Let me start by saying I love the franchise, I love the concept, and I love the routes, and have largely held off making a post like this. However. How on earth are we four TSW iterations in and the game still isn’t optimized properly? Why are there still immersion-breaking stutters when routes are played at low (and high) graphics settings? Why does it feel like the community is constantly pre-ship beta testing the software / product, and fixes are only effected well after the fact, if never at all? I sincerely regret purchasing TSW4 and uninstalling TSW3. TSW3 had finally gotten to a point for me where the stutters were minimal, the gameplay was largely well-optimized, and the game itself was in a relatively high state of playability. I’ve tried various permutations of .ini optimizations and graphics settings, and the game is still largely in a beta state. DTG - if you have any suggestions to optimize the gameplay, I’m all ears and all for solutions. Do better.
Optimization is one of the mantras people like to chant, but if you don't know exactly why it's stuttering on your system, and don't have access to the TSW source code and compilers, why invent a cause and blame instead of looking for actual solutions? Post your system specs, and maybe the community can offer suggestions.
If I switch the game to dx12 my cg goes from 54% to 95%!! I have an RTX 3060 not an RTX 4090.... In short, in dx11 I play in ultra at 1080p. In dx12 I play at average 1080p and my cg is at 95%... The game is sold in dx11 not dx12, I don't like tinkering with my games, I buy a "finished" game not a beta where I have to test a whole bunch of LOVE for it to work correctly. QNS Chris is right, it's up to DTG to make it work correctly, not us, or in this case, they'll reimburse me for the 1600 euros I spent on tsw1-tsw2-tsw3-tsw4 and a whole bunch of DLC if they want me or a “client” (pigeon) beta tester. They sell a game, they have to make it work properly. TSW3 was less choppy than tsw4! This is not logical, tsw4 should be better optimized. We waited almost 10 years for TSC to work smoothly without any jerks, there is no question of waiting that long for tsw! DTG makes it work as it should. STOP the tinkering bullshit.....
If I buy a new car in principle we are not going to make adjustments to the engine 10 seconds later. The car works as it was designed and sold. You buy a Ferrari, you get a Ferrari, not a Ford.
mate I have a GTX 1660 and some settings are on low and it runs fine for me. idk what CG and NaN% means but don't put up everything at ultra, it'll just make you able to bake a pizza with the thing you then used to play videogames on
After all this time it can only be 2 things. 1. Unreal engine is not the correct engine for this type of game 2. DTG do not have the skill sets to programme unreal
I understand the frustrations and I will not defend DTG what so ever. The progress they make is recognizable though. If you compare the hills from the TSW 2 routes with the new routes, it is stunning what they now achive. So I really don't think that there is a skill problem. But I really do think there is a priority problem. I really wish they would focus more on quality and not on quantity.
For me on a GTX1650 DX12 made no difference in fact it caused some of the in cab dials and text to become blurred. There are of course external reasons on PC at least that can affect performance of other software running, notably Windows Update. There was a huge one just a couple of days ago which crashed SimRail when I was playing it. However that does not excuse the fact that TSW does chug along with pauses and rubber banding affecting visuals and sound. Despite having run Vorarlberg a few times now across its entire length it still judders and pauses so can hardly be shaders still caching.
