• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

[Digital Foundry] Batman Arkham City PS4 Pro: What If Every Game Got A Pro Upgrade?

onQ123

Member
To answer your original question: comes to mind. But you can't just write off Bloodborne, because that game also dips in places on top of the framepacing issues (as do most if not all games).


So you would use a game that has a 4K mode , 60 fps mode & better graphics mode on PS4 Pro as a example of the games that are getting noticeable performance boost on Xbox One S as a result of doing a brute force upgrade vs patches that take advantage of the system?


I'm asking how much bigger is the list of Xbox One games that got a noticeable performance boost from Xbox One S brute force method vs games getting PS4 Pro enhancement.


Edit: never mind you said Tomb Raider Definitive Edition I was thinking about Rise Of The Tomb Raider.

Edit: how noticeable is the performance boost when played on Xbox One S? if things haven't been updated PS4 was closer to 60fps while Xbox One was closer to 30fps.

https://www.youtube.com/watch?v=RWWtm4Wq9QU
 

Synth

Member
I'm asking how much bigger is the list of Xbox One games that got a noticeable performance boost from Xbox One S brute force method vs games getting PS4 Pro enhancement.

This is a pointless thing to ask though. The only reason the Xbox One S is being brought up is to support the "concept" of passive improvements, not to show the amount of games that would benefit. The XB1S is a very marginal increase in performance, unlike the PS4 Pro that has a significant delta to the standard PS4. There would logically be far more games that would have been noticeably improved by that hardware jump, than the current XB1S. It's like comparing a GTX970 and a GTX1060, and using that to argue there wouldn't be much to gain from upgrading to a GTX1080. It's a silly argument to have.

Why not ask how many PSP games see a noticeable performance increase with an upclock? Would that inconvenience the point you're trying to shield yourself from?
 

onQ123

Member
This is a pointless thing to ask though. The only reason the Xbox One S is being brought up is to support the "concept" of passive improvements, not to show the amount of games that would benefit. The XB1S is a very marginal increase in performance, unlike the PS4 Pro that has a significant delta to the standard PS4. There would logically be far more games that would have been noticeably improved by that hardware jump, than the current XB1S. It's like comparing a GTX970 and a GTX1060, and using that to argue there wouldn't be much to gain from upgrading to a GTX1080. It's a silly argument to have.

Why not ask how many PSP games see a noticeable performance increase with an upclock? Would that inconvenience the point you're trying to shield yourself from?

On one hand people are using the fact that Xbox One S was able to brute force Xbox One enhancements without breaking anything as a reason why Sony should have done the same thing with the PS4 Pro & on the other hand you have people saying that it shouldn't be compared because Xbox One S is just a small up clock over the Xbox One vs PS4 Pro being a bigger upgrade over PS4.

My question was not pointless because if people are saying that Sony should have done what MS done with Xbox One S they should at least have a example of how much bigger the list of games that would get a noticeable performance boost would be vs the PS4 Pro approach.


So it's ok to say that PS4 Pro should have done it because Xbox One S did it but when I ask for the results of Xbox One S method it's pointless?


So no one has the actual results but keep using it as a talking point? Now I'm the bad guy for asking for the numbers?
 

Synth

Member
On one hand people are using the fact that Xbox One S was able to brute force Xbox One enhancements without breaking anything as a reason why Sony should have done the same thing with the PS4 Pro & on the other hand you have people saying that it shouldn't be compared because Xbox One S is just a small up clock over the Xbox One vs PS4 Pro being a bigger upgrade over PS4.

My question was not pointless because if people are saying that Sony should have done what MS done with Xbox One S they should at least have a example of how much bigger the list of games that would get a noticeable performance boost would be vs the PS4 Pro approach.

So it's ok to say that PS4 Pro should have done it because Xbox One S did it but when I ask for the results of Xbox One S method it's pointless?

So no one has the actual results but keep using it as a talking point? Now I'm the bad guy for asking for the numbers?

No, the amount of games that are noticeably improved on the XB1S is pointless. Nobody's ever going to go DF-style on every game to find out, so it can be either far more than you're assuming, or far less than others are... you're not going to get this number, so it's irrelevant. That the XB1S provides noticeable passive improvements is the point, and it doing so with such a small increase makes it logical that the PS4P with a more than 2x increase to the stats, would have seen far more noticeable performance gains, across more games (regardless of what number of XB1S games do).

