hehe - yeah... thats true... and with your second sentence you gave your self an answer, why UE4 could be interessting for S3S. I'm pretty sure that Kartkraft will have Wheel-Support with FFB and that they want to have physics that are close to the real thing as possible. In Addition: S3S isn't a newbi in developing a Sim and know what a sim needs. So if they evaluate the Unreal Engine 4 i'm pretty sure that they have an eye on those aspects. I just found another Sim that will use UE4. Ok, its not racing, but a least driving https://www.facebook.com/FernbusSimulator/
AFAIK you CAN implement a 3rd party physics engine in UE, but it definitely isn't a plug and play job.
This is 100% not true. Many games from various genres are made with Unreal, Unity, and CryEngine. You can even find the extensive lists on the engines' websites. (In case you don't want to look it up yourself, here's the wiki page for Unreal alone: https://en.wikipedia.org/wiki/List_of_Unreal_Engine_games#Unreal_Engine_4 ) Kunos made their own engine because it started as a one-off Ferrari simulator that ended up becoming Assetto Corsa after changes/updates. Codemasters use their own EGO engine because they've had it for quite a while (Since 2008) and it makes sense to keep using it. Also, Codemasters needed something they could leverage across both console and PC without compromising either so a specific engine for themselves made more sense. Accurate physics are available within all 3 of the engines you mentioned: Unreal, Unity, and CryEngine. There would be some changes/additions necessary for the complexity of more realistic racing titles, but saying you don't want to use any of these engines because they can't do the physics is vastly misunderstanding the current state of modern game engine physics.
I think the best way would be if the Sector 3 Studios would start to program their own sim engine like iRacing / Papyrus.
hm, i'm not sure if it is really better. As long it is possible to achieve the wanted results with a third-party-engine, it is always better to use that engine. In that way you can focus your development on the final product. If a third-party-engine has to be modified heavily to achieve the wanted results, than the development of an own engine could be the better way. But developing an own engine costs alot and needs much time. Another benefit of using a third-party-engine is, that it will be updated for free and so you might be able to support the latest technologies without changing anything in your own coding.
well put @m.bohlken While there is a lot of charm in having your own complete technology stack, it is also rather costly upkeep. Upgrading graphics also typically means shader/material/effect editors for artists that go along with it, there is a certain rat tail to making technology accessible. unity and unreal have large development teams, hell even pure research teams just for graphics. Tech investment alone won't rescue, crytek went almost bankrupt (had to sell the engine to amazon aka lumberyard) and most of their original tech gurus are gone. Then there is a couple big inhouse engines that are not open (sony's studios, EA, various japanese studios...) and their costs are amortized across multiple titles. Using unity/unreal/whatever you as small team benefit from those huge r&d budgets (epic is probably >1b$ worth company) and can focus more on your actual game, where you want to make the difference (AI, physics, "atmosphere"...). Proprietary technology investment makes sense if you already have something and it's in a good shape/have expertise to advance it. Or if you think you can do a better job than your competition and it makes an impact for the target audience, gets you more sales/attention. But that is tough, you are up against forza, gran turismo, nfs in the mainstream... pcars did well by moving super fancy graphics to an audience that wasn't as much used to that but that ship has sailed already and increased the pressure to deliver similar in the long run. s3s has a really high quality content team, they need technology that doesn't hold that back. 3rd party technology reduces the team to be "at risk" when key people of proprietary technology leave due to whatever reasons. And given that is the situation how s3s came along, they probably want to play it safer
Are you sure? I think racing (and games that simulate racing) is as much dynamic as a shooter. Just think of all the physics stuff that has to be calculated in real time. The more CPU cycles you can use the better you could calculate this or calculate more things and even give AI cars the same physics as the player itself has to fight with. One of the most reasons why AI is performing not very well in a lot of the racing sims and at least the most graphical ones like pCARS is that it has to use simplified physics because you have not enough CPU power left to spent on accurate physics modeling for AI, too. As a proper implementation of DX12 could free up CPU cycles to do other things it would be a big step ahead especially to racing games.
that is the point why my amd cpu sucks in raceroom. Next gen titles racing or anything else are performing with good fps on my system. In r3e with full grid of ai it is sometimes hard to get 60fps when ai is before or behind me. That's the thing why I must say I need that new engine as soon as possible or sector3 can fix that problem, but I don't think it is possible with that dx9 old engine
@LTC Mike You are right if leaving everything as is, you will gain some CPU time in dx12 over dx9. However imo the gain factor will not be as huge as for the cases where dx12 shines (see the benchmarks of some asteroid fields, or RTS games with tons of units, each in its own action cycle and could be damaged to various degrees). Your game simulation, AI, physics, ffb... can run fully decoupled on its own threads independent of rendering already, even in dx9. Being limited by rendering doesn't mean you have to cut back on simulation fidelity. What dx12 is good at is rendering lots of unique elements in a scene that can change every frame, and that are not easily "batchable" as a system solution. Then you can benefit of processing these many objects with their state combinations in parallel for rendering. System solutions (vegetation, special fx, tire marks, dirt...) will benefit of dx11 already a huge deal by offloading more to GPU (compute shaders...). At some point dx12 will be the standard, as dx11 is today, but using dx12/vk now feels less crucial for R3E. I may underestimate the impact of headlight-shadows and dynamic reflections for water puddles and all that.
I am big time pissed by Microsoft's GWX-related "policies", which are an abuse of and assault on PC users. (GWX: get windows 10 software intrusion campaign). I would love to see S3 becoming pioneer for somethign else than just moving in a DX12 engione. I would like to see them transfering the sim to Linux. With that move they could make me buying everything I own a second time again - and for three times the money. And I own all tracks and cars and most skins. I have abanonded all non-gaming stuff under Windows already, and the more of the remainign sims and games I like do not depend on Windows anymore as a launching platformn, the better - and the sooner I could turn my back on MS completely.
I hope that we don't have to wait for dx11/12 (but I think that we will have to......) in order to have some proper shaders and lighting, going back and forth from assetto corsa, there's a world of difference in cockpit shaders and overall lighting. In Ac at least in daylight condition it actually feels to be under the sun, racing the nordschleife and passing from shadow areas to light to shadow again is just amazing, while in raceroom everything feels fake. And it's a shame since RRE is one of my preferred sim atm. I'm not saying that the graphic is bad but their lighting system combined with cockpit shaders and in some cases low poly models feels a bit outdated. And yes, I care a lot of graphic because the more it looks real the better it is for immersion (together of course with physic, ffb, sound etc....)
@Skybird what is the situation of ffb wheel & game controller drivers on Linux? Being a OpenGL/Vulkan developer I would second non dx12 choice, but if they go with UE4 then they get all the apis and platforms and it's not really their problem anymore.
Any updates regarding the Unreal Engine 4? Work in Progress image? With more race cars on a Track? Please? Because i want it so bad...
Hopefully we get some more infos about the direction of the game (sim?) soon. (With "soon" i don't mean "september-patch-soon")
Most probably not, more like "next-year-soon" I guess. As Chris said, it's all in a very early stage.
Early stage, of course. But at least it would be interesting to know, where Simbin wants to place its new product on the market. Realistic approach or more casually focused? Even in early stage there should be an aim.