It’s embarrassing that this kind of talk is still the norm surrounding Unreal Engine in gaming forums.
Have you two thought about just how difficult it is to maintain a game based on an engine that is now 22 years old? Imagine hiring new developers who taught themselves or already have worked in other studios on UE4 and 5, but now have to go back to a significantly less advanced engine with far more archaic tools. The studio has to move on at some point.
Also, modern versions of UE can be incredibly performant - it entirely depends on the features you’re using. Going full-out on Nanite for example is demanding, but it’s supposed to be a high-end feature pushing the envelope on the latest hardware. You’re not forced to use it as a developer and I doubt they will for an arena-based sports title. Similarly, Lumen is entirely optional as well.
Yes, it is possible to make performant products in UE5. Nobody bothers, though. Far easier to just turn all the magic buttons on that Make Game Look Good™ and then shunt the downsides onto the consumers hardware. I want to be proven wrong, desperately - I would feel nothing but joy if the ue6 rocket league was equally performant compared to it’s current state. I’ll be overjoyed if the Witcher 4 doesn’t require AI upscaling to run at 60fps on the latest hardware because perhaps cdpr devs can give a damn about their optimisation despite using ue5.
While I agree that there are games that - mostly for a lack of time in an industry that puts ever increasing pressure on developers - are having hardware requirements that are higher than they need to be, the whole picture is a bit more nuanced.
DLSS, FSR 4.1 and (to a lesser degree) Xess are good enough that you’re just wasting performance (and even, in some games, image quality) for no reason by not activating them. They are an inherent part of modern rendering pipelines, not mere crutches, a way out of the fact that every tiny improvement to visual fidelity comes with a huge cost to performance. If you have hardware that is still limited to older versions of FSR - which are kind of shitty - then I can understand your sentiment, but if you’re not, then I invite you to have a friend help you with a blind test, activating and deactivating current upscalers to find out for yourself if you can spot the difference. You might be surprised.
Also, if you had paid attention to what CDPR developers have published and presented so far, they have optimized UE significantly, dramatically improving performance, both by rewriting parts of the engine entirely (like the geometry streaming system) and throwing out others that they have no need for (like Blueprints). Expect The Witcher 4 to have hardware requirements that are appropriate for the visuals. Note that I’m not saying it will have low hardware requirements - there’s a difference. Given the target audience, the expectations and the game’s status as a halo game for both CDPR and UE, this game will push technology forward, which is not cheap.
That said, this game has been announced for the current console generation, which includes Xbox Series S, so it’ll still have to scale. Expect fallbacks (like possibly software ray-tracing, since UE supports this feature - I doubt it’ll allow you to disable RT entirely) and plenty of options to dial it back to run on reasonably modest hardware. It’s likely going to be too much for the Steam Deck, but still run on e.g. an RTX 2080, a first-gen RT-capable card with 8GB of VRAM from eight years ago (ebay price: ~200 bucks), which in my experience can handle the most demanding current RT-only games remarkably well.
Do not expect the upgraded version of Rocket League to have hardware requirements as low as the current version. That game officially only needs a GeForce 760, which is an upper entry level card from thirteen years ago that is limited to DirectX feature level 11 and has only 2 GB of VRAM. I honestly doubt that many players are still playing the game on hardware this old (and the developers have the numbers). What we can reasonably hope for, just like with The Witcher 4, is that the new version has hardware requirements that are appropriate for the visuals. I think it’s realistic to expect them to be similar to Fortnite, given the developer being part of Epic, scaling on a wide variety of systems. It’s still a F2P e-sports title after all, so it will have to remain accessible to a large number of players on hardware ranging from toaster to space ship.
I absolutely can tell when DLSS is on vs off. The swimming artifacts are blatantly obvious. I think it is a fantastic technology, I do not agree with the direction of AAA games making non native resolution rendering part of their performance benchmark. Seems to be an issue here and there broadly in AAA games but it seems to be the default for UE5 devs. Also here we have the first fuck you for simply assuming I’m too out of the loop rather than just having an informed, differing, opinion.
