By using this site, you agree to our Privacy Policy and our Terms of Use. Close

Forums - Microsoft Discussion - Starfield will be 30 fps on Xbox Series X and S.

Otter said:

Priority in a pipeline is orientated around opportunity cost, the reward of actually taking on a task.

There are probably 100 of things that will be ahead of a unstable 60fps mode. And again, you're not talking about a reliable 60fps mode, you're talking about a mode which at least hits 45fps for VRR users which is simply not valuable for the vast majority of Xbox gamers who do not have a VRR display. Such a mode will be at the bottom of the bucket list for a game like Starfield which in terms of testing will be magnitudes higher than probably all of the games you're seeing launch with performance modes. 

And I feel like people are throwing around "open world" comparisons a lot recently as if they're all made the same. Simply having a sandbox does not make a game incredibly complex. Beyond loading environments, A game like Horizon or Spiderman (for example) is probably closer to a single player title like TLOU compared to something like Starfield. They're not games where half the objects you come across have a physics component, can be picked up, stored, utililised, with 1000s of NPCs and their dynamic impact on the environment &  story being managed in real time. Its really apples and oranges. I won't give Bethesda a pass for launching games in a buggy state, but I also think people are doing the kind of games they make a disservice by comparing them to otherwise linear sandbox games or last gen/cross gen titles.

Opportunity cost is the cost associated with performing that task rather than another task, not "the reward of actually taking on a task." This is why I mentioned that pipelining in software development (including game development) can work to reduce the effective total time lost -- aka opportunity cost. Think of it like pipelining in a CPU. If you aren't using the resources for something else because there is latency from another resource finishing up its step in the process (such as developing the content to be QA'd), then that resource could do a different thing (such as QA different performance configs in already produced content.) 

See: https://cs.stanford.edu/people/eroberts/courses/soco/projects/risc/pipelining/index.html

Given that Starfield's development team will likely be doing a lot of the necessary work to make performance modes already (on PC's of comparable specs to the Series X and S) a lot of the work that needs to be done is already being done for other purposes anyway. It's not as if Starfield is an Xbox exclusive and not already running on other performance-tiers.I don't know where you got the idea that creating a performance profile is some long arduous task with a significant marginal cost, but that isn't true at all. The PC version will have consumers optimizing it within a week on a multitude of platforms to hit different targets. Pick the best config that works well on comparable hardware to the Series X and S, then use maybe a bit more time to actually test it on a dev-kit, likely will get better results than those PC players got because of the closed-platform, and then there you go. Yes, it will likely take months for every person in the pipeline to get their part done, but not months worth of dedicated full-time effort on the part of everyone involved. It would be something one would work on while one is waiting for other more critical things.  

As to your second point, much of the interactivity we've seen so far in Starfield isn't much of a massive improvement over what we've seen in past Bethesda games. But even if it were, we've seen games with a high degree of interactivity and open-worlds that have multiple performance modes, and have had a long list of performance updates. Red Dead Redemption 2 and Cyberpunk 2077 are very notable examples, where interactivity is on-par if not in excessive of most Bethesda games. If you want to make your point stronger, it'd help if you were more specific about what in Starfield you think would add complexity to the addition of a performance mode and the actual technical reasons for it. 

Last edited by sc94597 - on 14 June 2023

Around the Network
sc94597 said:
Otter said:

Priority in a pipeline is orientated around opportunity cost, the reward of actually taking on a task.

There are probably 100 of things that will be ahead of a unstable 60fps mode. And again, you're not talking about a reliable 60fps mode, you're talking about a mode which at least hits 45fps for VRR users which is simply not valuable for the vast majority of Xbox gamers who do not have a VRR display. Such a mode will be at the bottom of the bucket list for a game like Starfield which in terms of testing will be magnitudes higher than probably all of the games you're seeing launch with performance modes. 

