is a fully linear game, UC4 is wide linear and HZD is open world. Streaming is a bigger limitation on UC4 and even more HZD.

TO1886, UC4 use baked ligthing. HZD is dynamic ligthng.

Okay, i was just looking at the overall fidelity of those titles. But lets say going from UC4 thiefs end to last of us 2, both are using baked lighting right. Both are ND (one of the better studios), one is early title, one is a last. Last of us 2 is doing things better obviously, but saying they are heaps and perhaps a generation apart isnt what most are going to agree with.
To this day, i still think UC4 looks amazing, in some scenes/environments.... the water, effects, lighting. While last of us did improve on the engine, their close enough to me. Its more in the details between those two. You'd also have to compare PS4 versions for this, the Pro obviously lifts the resolution and stabilizes the frame rates for the latter.

And again all this game are fully designed around PS4 I am sure we will see a gap when we will go out of this shitty cross gen period on PS5.

DS and rift apart are fully designed around and for the PS5 only. Its what i was comparing to. Theres a reason i from the start left out cross gen titles since they wont tax any new system to their capabilities. But since you went on BF4, well, there was my view on it.

Edit: as a fair sidenote, BF4 PS3 in todays standards would have been de-listed day one. Its performance is such a hot mess, and considering BF is a MP oriented title to start with (where systems get stressed, not SP), 24 player matches, 32 sized maps, lower then low settings accompanied by the framerate as it was, that had nothing to do with Battlefield anymore. And no, the 10fps comment isnt an over-stated. I personally had the game freezing to a complete halt. Its that typical CP2077 example of where a late-gen game doesnt cut it on old hardware.

https://answers.ea.com/t5/Battlefield-4/Battlefield-4-laggs-really-bad-on-ps3/td-p/1843901

 

[chris1515]

Yes, and? I never said they where. TF for TF, RDNA2 is doing much more. But so does GCN over G70, obviously. Backwards compatible, yes, not something that did exist on PS3 to PS4 exactly. The comparison is kinda flawed there i think.

Now you take on BF4 again, have you actually played the game on both PS3 and 4? Its a world of a difference, try multiplayer. On PS3, the settings equal to lower then low (yes, i have tested that backthen), the FPS is capped to 30 opposed to 60 on PS4. Further then that, map sizes and player slots are limited to 24 players on smaller maps. Thats a fckn huge difference to me. Aside from the rather laggy and stuttery experience the PS3 version was.

Its not just 'but with better settings' either, settings matter alot, going from low all the way to Ultra (what BF4 pS4 is running, about), is a huge difference to performance. Going from low to Ultra, while doubling the frame rate, and upping the resolution considerably.... and then also enabling 64 player matches as opposed to 24 on the PS3, all the while opening up the maps to the larger sizes. Sounds like a generation difference to me. Forget about the PS3s performance when things get hasty.



See what it does to the PS3.



Well, wherever forum that was you got banned for attacking members. Anyway, on to the technical discussion.
Again, i am not claiming that paper specs mean everything, they clearly do not. Im aligning those with what we see. Where did i state that RDNA flops equal GCN flops? Quote me on it.

Il try to explain again, we both are no english native speakers so it could be that. What i mean is, going from G70 to GCN, there was a huge architectural improvement, which means a better IPC for GCN. Its the same for GCN to RDNA2. Some argue though that the architectural improvements seen going from G70 to GCN1.1 where bigger then going from GCN to RDNA. But i cant prove it. Another member agreed on me on that one though.



Okay, i was just looking at the overall fidelity of those titles. But lets say going from UC4 thiefs end to last of us 2, both are using baked lighting right. Both are ND (one of the better studios), one is early title, one is a last. Last of us 2 is doing things better obviously, but saying they are heaps and perhaps a generation apart isnt what most are going to agree with.
To this day, i still think UC4 looks amazing, in some scenes/environments.... the water, effects, lighting. While last of us did improve on the engine, their close enough to me. Its more in the details between those two. You'd also have to compare PS4 versions for this, the Pro obviously lifts the resolution and stabilizes the frame rates for the latter.



DS and rift apart are fully designed around and for the PS5 only. Its what i was comparing to. Theres a reason i from the start left out cross gen titles since they wont tax any new system to their capabilities. But since you went on BF4, well, there was my view on it.