Now, I don't actually disagree with you that the XB1S and the PS4P situations are different, and just because things "just work" on XB1, attempting to do the same wouldn't blow a bunch of games up on PS4P. A lot of that is likely just down to forward planning, with MS having planned for XB1 games to potentially run on differing hardware configurations (with the whole Win 10 push and all). I'm not convinced "small increase vs large increase" is the difference, when you consider that the Scorpio is looking to do the same with a much more dramatic change to the hardware spec.. If you were simply stating that Sony likely can't just "flip a switch" and expect everything to work, then I'd agree with you... however if you're trying to suggest that there would be little noticeable differences to plenty of games with poor framerates (like, say Until Dawn), dynamic resolutions, etc with 2x the performance without patching.. then I think you're crazy.

As I said before, if you ignore the XB1S for a moment, to avoid looking at the situation from an Xbox vs PlayStation perspective, you can already see similar (and very plentiful) examples of what a passive upgrade can do with the PSP. The number of games is a meaningless deflection of the overall point.
 

onQ123

Member
No, the amount of games that are noticeably improved on the XB1S is pointless. Nobody's ever going to go DF-style on every game to find out, so it can be either far more than you're assuming, or far less than others are... you're not going to get this number, so it's irrelevant. That the XB1S provides noticeable passive improvements is the point, and it doing so with such a small increase makes it logical that the PS4P with a more than 2x increase to the stats, would have seen far more noticeable performance gains, across more games (regardless of what number of XB1S games do).

Now, I don't actually disagree with you that the XB1S and the PS4P situations are different, and just because things "just work" on XB1, attempting to do the same wouldn't blow a bunch of games up on PS4P. A lot of that is likely just down to forward planning, with MS having planned for XB1 games to potentially run on differing hardware configurations (with the whole Win 10 push and all). I'm not convinced "small increase vs large increase" is the difference, when you consider that the Scorpio is looking to do the same with a much more dramatic change to the hardware spec.. If you were simply stating that Sony likely can't just "flip a switch" and expect everything to work, then I'd agree with you... however if you're trying to suggest that there would be little noticeable differences to plenty of games with poor framerates (like, say Until Dawn), dynamic resolutions, etc with 2x the performance without patching.. then I think you're crazy.

As I said before, if you ignore the XB1S for a moment, to avoid looking at the situation from an Xbox vs PlayStation perspective, you can already see similar (and very plentiful) examples of what a passive upgrade can do with the PSP. The number of games is a meaningless deflection of the overall point.

How can the number of games that got a noticeable boost from Xbox One S be pointless in a thread where it's being said that PS4 Pro should have done what Xbox One S done?


How is the numbers pointless when that's the whole point of people asking for the brute force approach? if the numbers are pointless then it shouldn't matter if it's brute force or upgraded by devs.
 

Synth

Member
How can the number of games that got a noticeable boost from Xbox One S be pointless in a thread where it's being said that PS4 Pro should have done what Xbox One S done?


How is the numbers pointless when that's the whole point of people asking for the brute force approach? if the numbers are pointless then it shouldn't matter if it's brute force or upgraded by devs.

Because they wouldn't be equivalent anyway. The PS4P is so much more of a performance increase, that the number of games it would affect would be substantially higher. Again, why are the XB1S numbers so important, but not the PSP (which you also won't get a number for... but you can trust it's a large amount)?

People are talking about the brute force approach, because it would apply also to games that the developer doesn't want to apply an explicit update to. It wouldn't prevent such patches from existing also... so every explicitly updated PS4P game would be in addition to the passive improvements on games that aren't updated. It's not either/or so "these numbers vs those numbers" isn't the point.

How is that so difficult for you to understand?
 

Marmelade

Member
How can the number of games that got a noticeable boost from Xbox One S be pointless in a thread where it's being said that PS4 Pro should have done what Xbox One S done?


How is the numbers pointless when that's the whole point of people asking for the brute force approach? if the numbers are pointless then it shouldn't matter if it's brute force or upgraded by devs.

Because you're comparing a 7% or whatever GPU boost with something far bigger
That approach would obviously be far more noticeable on the Pro
 

Izuna

Banned
This very video shows what he might get in other games. Nothing else needs to be said other than exactly what DF says in the OP. :s
 

mattp

Member
Regardless of the game being shown here, I agree that it would be great to see an option to run standard PS4 games using the additional power of the system. As long as it's noted to be a "use at your own risk" kind of option, I don't see the issue.

The PS2 is a good example of this, I think, since faster disc speed definitely could break some games while texture filtering didn't always work as intended. They were optional and had to be engaged specifically when loading a PS1 game each time. Why not offer this Pro users?