And I feel like people are throwing around "open world" comparisons a lot recently as if they're all made the same. Simply having a sandbox does not make a game incredibly complex. Beyond loading environments, A game like Horizon or Spiderman (for example) is probably closer to a single player title like TLOU compared to something like Starfield. They're not games where half the objects you come across have a physics component, can be picked up, stored, utililised, with 1000s of NPCs and their dynamic impact on the environment &  story being managed in real time. Its really apples and oranges. I won't give Bethesda a pass for launching games in a buggy state, but I also think people are doing the kind of games they make a disservice by comparing them to otherwise linear sandbox games or last gen/cross gen titles.

Opportunity cost is the cost associated with performing that task rather than another task, not "the reward of actually taking on a task." This is why I mentioned that pipelining in software development (including game development) can work to reduce the effective total time lost -- aka opportunity cost. Think of it like pipelining in a CPU. If you aren't using the resources for something else because there is latency from another resource finishing up its step in the process (such as developing the content to be QA'd), then that resource could do a different thing (such as QA different performance configs in already produced content.) 

See: https://cs.stanford.edu/people/eroberts/courses/soco/projects/risc/pipelining/index.html

Given that Starfield's development team will likely be doing a lot of the necessary work to make performance modes already (on PC's of comparable specs to the Series X and S) a lot of the work that needs to be done is already being done for other purposes anyway. It's not as if Starfield is an Xbox exclusive and not already running on other performance-tiers.I don't know where you got the idea that creating a performance profile is some long arduous task with a significant marginal cost, but that isn't true at all. The PC version will have consumers optimizing it within a week on a multitude of platforms to hit different targets. Pick the best config that works well on comparable hardware to the Series X and S, then use maybe a bit more time to actually test it on a dev-kit, likely will get better results than those PC players got because of the closed-platform, and then there you go. Yes, it will likely take months for every person in the pipeline to get their part done, but not months worth of dedicated full-time effort on the part of everyone involved. It would be something one would work on while one is waiting for other more critical things.  

As to your second point, much of the interactivity we've seen so far in Starfield isn't much of a massive improvement over what we've seen in past Bethesda games. But even if it were, we've seen games with a high degree of interactivity and open-worlds that have multiple performance modes, and have had a long list of performance updates. Red Dead Redemption 2 and Cyberpunk 2077 are very notable examples, where interactivity is on-par if not in excessive of most Bethesda games. If you want to make your point stronger, it'd help if you were more specific about what in Starfield you think would add complexity to the addition of a performance mode and the actual technical reasons for it. 

Right, and when I mention that their will be way more aforementioned tasks that is exactly what I'm talking about.

"There are probably 100 of things that will be ahead of a unstable 60fps mode"

Testing a VRR mode on Xbox will be take time from things considered higher priority given the low reward of said mode. Critical bug/performance fixes and optimisation which is given at launch, followed by probably a haste to create & test further content. I don't see it being high in their pipeline. If you were talking about a mode relevant to more than 1% of Xbx user's, this could be different. 

Cyperpunk literally sunk CD Project red's stock price for a whole year. Simply offering performances patches is not the goal, especially if said patches get bad responses & there is no way to alter it on the player side, which is one of the reasons why console profiling & testing on the console side is way more dedicated then on PC. Cyperpunk is also like a  30hour game experience. You cannot build your own in-game forts, mass collect trickers/items that accomulate in a persistent environment, have those items be stolen & taken by NPCs, alter gravity, interact with a wider & dynamic ecosystem.

RD redemption is a last gen game, launched in 2018. When did it receive a performance mode on console? If we're taking about 4k modes on PS4 Pro & X1X, updating & supporting to 4k was massively more important from a consumer & marteking perspective than a VRR mode & also likely less CPU bound. Anytime your addressing a CPU limitation you're typically looking at more critical impacts on the game which require more testing.

And this is not me saying it'll never happen, but it'll probably not happen as a casual lauch window patch and is probably the least of their worries at the moment.