I brought attention to TW4 because I’ve seen the examples of optimisation cdpr are doing. Here is the second fuck you for simply assuming I’m out of the loop rather than having an informed opinion. If the Witcher 4 has hardware requirements that are appropriate for the visuals, then it will be the first blockbuster AAA in several years to do so, and I will celebrate their success and be glad to continue enjoying their art.
“I doubt it’ll allow you to disable RT entirely” is a huge problem. Rasterized lighting looks about 80-90% of the way there compared to full RT/PT with all the fancy pbr textures and all, but has a performance cost which is dwarfed in comparison. Offloading what could have been a dev’s handiwork onto the customer’s hardware willy nilly is part of the bigger problem. And we aren’t seeing cheaper games as a result of the corner cutting, despite the extra burden on consumers.
My experience doesn’t match yours wrt the 2080 being capable of handling current day rt only titles well, quite the opposite in the case of my friends with 2080s, but thats anecdotal from both sides so its moot.
I don’t expect RL to run as well on ue6 as it does currently. Obviously. That’s why I said it. If they do some big visual overhaul then we can’t compare apples to apples, sadly. But if it is just an engine swap without investing in PBR materials and stuff like that, then why would it be reasonable to expect UE6, a newer, hopefully more refined engine, to run worse than UE3?
Anyway. Despite your assumptions about me, we are both clearly informed on the topic(s) and have drawn differing conclusions. If you’re going to continue to engage with me with the same lack of good faith as your first reply, I’d kindly request that you simply do not. Have a good day.
To be clear, “here"s the Nth fuck you” was me referring to the fact that you insulted my intelligence. I was not saying fuck you to you, but pointing out that you had implicitly said it to me.
Because they would have to convince the existing player base to move to the new game, which is risky. By making it an update, you get most of them to continue playing the game.
Yeah, but what else can they do? There’s a point where supporting an ancient engine and code base are becoming nothing but a headache and I suspect they are long past this point. Like I said elsewhere in this thread, UE3 originally released in 2004. Most players of this game were likely not even born yet by that year. Even the last stable build is eleven years old now.
You don’t want two versions of the game either. This splits the player base in two and you still have to provide support for the old version. Sooner rather than later, this limb needs to be cut off. It’s the least terrible option.
I mean, it’s an arena-based game with a small handful of mostly static car models pushing around a ball. Even in the announcement trailer, the new car models aren’t exactly looking like they are made out of a billion polygons. This isn’t an open world game either, so traversal stutter is out of the question already. Load in the entire arena and done. They can get away with static GI too.
Hardware requirements will be higher, but since this will have to run on Switch 2 and your run of the mill laptop with iGPUs, they won’t be outrageous. I’ll eat my hat if it doesn’t run on the MacBook Neo with an old iPhone chip and 8 GB of total RAM. The overlap between the target audiences of this laptop and this game might as well be a circle.
Is RL still on Unreal 3? I know it was when it launched, but I hope this doesn’t completely break compatibility with the existing Steam Workshop content
Yeah, Psyonix actually did a reasonable job still supporting it for folks who bought it on Steam. If Sweeney didn’t have a stick so far up his ass about the value of other stores, it probably would continue too
Fortnite went from Unreal 4 to 5. They’re quite similar systems but Unreal changed pretty fundamentally between 3 and 4, so this is a big leap. We’re talking about an engine that was built in 2006. I’m sure Psyonix has the internal support required to port it, but there’s a lot more than a visual upgrade involved.
You are not logged in. However you can subscribe from another Fediverse account, for example Lemmy or Mastodon. To do this, paste the following into the search field of your instance: [email protected]
No game suggestions, friend requests, surveys, or begging.
No Let’s Plays, streams, highlight reels/montages, random videos or shorts.
No off-topic posts/comments, within reason.
Use the original source, no clickbait titles, no duplicates.
(Submissions should be from the original source if possible, unless from paywalled or non-english sources.
If the title is clickbait or lacks context you may lightly edit the title.)
I’m guessing the graphics will become marginally better, while the performance gets massively worse
Worse performance is always something I’m happy to see in my competitive online games!
It’s embarrassing that this kind of talk is still the norm surrounding Unreal Engine in gaming forums.