It would actually help out in the case of mediocre ports. If you have a game that is almost 30fps but often drops, this extra power might be enough to bring it up to a solid 30.

this seems like the ideal way to handle things
 

Panajev2001a

GAF's Pleasant Genius
I agree, but about the bolded - why do you think 'generations' approach is better than 'iterative'? With generations I think the risk is we'd lose forward compatibility, i.e. there's little chance that PS5 games will be playable on original PS4, whereas Microsoft seems to be going into the future where any Xbox title can be played on any Xbox console (or Xbox device - tablet, phone) just that the experience would be severely compromised on low-end machines. I don't think even Sony will once again try implementing some disruptive, innovative tech (like Cell) because the R&D going into that vs. the hurdles put on the developers to learn new paradigms are not really worth it. It's smarter to leverage the commoditised PC tech and - if possible - add some tweaks here & there, like they did for PS4 and now for PS4 Pro. I think Sony should - unfortunately for performance - also go further "away from the metal" to higher abstraction level APIs, because it's good for business. Devs can then maintain single code base for their game, with just separate compiler paths for specific devices. Games are nowadays so big and complex that adding specific solutions for their console would be counterproductive, IMO.

Obviously, annual console iterations would be overkill, but if it's happening every 3-4 years and I'd know my purchase could last me 6-7 years if I don't need the top model then I'd be fine with it.

I do not find value in forward compatibility as much as backward compatibility, which is a must for PS5, for what I want a console to be... quite the opposite actually, I think it would kill what the console experience is for customers and developers: a piece of fixed specs HW that developers can invest on and target knowing that they have enough time to get a good return on investment and the financial incentive to do so (the reset at launch helps developers differentiate themselves and keeps the developers and the console makers sectors more competitive instead of just locking people to the huge giant market leader more and more, it helps keeping the market competitive).

What you are essentially suggesting is that we could replace consoles with PC's to deliver something that at best is almost the same as we have now and at worst is a the kind of forever cross generation situation I find on mobile now (beside some built in apps, buying a new iPhone means waiting a long while before developers support it...).
When I bought a PS1, PS2, Xbox, GameCube, Wii, Xbox 360, Wii U, PS4, etc... I had purchased that could last me 5+ years too and in the case of Sony consoles I actually got lots of software almost till the day the console stopped being manufactured (amazingly close to it at least ;)).

What is stopping consoles a bit now, and one of the big reasons these stop gap consoles are here for, is that manufacturing node improvements are farther and farther away from each other and more and more expensive each time... this in turns hurts the old depreciation model driving console prices down over time while allowing the manufacturer to make money (more so than at launch too). It hurts its viability to the point that it is more lucrative for them to keep releasing new HW and keep people hungry for a new console instead of a healer Slim model.
 

antyk

Member
If you keep a purely iterative model, where is the incentive to invest in meaningful leaps forward? Why bother pushing to get an 8-16 core zen or HBM memory, if you're held back by the need for previous consoles to play the same games? We see this to an extent on iOS. The latest iPhones and iPads don't have many games that fully push the hardware because the developers rightly want to address the largest possible market which includes much older hardware.

I wouldn't worry about it, just like I'm certain PC tech will evolve as a result of requirements of 4K, VR, pretty soon 8K, then holograms or whatever :) What we'd probably avoid is going down the 'weird' avenues like Cell architecture, dedicated processors, etc. Sure, it's great to go there from the R&D perspective and e.g. Cell paved the way for multicore / wide processing and programming techniques, which now reaps the benefits in GPU-centric environments; but it costed Sony (and co.) a lot to develop, costed devs a lot to understand and master it, costed Sony a market share and costed the gamers few years of poor and inferior ports. WIth current complexity and scope of games it really doesn't make sense to invest in convoluted, proprietary tech just for the sake of being different. Gone are the times when consoles were ahead or at least on par with PCs at release, because Sony/MS/Nintendo are no longer seeing business incentive in subsidising the hardware and current generation is best example of what kind of "jump" we could expect from - theoretical, or not - PS5 and Xbox2.

I have a console simply because I don't have to think about system updates & maintenance, it plays all the new games with acceptable quality and my smaller kids can use it without my assistance - it's mostly plug and play. If it can perform its function better or cheaper using commoditised PC hardware, then I'm all for it.
 

antyk

Member
What you are essentially suggesting is that we could replace consoles with PC's to deliver something that at best is almost the same as we have now and at worst is a the kind of forever cross generation situation I find on mobile now (beside some built in apps, buying a new iPhone means waiting a long while before developers support it...).

Actually, yes. Developers already build their games to scale on PCs, so why not extend the same model to console? In fact I think this is (and has been in PS3/X360 generation) the case already, because if you look at games - and disregard resolution, texture quality, shadows, AF, AA methods, etc. - they're exactly the same, which wasn't the case before PS3/X360 with the same games looking and sometimes playing vastly different on different hardware.