Last edited by Otter - on 14 June 2023

Otter said:

Right, and when I mention that their will be way more aforementioned tasks that is exactly what I'm talking about.

"There are probably 100 of things that will be ahead of a unstable 60fps mode"

Testing a VRR mode on Xbox will be take time from things considered higher priority given the low reward of said mode. Critical bug/performance fixes and optimisation which is given at launch, followed by probably a haste to create & test further content. I don't see it being high in their pipeline. If you were talking about a mode relevant to more than 1% of Xbx user's, this could be different. 

You original statement was:

"Any additional performance modes probably take months of testing particularly for a game of this scale. They're not going to spend 1000s of man hours on a performance mode that will be shat on in gaming media for being unplayable, with the only saving grace being that the 1% of Xbox Series Owners with VRR displays can benefit. Especially considering the type of game it is, post launch content & bug fixing will be priority over a 60fps mode."

1. It won't take 1000s of man-hours. I'd be surprised if it took 100 man-hours. Why? Because A. They've already likely been doing continuous performance tests. That is part of game development. They know what likely has to change to get a 45 fps minimum. That reduces the number of iterations between development staff and testers. B. It really doesn't take 1000's of hours to change and test config settings, even if they didn't already have some active performance testing in place already. 

2. Post launch bug-fixing definitely would be something a QA team would be involved with, but post-launch content takes time to develop, and it could be months before a QA team even sees that content. There definitely will be times, as there is in any software development, where the QA team have more capacity to take on small tasks -- like adding a performance mode, while they wait for the development team to develop new content to test. 

Otter said:

Cyperpunk literally sunk CD Project red's stock price for a whole year. Simply offering performances patches is not the goal, especially if said patches get bad responses & there is no way to alter it on the player side, which is one of the reasons why console profiling & testing on the console side is way more dedicated then on PC. Cyperpunk is also like a  30hour game experience. You cannot build your own in-game forts, mass collect trickers/items that accomulate in a persistent environment, have those items be stolen & taken by NPCs, alter gravity, interact with a wider & dynamic ecosystem.

Imagine if CD Project didn't improve performance and the game remained unplayable on many SKU's until this day. Yes, their short-term profitability might've been better in that scenario, but their credibility as developers would be trashed. I don't know what you're referring to with "bad responses" here. The game released to a "bad response" and was improved over time. 

Being a "30 hour game experience" really tells us nothing about how hard it is it make different performance configs. Cyberpunk 2077, unlike Starfield, pushed graphics in a way that made finding a config more difficult (leaning on ray-tracing which many platforms don't quite support well), not less.

As for the other interactivity situations you describe, all of these would've been tested (if Bethesda is doing its due-diligence) in the development of the game, as the game isn't just targeting one platform. They should already have a good idea how the game scales with different hardware and what it would take to achieve 45-60 fps. 

Otter said:

RD redemption is a last gen game, launched in 2018. When did it receive a performance mode on console? If we're taking about 4k modes on PS4 Pro & X1X, updating & supporting to 4k was massively more important from a consumer & marteking perspective than a VRR mode & also likely less CPU bound. Anytime your addressing a CPU limitation you're typically looking at more critical impacts on the game which require more testing.

Not an official patch, but one developed by a single modder. 

https://screenrant.com/red-dead-redemption-2-mod-ps4-pro-4k/

"As reported by Ben T. on Twitter, modder illusion0001 has managed to push Red Dead Redemption 2 beyond its limits on the PlayStation 4 Pro, allowing the title to smoothly run around 60 frames per second and in native 4K. Ben claims that the unofficial patch unlocks the frame rate for the PlayStation Pro version, but due to the console having a weak CPU the frames hang in the 50s range and can't keep a steady 60. Additionally, another patch by the modder boosts the resolution beyond 1080p and allows native 4K with the exclusion of blurry textures."