Edit: as a fair sidenote, BF4 PS3 in todays standards would have been de-listed day one. Its performance is such a hot mess, and considering BF is a MP oriented title to start with (where systems get stressed, not SP), 24 player matches, 32 sized maps, lower then low settings accompanied by the framerate as it was, that had nothing to do with Battlefield anymore.

https://answers.ea.com/t5/Battlefield-4/Battlefield-4-laggs-really-bad-on-ps3/td-p/1843901

Video compare campaign between the two and like I said Some settings go higher on PS4 but the same for Destiny 2 higher settings on PS5 but the gap is higher between the PS3 and PS4 settings of Battlefield but it push less pixel. I think this is comparable and 60 fps with slowdown on PS4 for Battlefield 4, Destiny 2 is locked 60 fps on PS5.

Like I said DS is limited in what he does because of being a remake. Rift Apart is a true PS5 game but again cartoonish graphics are often not considered by many peoples as impressive as realistic game. For example, I see no interest to compare R&C Rift Apart and Killzone Shadow Fall too different.

Compare R&C PS4 and R&C PS5 will be more interesting, this was the first PS4 game of Insomniac Games.

We can compare Killzone Shadow Fall and the next Battlefield or the next FPS by Embark Studios. Battledield will release in 2021 and be the first DICE Game on PS5/Xbox Series. This will be a true PS5/Xbox Series/PC games no shitty cross gen. Like I said I think DICE and Embark Studios will save the beginning of the gen and will probably make all first party cross gen games look outdated.

Sorry but I see a huge gap between a cross gen game like TLOU remastered and UC4 and like I said we will see a huge gap in realistic graphic gameplay when we will see true PS5/Xbox Series X.

 

[BillSpencer]

Assassins Creed: Pirates
-vs-
Assasins Creed: Vikings

does not count because of reasons right?

 

[PSman1700]

Video compare campaign between the two and like I said Some settings go higher on PS4 but the

Some settings? The PS3 version is running lower then low for many (almost all settings). The PS4 is basically maxed to PC's Ultra (i think mesh and texture where lowered to high or medium, cant recall). All the while doubling frame rate and upping the resolution. If we dive into MP, things are getting even worse for the PS3. See my previous post for those. Like said, BF4 on PS3 would be de-listed day one. Also, settings impact performance alot, in special the higher ones, whilest the visual returns arent that big as performance would suggest. On the other hand, it shows how large a leap the PS4 version still was in terms of what it could do.

Like I said DS is limited in what he does because of being a remake. Rift Apart is a true PS5 game but again stylish graphics are often not considered by people as impressive as realistic game. For example, I see no interest to compare R&C Rift Apart and Killzone Shadow Fall too different.

You would have a point if DS was a remaster, its not. Its on a totally new engine on a totally new platform. DS is a true PS5 game, as said by many, its not cross gen. The latter of your senstences there i wont even comment, since were diving into personal preferences there.

We can compare Killzone Shadow Fall and the next Battlefield or the next FPS by Embark Studios. This will ve true PS5/Xbox Series/PC games no shitty cross gen. Like I said I think DICE and Embark Studios will save the beginning of the gen and will probably make all first party cross gen games look outdated.

Again, im looking at native PS5 (non cross gen) games for a reason.

Sorry but I see a huge gap between a cross gen game like TLOU remastered and UC4 and like I said we will see a huge gap in realistic graphic gameplay when we will see true PS5/Xbox Series X.

The difference between UC4/thiefs end and last of us 2 isnt a generation apart. They both look amazing, even though last of us 2 engine improved upon it. Its not the PS2 and PS3 days anymore where there where real massive differences between launch and late gen games.

 

[chris1515]

Assassins Creed: Pirates
-vs-
Assasins Creed: Vikings

does not count because of reasons right?

AC Black Flag and AC Valhalla is interesting too. Other things on PS4 and XB1 cross gen or not 60 fps was very often not stable at all.

 

[chris1515]

Some settings? The PS3 version is running lower then low for many (almost all settings). The PS4 is basically maxed to PC's Ultra (i think mesh and texture where lowered to high or medium, cant recall). All the while doubling frame rate and upping the resolution. If we dive into MP, things are getting even worse for the PS3. See my previous post for those. Like said, BF4 on PS3 would be de-listed day one. Also, settings impact performance alot, in special the higher ones, whilest the visual returns arent that big as performance would suggest. On the other hand, it shows how large a leap the PS4 version still was in terms of what it could do.



You would have a point if DS was a remaster, its not. Its on a totally new engine on a totally new platform. DS is a true PS5 game, as said by many, its not cross gen. The latter of your senstences there i wont even comment, since were diving into personal preferences there.