I don't think iterative model will kill progress - just look at how many games are already supporting PS4 Pro with higher resolutions, better details and more stable framerates. Do you think the same would be possible, if it was a more fundamental shift in underlying technology and the engines were not built with flexibility in mind?

BTW, we have opposite view on compatibility - why would I really want to play on new hardware games I finished years ago? What I care for is that I can play new games until I switch to newer console model. I'm not against backwards compatibility (in fact I expect Sony will have it this time around), but forward compatibility is I think more important - this is why I was wondering about Sony's policies regarding software abstraction if it changed over the last couple of months. This is also the unfortunate reason why most mobile games target mid-range phones, simply because there's more money in that segment.
 

Turrican3

Member
I'm *still* a bit confused by PS4 Pro default, "hi-compatibility" behaviour.

I understand they turn off half GPU, but what about clocks?
Is the CPU (and the other half GPU, memory, etc.) still running at the same speed as the standard PS4 or do they sport higher clocks as well, as per Pro specifications?
 

dex3108

Member
I'm *still* a bit confused by PS4 Pro default, "hi-compatibility" behaviour.

I understand they turn off half GPU, but what about clocks?
Is the CPU (and the other half GPU, memory, etc.) still running at the same speed as the standard PS4 or do they sport higher clocks as well, as per Pro specifications?

Everything goes down to PS4 specs as far as I know.
 

Turrican3

Member
Ok, thanks!

So, is there a general consensus on what's going on here with Batman?
I mean, I guess developers have been using the Pro for a while, so is it really *that* unlikely to think code was already intentionally there to detect and possibly leveraging on the extra power?

(sorry if the question sounds dumb, I haven't been following closely these Pro/Scorpio debates)
 
Regardless of the game being shown here, I agree that it would be great to see an option to run standard PS4 games using the additional power of the system. As long as it's noted to be a "use at your own risk" kind of option, I don't see the issue.

The PS2 is a good example of this, I think, since faster disc speed definitely could break some games while texture filtering didn't always work as intended. They were optional and had to be engaged specifically when loading a PS1 game each time. Why not offer this Pro users?

It would actually help out in the case of mediocre ports. If you have a game that is almost 30fps but often drops, this extra power might be enough to bring it up to a solid 30.

I would like to see Sony do this, even if you have to jump through a few hoops to enable it and have a disclaimer screen you have to OK before starting a non-Pro patched game in Pro mode.

The fact Sony hasn't done this may be because it really does break the majority of games they have already tested. Instead of looking at Pro mode as just an overclock, you should remember that the you are doubling the amount of GCN compute units as well. If you compare this to SLI/Crossfire on PC, you will know that this can cause a lot of problems in games that do not natively support SLI such as causing texture or shadow flickering, introducing micro-stutter / framerate or framepacing issues, lighting problems and more. These games need to be patched by the developer and/or Nvidia & AMD need to patch their drivers. I know SLI isn't the same thing as the Pro situation but it may result in similar issues depending on how the Pro's new architecture works.

There could also be issues with netcode on multiplayer games as result of these changes too.

This is just speculation of course. Providing a Pro-override on non-Pro games would be a good selling point for the Pro, I don't think Sony would play it so safe unless they really had to.
 
The fact Sony hasn't done this may be because it really does break the majority of games they have already tested. Instead of looking at Pro mode as just an overclock, you should remember that the you are doubling the amount of GCN compute units as well. If you compare this to SLI/Crossfire on PC, you will know that this can cause a lot of problems in games that do not natively support SLI such as causing texture or shadow flickering, introducing micro-stutter / framerate or framepacing issues, lighting problems and more. These games need to be patched by the developer and/or Nvidia & AMD need to patch their drivers. I know SLI isn't the same thing as the Pro situation but it may result in similar issues depending on how the Pro's new architecture works.

That is as a result of the pre-dx12 driver model and how AFR is basically a driver hack. I do not think it is a good direct analogy for enabling the other compute units on the GCN chip of the PS4pro APU or having the CPU cores run at the PS4pro base clock.
 

andshrew

Member
Ok, thanks!

So, is there a general consensus on what's going on here with Batman?
I mean, I guess developers have been using the Pro for a while, so is it really *that* unlikely to think code was already intentionally there to detect and possibly leveraging on the extra power?

(sorry if the question sounds dumb, I haven't been following closely these Pro/Scorpio debates)

All that is going on here is that the developers made the game Pro aware/compatible. That's where their efforts (seemingly) end. They have haven't included higher resolution support, or higher quality assets like other developers they're simply allowing their game to take advantage of the Pro's hardware.
 

antyk

Member
The fact Sony hasn't done this may be because it really does break the majority of games they have already tested. Instead of looking at Pro mode as just an overclock, you should remember that the you are doubling the amount of GCN compute units as well. If you compare this to SLI/Crossfire on PC, you will know that this can cause a lot of problems in games that do not natively support SLI such as causing texture or shadow flickering, introducing micro-stutter / framerate or framepacing issues, lighting problems and more. These games need to be patched by the developer and/or Nvidia & AMD need to patch their drivers. I know SLI isn't the same thing as the Pro situation but it may result in similar issues depending on how the Pro's new architecture works.