As a reminder, the RTX 3060 (12 GB) is normally made for ultra gaming at 1080p which in this case is the case. As a reminder of the specifications requested by dtg for the game: RECOMMENDED: Operating system: 64-bit Windows 10 or Windows 11 Processor: Intel Core i5-9600K @ 3.7 GHz or AMD Ryzen 5 3600 @ 3.6 GHz RAM: 16 GB of memory Graphics: NVIDIA GeForce RTX 2060 6 GB or AMD Radeon RX 5700 with 8 GB VRAM or more DirectX: Version 10 Network: high-speed internet connection Disk space: 80 GB of available disk space Sound card: DirectX Compatible Additional Notes: Requires mouse and keyboard or Xbox Controller It seems to me that the RTX 3060 is superior to the 2060 when it comes to the processor I have an i5 12400 and 32 GB of RAM and an SSDM2 and I am on Win 11 Pro. So I have a PC higher than the request. in short a PC from last year. The game is in ultra in dx11 and my PC is shaking its balls at 54% for the cg and 12% for the proc. Don't tell me the problem is with my hardware. You have to ask the right questions... I'm not complaining, I'm very happy with the performance of my cg and my PC in general, it's TSW4 that I'm not happy with. If I activate dx12 it's no longer possible to play ultra with this card which is completely crazy!!! I have plenty of other games in DX12 (the latest Forza is great for example) and I play in ultra without any problems. The problem is not even dx12 since the game is sold in dx11 and it works in dx11 natively (dtg does not recommend activating dx12). The problem is the catastrophic optimization of the game. The jerks every 5 seconds is not possible without counting the rather low graphics of the game, the trees which pop 50 meters in front of you, it is totally unacceptable. In short, I don't even try to defend dtg anymore, they are either incompetent or ue4 is simply not the right engine to use for tsw. For my part, ue4 is not good for this kind of game, I am even ready to buy a tsw5 and start from scratch if a better, better optimized game engine was sold for a tsw5. I like the game, I've been playing train with dtg since 12 years ago and on trainz before him but I think that staying on u4 is a huge mistake. DTG should make a radical choice, change engines and start from scratch or the problems will only get worse over time. In the case of a change of engine we could anyway keep tsw4 and continue to play it as long as we want (we bought it, it seems to me...). TSW5 with its new beginning will only be a new train game like any other and we are leaving for a new adventure with this time a good game well optimized under penalty of death for DTG.... Because in this case they will no longer make any excuses. I would add that half of the game is broken, look at the different posts on the forum, obviously not much is working correctly, even you have problems with the editor obviously, finally do you want it to work so you can enjoy it yes or no? ...
I would not say they do not have the skills. While I agree that they have made some questionable programming decisions, the basics of TSW is pretty good. They just need to take more time and care when creating content. Because most of the issues come simply from badly setup and unoptimized assets, which could be easily avoided by a bit more care and polish. Just stop rushing things.
Unreal Engine 4 is not the problem. Switching to a new engine/UE5 won't make the game better by default. They need to invest in developer resources to optimise, rather than investing in an orchestral score.
No, they either have to change the engine (not sure that UE5 is the best choice again) or they definitely stop making train games..... They don't know how to optimize, the game was released 6 years ago!!! 6 years of non-optimization, please stop dreaming He invests in promotion not in quality or optimization, they already can't do an update without breaking the game.... I'm not adding that on TSW, we all saw the mess on TSC too.....
So if they don't know how to optimise, another engine wouldn't necessarily be any better, no matter what it is. No existing commercial engine will be built for TSW, so they would need to customise and optimise it. They haven't yet done everything they can with UE4. Even Will acknowledged they can try certain things, but he has no intention of doing it on this iteration because of the way the game is built meaning significant extra work. Switching to a new engine results in just as much and potentially more extra work (especially if it isn't a version of Unreal, it would definitely be starting from scratch).
That's exactly what I'm saying. Starting from scratch would be more beneficial than trying to “optimize” something that doesn’t work well. I repeat, 6 years since tsw was released and still no optimization for jerks!!! So, either they are incompetent or the game engine is not good and they are in a dead end!
The problem with starting from scratch is that everything in the game will be immediately obsolete and not work. That would likely alienate the majority of the player base.
Whatever engine dtg uses requires them to modify it heavily to run a train simulation. Dtg had to modify unreal a lot for it to do what they wanted it to do
I don't see why tsw4 would become obsolete?!! If a new game is created on a new engine, tsw4 remains on sale and available for those that have been purchased! When FIFA or Forza releases a new game the old one is still played! I have Forza Horizon 3-4 and 5 myself and I still play 3! I remember when tsw1 came out everyone said that tsc would be dead within the year! And obviously this is not the case, the proof is that they continue to sell it and update it... In short, anyone who doesn't want to buy tsw4 doesn't do it, I don't see why players would be "crazy" at the idea of having another tsw with a better engine and more than good optimization. A whole bunch of people including me bought tsw4 and it's not that I cried to death the loss of tsw3. A new game is a new game, nothing stops me from continuing to play the old one, I still play tsc.....