Again, im looking at native PS5 (non cross gen) games for a reason.



The difference between UC4/thiefs end and last of us 2 isnt a generation apart. They both look amazing, even though last of us 2 engine improved upon it. Its not the PS2 and PS3 days anymore where there where real massive differences between launch and late gen games.

Lol Cyberpunk 2077 was not delisted because of the performance but because of CDprojek msimanagement and lies. EA did not hide Battlefield 4 PS3 to reviewer, CEO did not lie to shareholder and they did not promise refund of the game without asking Sony or MS approval. CDProjekt executives management of the situation is a joke, they deserve everything they receive but it is another problem.

BF4 shadow settings is behind ultra too.

No it was design with PS3 in mind, and a very linear game with tiny environnements not comparable to PS4 design no need to go on PS5. They would have design the game differently if it was a true PS5 game or a remake of a PS4 game.

The difference is that first UC4 release three years after the launch. There is a good reason why ND never release game at a beginning of a generation. UC4 development was make fully on final devkit for example*, this is not the case of all launch period PS5 or PS4 games. If we compare Killzone Shadow Fall and TLOU2 the gap is pretty big not a generation apart but huge.

*And they receive some advices of first party studio doing some launch game.

 

[techuse]

I'm not sure the comparison is entirely fair though as we're much further down the road of diminishing returns today than we were in 2007. Back then, it was relatively easy to turn superior GPU power into superior visual results. Today it's much less trivial, especially when you look at how resource intensive RT is.

I do agree with you though that CB2077 isn't as revolutionary as Crysis. That was never really my point. Only that it's a clear graphical step above everything that's come before - at least in it's genre.



I agree on this, but like I said in my last post, if we're talking about how well CB2077 utilises the available power then the comparison is much more favourable if you discount RT which is obviously a massive resource hog when used heavily. Removing RT, equalising resolution to the PS4 norm, and using more optimised standard settings still gives you a game that looks superb but runs fine (by console standards) on hardware that is only twice as fast as the base PS4. You may still argue that the graphics at those settings don't warrant double the PS4's horsepower but that's a much more ambiguous argument that the 35-40x argument above. So yes, while the game can look much better at Ultra settings with Ultra RT at 4K than at 1080p medium without RT, you're definitely well into the point of diminishing returns at that stage.

I don't believe it's an issue of diminishing returns. The generation may prove me wrong but I think we have quite a ways to go before diminishing returns start to become a legitimate concern. I do think from a budgetary stance it becomes increasingly expensive however. The jury is still out for me on whether RT is the most effective use of resources with todays performance levels. I've not seen much of Cyberpunk without ultra settings and ultra + RT so I'm unable to give my opinion on how it looks at lower settings.

 

[PSman1700]

Lol Cyberpunk 2077 was not delisted because of the performance but because of CDprojek msimanagement and lies. EA did not hide Battlefield 4 PS3 to reviewer, CEO did not lie to shareholder and they did not promise refund of the game without asking Sony or MS approval. CDProjekt executives management of the situation is a joke, they deserve everything they receive but it is another problem.

Ah well yes, true to that then. My point was barely performance though, BF4 actually ran worse on PS3 then 2077 does/did on base PS4. On top of that, it didnt even resemble anything 'Battlefield' anymore.

The difference is that first UC4 release three years after the launch. There is a good reason why ND never release game at a beginning of a generation. UC4 development was make fully on final devkit for example*, this is not the case of all launch period PS5 or PS4 games. If we compare Killzone Shadow Fall and TLOU2 the gap is pretty big not a generation apart but huge.

Well, ND is actually launching rift apart soon. And from what ive seen, its not looking any more impressive then say demon souls. Going from 1886, SF, UC4 to last of us 2 and GoT, i dont find that gap that huge. There are neat improvements, but nowhere near 'a massive leap'. I think that what we got in fidelity today with demon souls is quit neat, dont expect massive leaps in fidelity going from there. It didnt happen for the PS4 either.

But, time will tell. Hardware didnt see as large a increase as PS3 to 4 did, but maybe the software did/does. Intresting times ahead indeed, i too think that DICE's next Battlefield could be a real showstopper. Its assumably not going to be a PS4 game.

 

[chris1515]

Ah well yes, true to that then. My point was barely performance though, BF4 actually ran worse on PS3 then 2077 does/did on base PS4. On top of that, it didnt even resemble anything 'Battlefield' anymore.