There could also be issues with netcode on multiplayer games as result of these changes too.

It has been already mentioned to someone else in this topic, but there's no 2nd GPU in Pro's. Mark Cerny only used an analogy (of mirroring the existing part) that people got carried away with - they simply doubled the amount of CUs (+ they're newer architecture & higher clock) and not somehow doubled the GPU with all it's buses, schedulers, memory controllers, etc. SLI is a bad analogy, it's just like the same, but bigger GPU.
 
This very video shows what he might get in other games.
No it doesn't. It shows what you get when you use Pro mode without altering your render targets. The developer specifically had to code for this, so we have no idea how the game would've reacted if the higher clocks were active for the old code. Maybe the game would be improved just like this. Or maybe something would break, and what he'd get is a game that runs worse than before.
 

Lord Error

Insane For Sony
So no one has the actual results but keep using it as a talking point? Now I'm the bad guy for asking for the numbers?
I can think of a few really good games that would have frame drops eliminated, and probably run at locked framerate if the "Pro" resources were unlocked to them:

Witcher 3, Bloodborne (still would have frame pacing problem but no drops), GTAV, AC:Unity, Life is Strange, every single Unity engine game except Inside, armageddon sequence at the end of a Resogun level. Just on top of my head.

Then again, for all I know, those games would simply crash trying to run on a new hardware, so the point I'm trying to make is worthless.

No it doesn't. It shows what you get when you use Pro mode without altering your render targets. The developer specifically had to code for this, so we have no idea how the game would've reacted if the higher clocks were active for the old code.
A actually ascribe to Richard's theory that a "Pro mode" compiler switch was enabled on this game by accident, and I seriously doubt anything was actually coded specifically for it. The reason being that they didn't even enable it for the other game in the collection.
 

Shari

Member
No it doesn't. It shows what you get when you use Pro mode without altering your render targets. The developer specifically had to code for this, so we have no idea how the game would've reacted if the higher clocks were active for the old code. Maybe the game would be improved just like this. Or maybe something would break, and what he'd get is a game that runs worse than before.

Nope, this is just activating pro mode for the game, that is all, no code related except a flag.
 
I actually ascribe to Richard's theory that a "Pro mode" compiler switch was enabled on this game by accident, and I seriously doubt anything was actually coded specifically for it. The reason being that they didn't even enable it for the other game in the collection.
Nope, this is just activating pro mode for the game, that is all, no code related except a flag.
This isn't fact, it's a theory. It's equally plausible that there was no accident; that this was an easy "get" for a port that didn't have the budget to develop a more enhanced Pro version. And the other game in the collection doesn't do this because they tried, and it broke, so they had to revert it.
 

Izuna

Banned
No it doesn't. It shows what you get when you use Pro mode without altering your render targets. The developer specifically had to code for this, so we have no idea how the game would've reacted if the higher clocks were active for the old code. Maybe the game would be improved just like this. Or maybe something would break, and what he'd get is a game that runs worse than before.

I think you're misunderstanding here. This wasn't coded specifically for Pro mode. As DF assumes, it's just not downclocking to the OG PS4.

If you think they're wrong, then you disagree with the ENTIRE premise of this thread.
 

Interfectum

Member
Yeah I def agree with DF here.

Let us toggle on Pro mode for all our games and use at our own risk.

Also... that PS4 Pro CPU. -_-
 

andshrew

Member
I think you're misunderstanding here. This wasn't coded specifically for Pro mode. As DF assumes, it's just not downclocking to the OG PS4.

If you think they're wrong, then you disagree with the ENTIRE premise of this thread.

What you're saying doesn't make sense. You can't just not downclock the Pro, you have to specifically enable Pro support or 'awareness' in your title. In this case they've done just that, but kept everything about their game the same. Same resolution, same assets; and so unsurprisingly on much faster hardware they get much higher performance.

Everyone was wondering why Arkham City ran at an unlocked frame rate compared with Arkham Asylyum which did not. I'd say this is a pretty clear indicator as to why.
 

NXGamer

Member
I think you're misunderstanding here. This wasn't coded specifically for Pro mode. As DF assumes, it's just not downclocking to the OG PS4.