Which substantiates the point that this isn't something that takes 1000's of man-hours.



sc94597 said:
Otter said:

Right, and when I mention that their will be way more aforementioned tasks that is exactly what I'm talking about.

"There are probably 100 of things that will be ahead of a unstable 60fps mode"

Testing a VRR mode on Xbox will be take time from things considered higher priority given the low reward of said mode. Critical bug/performance fixes and optimisation which is given at launch, followed by probably a haste to create & test further content. I don't see it being high in their pipeline. If you were talking about a mode relevant to more than 1% of Xbx user's, this could be different. 

You original statement was:

"Any additional performance modes probably take months of testing particularly for a game of this scale. They're not going to spend 1000s of man hours on a performance mode that will be shat on in gaming media for being unplayable, with the only saving grace being that the 1% of Xbox Series Owners with VRR displays can benefit. Especially considering the type of game it is, post launch content & bug fixing will be priority over a 60fps mode."

1. It won't take 1000s of man-hours. I'd be surprised if it took 100 man-hours. Why? Because A. They've already likely been doing continuous performance tests. That is part of game development. They know what likely has to change to get a 45 fps minimum. That reduces the number of iterations between development staff and testers. B. It really doesn't take 1000's of hours to change and test config settings, even if they didn't already have some active performance testing in place already. 

2. Post launch bug-fixing definitely would be something a QA team would be involved with, but post-launch content takes time to develop, and it could be months before a QA team even sees that content. There definitely will be times, as there is in any software development, where the QA team have more capacity to take on small tasks -- like adding a performance mode, while they wait for the development team to develop new content to test. 

Otter said:

Cyperpunk literally sunk CD Project red's stock price for a whole year. Simply offering performances patches is not the goal, especially if said patches get bad responses & there is no way to alter it on the player side, which is one of the reasons why console profiling & testing on the console side is way more dedicated then on PC. Cyperpunk is also like a  30hour game experience. You cannot build your own in-game forts, mass collect trickers/items that accomulate in a persistent environment, have those items be stolen & taken by NPCs, alter gravity, interact with a wider & dynamic ecosystem.

Imagine if CD Project didn't improve performance and the game remained unplayable on many SKU's until this day. Yes, their short-term profitability might've been better in that scenario, but their credibility as developers would be trashed. I don't know what you're referring to with "bad responses" here. The game released to a "bad response" and was improved over time. 

Being a "30 hour game experience" really tells us nothing about how hard it is it make different performance configs. Cyberpunk 2077, unlike Starfield, pushed graphics in a way that made finding a config more difficult (leaning on ray-tracing which many platforms don't quite support well), not less.

As for the other interactivity situations you describe, all of these would've been tested (if Bethesda is doing its due-diligence) in the development of the game, as the game isn't just targeting one platform. They should already have a good idea how the game scales with different hardware and what it would take to achieve 45-60 fps. 

Otter said:

RD redemption is a last gen game, launched in 2018. When did it receive a performance mode on console? If we're taking about 4k modes on PS4 Pro & X1X, updating & supporting to 4k was massively more important from a consumer & marteking perspective than a VRR mode & also likely less CPU bound. Anytime your addressing a CPU limitation you're typically looking at more critical impacts on the game which require more testing.

Not an official patch, but one developed by a single modder. 

https://screenrant.com/red-dead-redemption-2-mod-ps4-pro-4k/

"As reported by Ben T. on Twitter, modder illusion0001 has managed to push Red Dead Redemption 2 beyond its limits on the PlayStation 4 Pro, allowing the title to smoothly run around 60 frames per second and in native 4K. Ben claims that the unofficial patch unlocks the frame rate for the PlayStation Pro version, but due to the console having a weak CPU the frames hang in the 50s range and can't keep a steady 60. Additionally, another patch by the modder boosts the resolution beyond 1080p and allows native 4K with the exclusion of blurry textures."

Which substantiates the point that this isn't something that takes 1000's of man-hours.