Have you two thought about just how difficult it is to maintain a game based on an engine that is now 22 years old? Imagine hiring new developers who taught themselves or already have worked in other studios on UE4 and 5, but now have to go back to a significantly less advanced engine with far more archaic tools. The studio has to move on at some point.
Also, modern versions of UE can be incredibly performant - it entirely depends on the features you’re using. Going full-out on Nanite for example is demanding, but it’s supposed to be a high-end feature pushing the envelope on the latest hardware. You’re not forced to use it as a developer and I doubt they will for an arena-based sports title. Similarly, Lumen is entirely optional as well.
Yes, it is possible to make performant products in UE5. Nobody bothers, though. Far easier to just turn all the magic buttons on that Make Game Look Good™ and then shunt the downsides onto the consumers hardware. I want to be proven wrong, desperately - I would feel nothing but joy if the ue6 rocket league was equally performant compared to it’s current state. I’ll be overjoyed if the Witcher 4 doesn’t require AI upscaling to run at 60fps on the latest hardware because perhaps cdpr devs can give a damn about their optimisation despite using ue5.
While I agree that there are games that - mostly for a lack of time in an industry that puts ever increasing pressure on developers - are having hardware requirements that are higher than they need to be, the whole picture is a bit more nuanced.
DLSS, FSR 4.1 and (to a lesser degree) Xess are good enough that you’re just wasting performance (and even, in some games, image quality) for no reason by not activating them. They are an inherent part of modern rendering pipelines, not mere crutches, a way out of the fact that every tiny improvement to visual fidelity comes with a huge cost to performance. If you have hardware that is still limited to older versions of FSR - which are kind of shitty - then I can understand your sentiment, but if you’re not, then I invite you to have a friend help you with a blind test, activating and deactivating current upscalers to find out for yourself if you can spot the difference. You might be surprised.
Also, if you had paid attention to what CDPR developers have published and presented so far, they have optimized UE significantly, dramatically improving performance, both by rewriting parts of the engine entirely (like the geometry streaming system) and throwing out others that they have no need for (like Blueprints). Expect The Witcher 4 to have hardware requirements that are appropriate for the visuals. Note that I’m not saying it will have low hardware requirements - there’s a difference. Given the target audience, the expectations and the game’s status as a halo game for both CDPR and UE, this game will push technology forward, which is not cheap.
That said, this game has been announced for the current console generation, which includes Xbox Series S, so it’ll still have to scale. Expect fallbacks (like possibly software ray-tracing, since UE supports this feature - I doubt it’ll allow you to disable RT entirely) and plenty of options to dial it back to run on reasonably modest hardware. It’s likely going to be too much for the Steam Deck, but still run on e.g. an RTX 2080, a first-gen RT-capable card with 8GB of VRAM from eight years ago (ebay price: ~200 bucks), which in my experience can handle the most demanding current RT-only games remarkably well.
Do not expect the upgraded version of Rocket League to have hardware requirements as low as the current version. That game officially only needs a GeForce 760, which is an upper entry level card from thirteen years ago that is limited to DirectX feature level 11 and has only 2 GB of VRAM. I honestly doubt that many players are still playing the game on hardware this old (and the developers have the numbers). What we can reasonably hope for, just like with The Witcher 4, is that the new version has hardware requirements that are appropriate for the visuals. I think it’s realistic to expect them to be similar to Fortnite, given the developer being part of Epic, scaling on a wide variety of systems. It’s still a F2P e-sports title after all, so it will have to remain accessible to a large number of players on hardware ranging from toaster to space ship.
I absolutely can tell when DLSS is on vs off. The swimming artifacts are blatantly obvious. I think it is a fantastic technology, I do not agree with the direction of AAA games making non native resolution rendering part of their performance benchmark. Seems to be an issue here and there broadly in AAA games but it seems to be the default for UE5 devs. Also here we have the first fuck you for simply assuming I’m too out of the loop rather than just having an informed, differing, opinion.
I brought attention to TW4 because I’ve seen the examples of optimisation cdpr are doing. Here is the second fuck you for simply assuming I’m out of the loop rather than having an informed opinion. If the Witcher 4 has hardware requirements that are appropriate for the visuals, then it will be the first blockbuster AAA in several years to do so, and I will celebrate their success and be glad to continue enjoying their art.