If you think they're wrong, then you disagree with the ENTIRE premise of this thread.
That is not what has happened here, they must have compiled Pro binaries for it to use the resource, otherwise the Hvisor would allocate base resource and use those instead. And, even if the game was rushed and had issues, I seriously doubt this was all done by mistake.
 

dark10x

Digital Foundry pixel pusher
It seems like some other games may have received silent support.

So I have Moto Racer 4 (which I expected more from but ultimately it's an average game).

On the standard PS4 it seems to average ~45-50fps during races. It isn't smooth at all.

Now, there's no mention of Pro support in this game but, on the Pro, it basically holds a solid 60fps now (outside of intro fly-bys).

The resolution hasn't changed from the base mode (which appears to be 900p) but the performance is so much better.

Leads me to believe that there could be other "secret" Pro games out there where developers implemented support for the system without mentioning it.
 
I think you're misunderstanding here. This wasn't coded specifically for Pro mode. As DF assumes, it's just not downclocking to the OG PS4.

If you think they're wrong, then you disagree with the ENTIRE premise of this thread.
You're the one misunderstanding. We all agree the game is merely using the increased clockrate of the Pro without attempting to improve graphics through other means. The disagreement is about why this is, and what it means.

Digital Foundry guesses that it's an accident, coming at the very end of development, which took literally zero effort. From this they conclude there would be no drawbacks to letting every game do this by default. But their entire scenario is just a guess, with absolutely no evidence pro or con. There are multiple other possibilities which would contravene their conclusion (and that of the folks who agree):

1. It was an accident, but it was very fortunate the game still works because most break if you do this.
2. It wasn't an accident, it was an explicit choice from the start. Recoding and lengthy testing was necessary to make sure the boost worked.
3. It wasn't an accident, and it was a trivial amount of work. But it only worked on this game, and the other game in the collection had to remain locked to original mode so it wouldn't break.
4. It was an accident, and the game does start glitching because of it, but only later in the runtime where Digital Foundry didn't examine.
Etc.

Just choosing one possible scenario that sounds plausible doesn't make it true.
 

Izuna

Banned
You're seriously only considering that because it suits your argument, where there is no indication that it would be the case. Lengthy testing for Arkham City to run at an uneven framerate? Hah.

And really, PSP, 3DS, Vita etc. shows that increasing the clock speeds of games in unintended ways doesn't break anything if at all. Cue the disbelief that it's a huge problem and the want for the option.

That is not what has happened here, they must have compiled Pro binaries for it to use the resource, otherwise the Hvisor would allocate base resource and use those instead. And, even if the game was rushed and had issues, I seriously doubt this was all done by mistake.

Or it could be as simple as the game being "flagged" to allow Pro mode. Like certain games having a number in an .NFO to dictate what software version it requires to run, this game could just have pro_mode = 1.

It's besides the point anyway, it doesn't matter how it is done, it indicates that this game was just allowed to run with the extra hardware in mind with nothing else done.

Otherwise, Cerny COULDN'T have allowed people to use the Pro hardware on unpatched games.

What you're saying doesn't make sense. You can't just not downclock the Pro, you have to specifically enable Pro support or 'awareness' in your title. In this case they've done just that, but kept everything about their game the same. Same resolution, same assets; and so unsurprisingly on much faster hardware they get much higher performance.

Everyone was wondering why Arkham City ran at an unlocked frame rate compared with Arkham Asylyum which did not. I'd say this is a pretty clear indicator as to why.

Perspective misunderstanding. I'm making the assumption that the Pro will automatically go into "compatibility" mode without a flag, not that it's specifically enabled per title. My comment entirely agreed with you.

--

Um, reading those posts back NXGamer and andshrew, I'm struggling to see where the misunderstanding came from.
 

cakely

Member
You're seriously only considering that because it suits your argument, where there is no indication that it would be the case. Lengthy testing for Arkham City to run at an uneven framerate? Hah.

And really, PSP, 3DS, Vita etc. shows that increasing the clock speeds of games in unintended ways doesn't break anything if at all. Cue the disbelief that it's a huge problem and the want for the option.

You can't really make that conclusion based on those platforms.

I agree with the other posters in this thread that somehow Arkham City was compiled with some Pro settings, and that as a result it's running with the full Pro GPU/CPU. I assume that it was intentional, but I'm really unsure why.

I also feel that it would be very unwise to simply enable those Pro settings by default for the 1100 existing PS4 games.

Unrelated: You're being really combative in this thread for no particular reason,
 

JP

Member
It seems like some other games may have received silent support.

So I have Moto Racer 4 (which I expected more from but ultimately it's an average game).

On the standard PS4 it seems to average ~45-50fps during races. It isn't smooth at all.

Now, there's no mention of Pro support in this game but, on the Pro, it basically holds a solid 60fps now (outside of intro fly-bys).