Possibly but it would have been a drop in the ocean, I suspect the money it cost would not have gone far on development costs. Maybe they have a marketing budget of which it came out of anyway. There have been many positive comments about the music so I suspect they will feel it was worth the investment.
Umm, no. Most of the performance issues probably come from simply badly set up assets. Like Clinchfield Railroad box cars causing framerate drop - probably due to something set up wrong in the SimuGraph. Another good example East New York station on LIRR causing framerate drops due to many lights being set to "cast shadow". The whole way of handling content by everything being entirely separate is also horrible for storage space and in certain cases performance as well. Like on a german route, when 5 different copies of the exact same locomotive can substitue, then that means the same locomotive might take up 5 times more memory for no real reason. And let's not even mention all the scenery objects that are used on all routes, yet still every single route have their own copy of them instead of them being a shared asset library. Sometimes scenery tiles also cause performance issues, because they get loaded at full LOD even if they don't have to be. This has been tested and proved by the community, as far as I remember. But all of these issues have nothing to do with the engine. It's just badly made objects or bad route building practices adding up one by one. A bit more care when creating these content and performance issues could be avoided. Now these might not affect the jerking you are talking about when loading scenery tiles during driving, but still, you get my point. Engine limitations aside, there are many optimalization problems in TSW that are not tied to the game engine used.
And I don't think DTG are the only ones in this dilemma. Run 8 desperately needs a better graphics engine, one that doesn't look like early MSTS, but the longer they leave it and the more of a catalogue of content the harder it is to make a clean break. Unless you can come up with a conversion utility to bring the old stuff across.
Do you mean literally creating their own game engine? Although even a jump to UE5 is effectively starting from scratch. Either way, if they don't currently have the resources to optimise the existing engine in known ways (Will, the lead software engineer, effectively said so in another thread) then they don't have the resources to build a UE5 version, and they certainly don't have the resources to create their own game engine from scratch (there are good reasons why most development studios use off the shelf engines). I am not against them doing either of these things, I just don't perceive either scenario as likely any time soon. And there is room for improvement using the current engine (again, acknowledged by Will in another thread), which retains backwards compatibility with players'/simmers' DLCs too, so I encourage them to do that first and get TSW to be the best it could be using what they already have.
I have several games that use the Unreal Engine and they all run fine to the point that I have never had to complain about them and/or request immediate improvements. TSW is the ONLY game that has such unacceptably poor performance that I refuse to support the franchise any further. Did they want it to run poorly, shuddering and stuttering in a very annoying and immersion shattering way? I don't think so. The fact that it does is proof that DTG have failed where many other software houses have succeeded in using the same engine to produce top notch products. I will never reward failure.
It seems a safe assumption that when DTG looked for a new engine to create their alternative to TSC, they settled on UE because it was the best fit. Which doesn't mean a perfect one, but nothing's perfect. DTG staff and third parties like Maik have explained in streams and on the forum some of the causes of problems that people on the outside like to put down to "optimization". It's such a small step from "This seems simple, so why aren't they doing it?" to "If they aren't doing it, maybe it's not simple", but so many people are unwilling to make that step.
Obsolete as in they wouldn't support it any further, regardless of whether or not it would still be new. Changing engines, or even upgrading to UR5 for that matter, makes everything on the current engine obsolete because it would be too time consuming to upgrade all DLCs to the new engine.
I don't understand why you want him to update tsw4 if he releases a new game on a new engine? A new game is a new game!! Why continue to support the old one if nothing more comes out on it in DLC? This doesn't stop people from continuing on the old game, however the new DLC will be on the new one. This is already the case with tsw4 as it was with tsw3 and tsw2 and tsw1 it seems to me. On tsw3 the last 3 routes do not work! You need to buy tsw4 to play it. So tsw3 is already just as obsolete as tsw2 and tsw1... Face the truth, no more DLC will be released on tsw3 and besides I think that the famous update will not see the light of day as was the case for tsw2.... So yes a new game would be beneficial because the current one is too "bugged" and is becoming more and more unrepairable. But hey, we'll talk about it again next year with tsw5, not true! Postponing the engine change again and again only accentuates future problems on the next tsw (5 - 6 - 7 etc etc)...