“I doubt it’ll allow you to disable RT entirely” is a huge problem. Rasterized lighting looks about 80-90% of the way there compared to full RT/PT with all the fancy pbr textures and all, but has a performance cost which is dwarfed in comparison. Offloading what could have been a dev’s handiwork onto the customer’s hardware willy nilly is part of the bigger problem. And we aren’t seeing cheaper games as a result of the corner cutting, despite the extra burden on consumers.
My experience doesn’t match yours wrt the 2080 being capable of handling current day rt only titles well, quite the opposite in the case of my friends with 2080s, but thats anecdotal from both sides so its moot.
I don’t expect RL to run as well on ue6 as it does currently. Obviously. That’s why I said it. If they do some big visual overhaul then we can’t compare apples to apples, sadly. But if it is just an engine swap without investing in PBR materials and stuff like that, then why would it be reasonable to expect UE6, a newer, hopefully more refined engine, to run worse than UE3?
Anyway. Despite your assumptions about me, we are both clearly informed on the topic(s) and have drawn differing conclusions. If you’re going to continue to engage with me with the same lack of good faith as your first reply, I’d kindly request that you simply do not. Have a good day.
And I kindly request you to not insult me.
To be clear, “here"s the Nth fuck you” was me referring to the fact that you insulted my intelligence. I was not saying fuck you to you, but pointing out that you had implicitly said it to me.
I think some of it is just aging hardware too, a lot of problems are ironed out after launch by adding options and changing defaults.
I was mostly surprised to see the plan to port the game over at all! It does make for a good test of the new engine though
Minimum requirements: nvidia H100 for FullHD 30 fps.
Why not leave it as it is and create Rocket League 2 with the new engine?
Because they would have to convince the existing player base to move to the new game, which is risky. By making it an update, you get most of them to continue playing the game.
FTFY lol
Yeah, but what else can they do? There’s a point where supporting an ancient engine and code base are becoming nothing but a headache and I suspect they are long past this point. Like I said elsewhere in this thread, UE3 originally released in 2004. Most players of this game were likely not even born yet by that year. Even the last stable build is eleven years old now.
You don’t want two versions of the game either. This splits the player base in two and you still have to provide support for the old version. Sooner rather than later, this limb needs to be cut off. It’s the least terrible option.
for the sake of the existing player base, I just hope they don’t add tons of stutters and nuke the system requirements into orbit
I just don’t have a lot of hope for UE6 lol
I mean, it’s an arena-based game with a small handful of mostly static car models pushing around a ball. Even in the announcement trailer, the new car models aren’t exactly looking like they are made out of a billion polygons. This isn’t an open world game either, so traversal stutter is out of the question already. Load in the entire arena and done. They can get away with static GI too.
Hardware requirements will be higher, but since this will have to run on Switch 2 and your run of the mill laptop with iGPUs, they won’t be outrageous. I’ll eat my hat if it doesn’t run on the MacBook Neo with an old iPhone chip and 8 GB of total RAM. The overlap between the target audiences of this laptop and this game might as well be a circle.
Damn I thought UE5 like just came out wtf
Edit: Oh Tim says maybe perview versions in 1 to 2 years okay
Version 1.0 of UE5 was released over four years ago.
buh
Is RL still on Unreal 3? I know it was when it launched, but I hope this doesn’t completely break compatibility with the existing Steam Workshop content
The workshop still works? I would’ve expected them to drop it when epic bought them
Yeah, Psyonix actually did a reasonable job still supporting it for folks who bought it on Steam. If Sweeney didn’t have a stick so far up his ass about the value of other stores, it probably would continue too
Wow that’s nuts, I wonder if the new version keeps it 😬 or if it’ll be full exclusive
So far all updates made it to Steam. The engine update will likely be barely noticeable. Wasn’t when Fortnite did it.
Fortnite went from Unreal 4 to 5. They’re quite similar systems but Unreal changed pretty fundamentally between 3 and 4, so this is a big leap. We’re talking about an engine that was built in 2006. I’m sure Psyonix has the internal support required to port it, but there’s a lot more than a visual upgrade involved.
Oh dear god. I’ve not even heard of UE6.
More slop, I guess.