A modder making something as a proof of concept is completely different from testing the game top to bottom making sure it works everywhere, no severe dips anywhere and stable throughout whatever you do in game. That's what takes 1000's of man-hours. Unlocking the frame rate or resolution is the easy part. He didn't optimize the engine to get a stable 60fps from start to finish, that's where the real work is.



sc94597 said:
Otter said:

Right, and when I mention that their will be way more aforementioned tasks that is exactly what I'm talking about.

"There are probably 100 of things that will be ahead of a unstable 60fps mode"

Testing a VRR mode on Xbox will be take time from things considered higher priority given the low reward of said mode. Critical bug/performance fixes and optimisation which is given at launch, followed by probably a haste to create & test further content. I don't see it being high in their pipeline. If you were talking about a mode relevant to more than 1% of Xbx user's, this could be different. 

You original statement was:

"Any additional performance modes probably take months of testing particularly for a game of this scale. They're not going to spend 1000s of man hours on a performance mode that will be shat on in gaming media for being unplayable, with the only saving grace being that the 1% of Xbox Series Owners with VRR displays can benefit. Especially considering the type of game it is, post launch content & bug fixing will be priority over a 60fps mode."

1. It won't take 1000s of man-hours. I'd be surprised if it took 100 man-hours. Why? Because A. They've already likely been doing continuous performance tests. That is part of game development. They know what likely has to change to get a 45 fps minimum. That reduces the number of iterations between development staff and testers. B. It really doesn't take 1000's of hours to change and test config settings, even if they didn't already have some active performance testing in place already. 

2. Post launch bug-fixing definitely would be something a QA team would be involved with, but post-launch content takes time to develop, and it could be months before a QA team even sees that content. There definitely will be times, as there is in any software development, where the QA team have more capacity to take on small tasks -- like adding a performance mode, while they wait for the development team to develop new content to test. 

Otter said:

Cyperpunk literally sunk CD Project red's stock price for a whole year. Simply offering performances patches is not the goal, especially if said patches get bad responses & there is no way to alter it on the player side, which is one of the reasons why console profiling & testing on the console side is way more dedicated then on PC. Cyperpunk is also like a  30hour game experience. You cannot build your own in-game forts, mass collect trickers/items that accomulate in a persistent environment, have those items be stolen & taken by NPCs, alter gravity, interact with a wider & dynamic ecosystem.

Imagine if CD Project didn't improve performance and the game remained unplayable on many SKU's until this day. Yes, their short-term profitability might've been better in that scenario, but their credibility as developers would be trashed. I don't know what you're referring to with "bad responses" here. The game released to a "bad response" and was improved over time. 

Being a "30 hour game experience" really tells us nothing about how hard it is it make different performance configs. Cyberpunk 2077, unlike Starfield, pushed graphics in a way that made finding a config more difficult (leaning on ray-tracing which many platforms don't quite support well), not less.

As for the other interactivity situations you describe, all of these would've been tested (if Bethesda is doing its due-diligence) in the development of the game, as the game isn't just targeting one platform. They should already have a good idea how the game scales with different hardware and what it would take to achieve 45-60 fps. 

Otter said:

RD redemption is a last gen game, launched in 2018. When did it receive a performance mode on console? If we're taking about 4k modes on PS4 Pro & X1X, updating & supporting to 4k was massively more important from a consumer & marteking perspective than a VRR mode & also likely less CPU bound. Anytime your addressing a CPU limitation you're typically looking at more critical impacts on the game which require more testing.

Not an official patch, but one developed by a single modder. 

https://screenrant.com/red-dead-redemption-2-mod-ps4-pro-4k/

"As reported by Ben T. on Twitter, modder illusion0001 has managed to push Red Dead Redemption 2 beyond its limits on the PlayStation 4 Pro, allowing the title to smoothly run around 60 frames per second and in native 4K. Ben claims that the unofficial patch unlocks the frame rate for the PlayStation Pro version, but due to the console having a weak CPU the frames hang in the 50s range and can't keep a steady 60. Additionally, another patch by the modder boosts the resolution beyond 1080p and allows native 4K with the exclusion of blurry textures."

