Run much better? UE5 is God awful for performance. Perhaps one of the biggest disasters in this generation of gaming. Why people want it... I don't know. I guess it's slightly more capable than ue4 in larger environments but you'd struggle to tell the difference when it's running sub 30fps most of the time.
Intrigued to know why people think UE5 would be a good thing. Matt already debunked the theory that you can just transfer it across in a previous roadmap and I’m pretty sure in this forum as well. He said all you would get is a UE4 game running under UE5 at a very high cost to transfer anything + they already customised UE4 so much that it wasn’t viable. I would see it as a very bad thing. It would likely be a new game, so no existing DLC and they haven’t got the game working properly in UE4 yet, so were going to see a whole new load of bugs on top of the current ones. I hope we never see TSW on UE5 until they master what we’ve currently got. I’m prepared to be wrong but I still think people are building themselves up for a huge disappointment. Expect nothing and then be pleasantly surprised if it’s something good.
Just want to chip in on this point (fascinating thread overall folks ) For the purposes of TSW it is not possible to transfer. We have a vast array of significant modifications to Unreal Engine that would require porting and re-doing to fit the new way UE5 works with some of its core systems. Some of them will come over relatively easily, some of the more fundamental ones would require quite significant rework. Aside from the core engine: - the fundamental way maps work in UE5 is different - if you port what you've got over (which is a deprecated, or discontinued, tech in UE5) then you don't gain any of the benefits the maps from going to UE5. - The way lighting works is fundamentally different using lumen, the rules and constraints are different and there are significant challenges to be solved there (lumen doesnt like things that move much because it resolves things over multiple frames), so you're redoing all the lighting and still resolving some major core issues. Lumen is what gives UE5 games "that look", if you go to UE5 and don't use Lumen, you may as well not bother. - The way models are built is fundamentally different, using Nanite. That's how you get super detailed meshes and no jumps in lods, lower ram usage because of no lods. Again, don't use Nanite and why are you bothering. Oh, and nanite doesn't support skeletal meshes, which is what all the trains are. UE5 train sims are certainly possible. All these things are fixable. But, while straight forward projects CAN just be ported from 4 to 5 with their import tool, there is nothing about TSW that is straight forward and if you want to actually get the benefits you need to design to use them and build content that uses them. If you just port it, it will just be "TSW UE4 in a UE5 engine" which, frankly, is a "marketing release" not a meaningful one. If / when we do UE5, personally, I want it to be a meaningful upgrade, or why bother, just so you can put "powered by UE5" logo on it? Matt.
There’s a whole lot that is better with UE5, the game could definitely look a lot better but UE5 has a lot more improvements like Lumen for dynamic lighting and reflections and Nanite that gives way more detail in things like train models and scenery without killing performance. The game could also support bigger maps with World Partition, that would let them load parts of the world quicker this could mean allowing mergers or even larger timetables with more AI trains running at once. Performance should improve too, because of better optimization and features like Temporal Super Resolution that help give more FPS without losing visual quality. It could also bring more realistic physics, better weather effects, improved sounds, and smarter passenger behavior. So a whole lot of improvements.
Thanks for chiming in Matt and certainly helps for a lot of insight i hope it’ll come to TSW as it will be a big upgrade but certainly has to be done right I agree.
On systems that aren’t built to handle that new engine performance of course you won’t get a good frame rate but the ps5 and xbox x are built to handle stuff like UE5 like in a performance mode that focuses more on better frames or quality mode for better graphics
Good because of all the benefits that come from it bad if one game gets abandoned for another and we loose everything they’ll have to get it right but if they do it’ll be a great benefit for the game.
What? UE5 is infamous for running poorly on current gen consoles and top end pcs alike. Dropping internal resolution to below 1080p in order to get a stable framerate does not constitute a reasonable performance mode especially on hardware that lacks sufficient AI upscaling capability (like the ps5 and xbox series). I have a long list of digital foundry videos you can peruse to see the depths of trouble UE5 causes developers. Look at Stalker 2, most of the AI systems had to be ripped out to get that thing running between 40 and 50fps. So a demonstrably less performant engine, where you'd have to retool all your assets with nanite and lumen would have to be cut to achieve any kind of reasonable performance. Dovetail would be better off making their own engine at that point, they basically already have with the customisations to ue4.
UE5 (now at version 5.2) still suffers from stuttering. There are some very nice videos on YouTube about this engine, however a video (even and rightly produced for propaganda) is a video while a game, perhaps on a not too recent system, is something else entirely. Even a very high-end notebook (to make the most of UE, assuming it has all these possibilities) is a problem. The cost is around 7000.00 euros. Not exactly small change.