The resolution hasn't changed from the base mode (which appears to be 900p) but the performance is so much better.

Leads me to believe that there could be other "secret" Pro games out there where developers implemented support for the system without mentioning it.
I'll be honest, I was half expecting stuff like this to happen. Developers would surely have to contain some sort of mode (PS4 or PS4 Pro) selection in their games for the actually hardware and although I don't think we'll automatically get improvements on games but I wouldn't be that surprised to see some developers doing pushing the extra power into some less demanding games.

Again, I'm not suggesting it will be a Xbox One(s) type situation of it happening automatically but I would expect, relatively speaking, that the sort of improvements you mention aren't going to involve companies having to invest large bundles of money into getting it working. I'd expect it more on games like the one you mention where poor performance is way enough to remedy by doing such a thing.
 

andshrew

Member
Perspective misunderstanding. I'm making the assumption that the Pro will automatically go into "compatibility" mode without a flag, not that it's specifically enabled per title. My comment entirely agreed with you.

--

Um, reading those posts back NXGamer and andshrew, I'm struggling to see where the misunderstanding came from.

The confusion is arising because you're saying "this wasn't coded specifically for Pro mode". It's accessing the Pro's additional resources, and so it was coded specifically to enable Pro support.

On the subject of 'stealth' upgrades and support I believe it is actually possible to determine a games support for the Pro from the JSON files which describe PS4 games to the system. If you run your PS4 through a proxy you can see the URLs that it is accessing, and whenever you install a game or patch it pulls the JSON file for the game. Within this file is new element called neoEnable which is set to 1 on games which support the Pro.

Looking at Skyrim's file you can see this. I don't have the Batman games but if you did this you would no doubt see a 1 here too.
URL to the Skyrim JSON
Contents:
http://tmdb.np.dl.playstation.net/tmdb2/CUSA05486_00_81B3B202FE5BE6FC6443DFA10987B8D8F539B6B4/CUSA05486_00.json said:
{"revision":2,"formatVersion":4,"npTitleId":"CUSA05486_00","console":"PS4","names":[{"name":"Skyrim"}],"icons":[{"icon":"http://gs2-sec.ww.prod.dl.playstation.net/gs2-sec/appkgo/prod/CUSA05486_00/2/i_fedd9fe4b268454faf8d4c0c8e733c27a597990304e0894f58fb39705e322e26/i/icon0.png","type":"512x512"}],"parentalLevel":9,"pronunciation":"http://gs2-sec.ww.prod.dl.playstation.net/gs2-sec/appkgo/prod/CUSA05486_00/2/i_fedd9fe4b268454faf8d4c0c8e733c27a597990304e0894f58fb39705e322e26/i/pronunciation.xml","contentId":"EP1003-CUSA05486_00-SKYRIMHDFULLGAME","bgm":"http://gs2-sec.ww.prod.dl.playstation.net/gs2-sec/appkgo/prod/CUSA05486_00/2/i_fedd9fe4b268454faf8d4c0c8e733c27a597990304e0894f58fb39705e322e26/i/snd0.at9","category":"gd","psVr":0,"neoEnable":1}
 
Sony needs to be pressured to allow this. And of course devs might have to do something to enable brute force boost. I am guessing they have to have unlocked framerate so that Pro can brute force the framerate? If so all games have to have an option to enable unlocked framerate when running on Pro and let the user use it if they choose to. I am not sure how willing Sony will be to allow this.
 

5taquitos

Member
Sony needs to be pressured to allow this. And of course devs might have to do something to enable brute force boost. I am guessing they have to have unlocked framerate so that Pro can brute force the framerate? If so all games have to have an option to enable unlocked framerate when running on Pro and let the user use it if they choose to. I am not sure how willing Sony will be to allow this.
It's my understanding that Sony does allow it. It's up to the devs to enable access to the Pro power. Even in Batman's case in this video. Every dev could go back and do this if they wanted.
 

onQ123

Member
It seems like some other games may have received silent support.

So I have Moto Racer 4 (which I expected more from but ultimately it's an average game).

On the standard PS4 it seems to average ~45-50fps during races. It isn't smooth at all.

Now, there's no mention of Pro support in this game but, on the Pro, it basically holds a solid 60fps now (outside of intro fly-bys).

The resolution hasn't changed from the base mode (which appears to be 900p) but the performance is so much better.

Leads me to believe that there could be other "secret" Pro games out there where developers implemented support for the system without mentioning it.



Maybe it's just that some games use fp16 more & the fact that PS4 Pro GPU is double rate fp16 give them a boost.
 