Which substantiates the point that this isn't something that takes 1000's of man-hours.

The game launched in an aweful state, what use is all these performance modes if not one of them is well optimised? The monthly patches didn't get the game in a decent state until about a year, so instead of worrying about increasing the number of modes and optimising for them, prioritising one makes more sense if performance is a concern. For that reason you using Cyperpunk is counter productive to your argument.

1. Given the performance profile of the Xbox version is currently 30fps with drops they likely do not have a full config for how it can maintain a solid 45fps throughout the experience on said hardware, 45 has likely never been a target because it is not a desirable frame rate. They're not going to whack on a low end PC profile onto Xbox Series X & S and call it a day without a thorough round of retesting.

We've already seen with plenty of games which require work to reach higher FPS that they are not next day patches. Plagues tale launched with 40 but took an extra  6months to get it up to 60fps. The logic you're working with is that the game is easily scalable for low-end CPUs when it being capped with drops below 30 suggest otherwise.

2. Early post game content can be approaching finalisation before a game is even launched but even in scenarios where it is not, testing is continuous throughout development & post game fixes in this kind of game will be a year long endeavour. Again it feels like you're not getting the point that it's not that it being impossible to do, but simply it is not something they will put high up on the list. It took Geurilla games a whole year to get around to fixing the AA in the 60fps mode in Horizon. There are often vital patches which don't get addressed months after launch.

3. The length and complexity of game interactions absolutely impacts the amount of QA needed which is what additional performance modes is competing for resources against. 

Especially for 1st party titles console performance profiles are often bespoke & there are not always 1:1 matches with available PC settings.

Lastly a single modder unlocking a frame rate and calling it a day is not the same due diligence a developer will take before putting out a patch out on millions of systems. It harkens back to Sony originally underplaying PS5's BC, for QA purposes they take a formalised route which is where the man hours come in. And again, configs for PC aren't going to just get thrown onto Xbox and ticked off as a new mode.

I would bet money on it not happening anytime soon after launch and the reason will be exactly as I'm saying, lack of priority

Last edited by Otter - on 14 June 2023

Around the Network

No suprise for a console versions



I'm hoping they will change their mind and offer 40 fps mode at some point after launch.

Surely CPU can handle 40 fps, that's 20 fps less than 60 fps there must be enough headroom from 30 fps.



Leynos said:
Mnementh said:

If you start from scratch, your product will be more outdated, because implementing all the stuff seen as modern takes time. So I think you wish they switch to another engine like Unreal or id tech. But chances are they would struggle with the big seemless open worlds that Bethesda does since ages ago. I am not convinced at all, that switching the engine solves the performance issues.

 Listening to their QA team for once might.

Yeah, I translate that comment into: putting focus (which means money) on quality control and polish. And yeah, that indeed would help. Not sure if they could squeeze out more fps with the graphics and world size they want to do, but certainly they could provide a more stable experience. As it is Bethesda I kinda expect some glitches and hiccups at launch.

Last edited by Mnementh - on 15 June 2023

3DS-FC: 4511-1768-7903 (Mii-Name: Mnementh), Nintendo-Network-ID: Mnementh, Switch: SW-7706-3819-9381 (Mnementh)

my greatest games: 2017, 2018, 2019, 2020, 2021, 2022, 2023

10 years greatest game event!

bets: [peak year] [+], [1], [2], [3], [4]



...to avoid getting banned for inactivity, I may have to resort to comments that are of a lower overall quality and or beneath my moral standards.

I think we are gonna start seeing 30fps games again on current gen. final fantasy 16 already can't hold 60fps, and star wars looks like its doing too much graphically to be 60fps. hope we get a ps5 pro soon and i will sell my ps5.