UE5 isn’t as bad as people think it just comes down to how developers use it. Look at games like Fortnite and the Matrix Awakens demo. Both run on current-gen consoles with Nanite and Lumen, and they run fine when optimized. The real problem is developers not taking the time to properly optimize for consoles. Take STALKER 2, for example yeah, it’s struggling, but that’s not UE5’s fault. They’re pushing raw assets without adjusting them for performance. Dropping resolution below 1080p isn’t great, but that’s exactly why Temporal Super Resolution (TSR) exists. It upscales lower resolutions to look sharp while keeping framerates stable. If a game runs poorly, it’s usually a developer issue, not the engine. And saying Dovetail should “just make their own engine” is wild. Building a custom engine is insanely expensive and takes years to develop. No matter how much they tweak UE4, it’s still an aging engine. UE5 brings World Partition for huge maps, Chaos Physics for better interactions, and faster tools that save dev time. At the end of the day, UE5 has way more potential than UE4 ever did. Games struggling right now are just figuring it out it’s part of the process. Blaming the engine instead of focusing on proper optimization isn’t it.
[QUOTE="jay#4862, post: 943292, Blaming the engine instead of focusing on proper optimization isn’t it.[/QUOTE] Exactly, but how much time does all this cost developers? If the graphics engine is too complex to optimize, it means that it is not a very good graphics engine. Unless then a video game is sold at a stratospheric price.
Not sure what you mean exactly, but I think you’re saying that videos showcasing UE5 look great, but that’s just marketing actually running it in a real game on older systems or even high-end hardware is something else entirely. And yea if you need a 7,000 euro setup to “make the most” out of UE5, it’s not practical for most people. The thing is, though, a lot of those videos are pushing max settings and showing off the engine’s full potential. No one expects every game to run like that on older systems. UE5 has tools like Nanite and TSR that let developers scale performance down without losing much visual quality. It’s more about how they optimize it. Still, I get your point just because the tech can look amazing doesn’t mean it will run smoothly on everything.
I get where you’re coming from, but I don’t think the complexity of optimization makes UE5 a “bad engine.” Any powerful engine is going to come with a learning curve, especially one with features like Nanite and Lumen. The tools are there to scale performance, but developers have to actually use them properly. It’s true that optimization takes time, but that’s part of modern game development. You can’t expect super clean visuals without putting in the work to make it run well. The payoff is massive though games that use UE5 properly look and feel better than what most other engines can pull off right now. And about price no one’s selling games at some crazy “stratospheric” price. Games are still launching at the same $60–$70 range. If developers put the effort in, UE5 gives them the tools to make something worth that price tag. It’s just about adapting to the new tech. (ps: DTG do i get a job? lol)
I guess it will be something from the list below(from unlikely to most likely): - TSW∞. The game will no longer receive new numbers in its title and will be updated annually with major core updates and a Rush Hour-level expansion packs. - Multiplayer Beta. - US iconic freight route. - New country. - Just some new DB or British route(perhaps even from the list of the most anticipated) with new rolling stock. - Class 66.
Actually, a clue was given by DTG. I think the sentence was: "something we've wanted to do/announce for a long time..." What could have been in the works for a long time? I think it's a special feature, for the game in general.
For me "something we've wanted to do/announce for a long time..." sounds like multiplayer. Or just a new DLC but special(US main freight, Shinkansen, DB dostos).
Agreed on that one. Something they have wanted to do / announce for a “long time”. DTG are finally able to do the NYC Subway in TSW ??. Now that would be a reveal and a half !!!.
I have no idea. The only thing I am sure of (99%) is that it will be a feature for all users. Nothing exclusive to one of the three countries mainly represented (DE, UK, USA).
NYC Subway would certainly be a good add on but why weren’t they able to do it, because NYC is such a big city maybe?
Another country? (Please be Italy or Australian) Steam? Free DLC? (no chance) TSW 6? Fred (not Jay) gets his 12 miles free? (No seriously, an extension to a DLC, like ECML to Kings Cross/ York) Big sale? Route merging? (Not “route hopping”) Signalling mode? (become a signaller) The game getting upgraded to the newest Unreal Engine 5 (5 for TSW 5….. or 6) Physical train controllers? (like rail driver) Route creation to console (over my dead body will they) Cross over between the fishings sims and TSW (imagine fishing at a pier and watching the trains go bye) Fortnite skin? (Jokes, Jokes) An advanced subscription (individual routes to gamepass) Continuation of the Rush Hour routes? A Museum modeled? (NRM or that one in London) Something to do with mobile? (companion app that tracks your progress) Mastery actually working? (Never) Port to Nintendo/ a handheld/ phone/ tablet ? (Could it actually be done) NPCs updated/ variation/ regionalised. anyways placing my bets here
one of these has to be right lol and don’t forget UE5 there’s still a chance, but what 12 miles are you talking about exactly?
Out of your list I'm going to have to go with Fortnite, clearly it's the only sensible option on the list, my bet is definitely going to make me a millionaire, I'm sure of it.......
At the start of this thread you said that the north London line should be free which is a 12 (don’t quote me) mile long route. A 12 mile extension to any route would be amazing. Nothing personal by the way, just a reference.
I hope so too. I really want more MTA content, especially the Metro North Hudson and New Haven lines to appear in TSW
I reckon something like a signalman, especially considering that some of the new routes (looking at you JT) have modeled boxes.
What about updated NPCs. It would be cool to see them regionalised (Japanese, American, Brittanians, Aus Bogans and eshays bah) Time to add this to the list.
I suggested that in the survey actually and had a conversation about it in discord. it will probably take a full rework of how trains get dispatched over the route and even if only as a scenario take years of work. I would however be very interested in it though.