Izuna

Banned
The confusion is arising because you're saying "this wasn't coded specifically for Pro mode". It's accessing the Pro's additional resources, and so it was coded specifically to enable Pro support.

On the subject of 'stealth' upgrades and support I believe it is actually possible to determine a games support for the Pro from the JSON files which describe PS4 games to the system. If you run your PS4 through a proxy you can see the URLs that it is accessing, and whenever you install a game or patch it pulls the JSON file for the game. Within this file is new element called neoEnable which is set to 1 on games which support the Pro.

Looking at Skyrim's file you can see this. I don't have the Batman games but if you did this you would no doubt see a 1 here too.
URL to the Skyrim JSON
Contents:

Fair enough.

That's about what I thought, nice find. I wish there was a jailbreak just so we could see what that would do for other games, if it'd enable the full Pro hardware or not. We won't know until its too late I guess.

You can't really make that conclusion based on those platforms.

I agree with the other posters in this thread that somehow Arkham City was compiled with some Pro settings, and that as a result it's running with the full Pro GPU/CPU. I assume that it was intentional, but I'm really unsure why.

I also feel that it would be very unwise to simply enable those Pro settings by default for the 1100 existing PS4 games.

Unrelated: You're being really combative in this thread for no particular reason,

How am I being combative? :s

That example with the other systems (along with the S) is to explain why the average person doesn't know what "breaking the game" would even be, and would otherwise not expect it to even be a thing until Cerny said so.

Maybe it's just that some games use fp16 more & the fact that PS4 Pro GPU is double rate fp16 give them a boost.

Please don't start this again...
 

andshrew

Member
It is because what I said is just as possible as what they said about the game running in Pro mode

What you're suggesting (and I'm not even sure what you're suggesting makes any sense at all) goes directly against how Sony have explained how non-Pro games work on the console. So no it isn't just as possible as the game simply being Pro enabled.

If you want 'proof' that the Batman game is legitimately a Pro enabled title someone who has it just needs to capture the URL to the JSON as I described here which will confirm how the game has been submitted to Sony (ie. neoEnable will be 1).
 

onQ123

Member
What you're suggesting (and I'm not even sure what you're suggesting makes any sense at all) goes directly against how Sony have explained how non-Pro games work on the console. So no it isn't just as possible as the game simply being Pro enabled.

If you want 'proof' that the Batman game is legitimately a Pro enabled title someone who has it just needs to capture the URL to the JSON as I described here which will confirm how the game has been submitted to Sony (ie. neoEnable will be 1).

I thought it was said that Pro enable games had to be 1080P? The post I responded to was about Moto Racer 4 which dark10x believes to be running in 900P.


That to me seems like a game that is just running better because of fp16 & not actually using PS4 Pro mode.
 
You're seriously only considering that because it suits your argument, where there is no indication that it would be the case. Lengthy testing for Arkham City to run at an uneven framerate? Hah.
No, I'm seriously considering it because I'm not assuming a specific scenario to be true. Whereas you have selected from the available choices, without any particular warrant for that selection being apparent.

For example, note that the absence of lengthy testing--even if we assume that--only invalidates one of the alternates I proposed.

It's besides the point anyway, it doesn't matter how it is done, it indicates that this game was just allowed to run with the extra hardware in mind with nothing else done.
But if it's easy and free, why doesn't the other game in the collection take the same advantage? Why doesn't every game released?

Clearly something else makes that unfeasible. So while I understand wanting this to be a simple, effective route to across-the-board improvements, I don't see how some people have convinced themselves that it is.

Note that I can't really argue with folks who say "Sure some games would break, do it anyway!" That's a personal preference.
 

mrklaw

MrArseFace
How come Sony cheaped out and didn't use Zen? Then everything would be 4K, 60 fps.

Age -> Sony -> done.

Because zen isn't ready for APUs yet? Because even if it was, the increased die size would mean reduced area for the GPU so you'd have less GPU power than the current pro? Because then it'd be disproportionately CPU heavy vs the current PS4 which would be a waste given the need to be BC and FC?
 

andshrew

Member
But if it's easy and free, why doesn't the other game in the collection take the same advantage? Why doesn't every game released?

Clearly something else makes that unfeasible. So while I understand wanting this to be a simple, effective route to across-the-board improvements, I don't see how some people have convinced themselves that it is.

Note that I can't really argue with folks who say "Sure some games would break, do it anyway!" That's a personal preference.

We can only speculate but given that this collection seems to have been through some development hell, maybe Asylum was finished much earlier and while they were finishing up City the Pro stuff became available so they threw it in.

Or maybe Asylum is configured in exactly the same way but because it is already capped to 30FPS (and I believe the base PS4 pretty much sticks to that?) there's essentially no improvement to see.
 
Top Bottom