Well, ND is actually launching rift apart soon. And from what ive seen, its not looking any more impressive then say demon souls. Going from 1886, SF, UC4 to last of us 2 and GoT, i dont find that gap that huge. There are neat improvements, but nowhere near 'a massive leap'. I think that what we got in fidelity today with demon souls is quit neat, dont expect massive leaps in fidelity going from there. It didnt happen for the PS4 either.

But, time will tell. Hardware didnt see as large a increase as PS3 to 4 did, but maybe the software did/does. Intresting times ahead indeed, i too think that DICE's next Battlefield could be a real showstopper. Its assumably not going to be a PS4 game.

BF4 run under 30 fps on PS3 but it doesn't run at all at 60 fps on PS4. It runs at 60 fps Jaguar style tablet framerate shit. Like I said and I will repeat it again, this is comparable at least on GPU side, all the other things out of memory bandwidth leap are better from PS4 to PS5. And for comparison purpose you use exactly the same situtation, this is why the benchmark is the campaign.

And the memory bandwith difference is not big knowing better delta color compression on RDNA 2 GPU and less bandwith contention on unified memory side from an AMD patent where they improve memory controller giving bigger priority to CPU call knowing it is more latency sensitive.

Ratchet is made by Insommiac. You don't even know who develop games.

Again false and Jaguar was not better than CELL in many workloads and GPU leap is not better again realworld performance but maybe saying the same bullshit again and again will do it and there is a chance this false assumption will be true.

<ModEdit>

 

[PSman1700]

Again false and Jaguar was not better than CELL

Wow, now thats you actually being wrong.

GPU leap is not better again realworld performance but maybe saying the same bullshit again and again will do it and there is a chance this false assumption will be true.

The GPU leap in pure numbers is only five times for the PS5. For the PS4, this was actually ten times. Thats factual numbers. Architectural improvements happened to both shifts.
Looking at real world performance, that what we got today, Shadowfall and Order 1886 where a bigger leap from PS3s best then what DS and rift apart are to the PS4s best (at launch).
You failed to get anywhere with the BF4 example. It actually was one of the biggest improvements ever going from PS3 to 4. From a mess to a near to pc experience. The PS3 was running 30fps, a much lower resolution, and lower then lowest settings. On top of that, the game was lagging so much it was unplayable. If it was playable, you'd be enjoying 24 player matches as opposed to 64 ones on the PS4. No idea what that equals in metrics, but thats going to be a huge one, a hell of a leap.
Its even more of an achievement if the game actually was done for the PS3 in mind, which i know it was not. Again, your the one coming with false claims here, and acting like a child with GIFs that certainly dont belong here, aside from big mouthing.

 

[chris1515]

Wow, now thats you actually being wrong.



The GPU leap in pure numbers is only five times for the PS5. For the PS4, this was actually ten times. Thats factual numbers. Architectural improvements happened to both shifts.
Looking at real world performance, that what we got today, Shadowfall and Order 1886 where a bigger leap from PS3s best then what DS and rift apart are to the PS4s best (at launch).
You failed to get anywhere with the BF4 example. It actually was one of the biggest improvements ever going from PS3 to 4. From a mess to a near to pc experience. The PS3 was running 30fps, a much lower resolution, and lower then lowest settings. On top of that, the game was lagging so much it was unplayable. If it was playable, you'd be enjoying 24 player matches as opposed to 64 ones on the PS4. No idea what that equals in metrics, but thats going to be a huge one, a hell of a leap.
Its even more of an achievement if the game actually was done for the PS3 in mind, which i know it was not. Again, your the one coming with false claims here, and acting like a child with GIFs that certainly dont belong here, aside from big mouthing.

Again being sub 30 fps doesn't matter, what matter is the leap and like I said in theory it push only 3.19 more pixel but the settings gap is bigger making the deficit in pixel. And same PS4 framerate is more in the 40 to 50 fps this is far from being a locked 60 fps, the reason is mostly the weak Jaguar CPU. And again you compare two different workloads in multiplayer <ModEdit>. Imagine scientific people trying to reproduce an experiment but doing it in different conditions.

The important place is the campaign. This is exactly the same reason dictator choose one PS5 cutscene to compare AC Valhalla to a PC build where the PS5 build run at 1440p and the reason the video of Digitalfoundy use the campaign as a comparison between PS3 and PS4.

Destiny 2 run at 30 fps a much lower resolution(the gap is higher than BF4) and lower settings too, high on PS5 like PS4 Pro not ultra but high. The gap settings advantage is on the leap between PS4 and PS3 but there is a pixel and settings advantage too for PS5. I think the gap is comparable on GPU side.