I understand that some of the improvements with Direct X is efficiency. To me the game looks good so I'm not really interested in better graphics but I understand with the efficiency gains you could boost performance and also gain poly count. For you experts out there, how hard is it to upgrade Direct X versions? Does it basically require creating a new engine or is it more manageable - maybe modular where you can replace old code and such? I wonder how much improvement would be expected? Is this something S3 is looking at or is it just not feasible? Thanks
Do to that Oculus won't have DX9 support, I really hope S3S will make that change! R3E is my favorite sim, and would be a shame if I won't be able to use it with my future VR rig.. . Just a side note: Next Car Game now have two launch options on steam. One for the old version, and one for the DX11 version. (if this would be a possibility to still support players with older hardware?)
Its difficult but not improbable to change from DX9 to DX11.The main difference between the two is shaders and textures.The main graphic code is quite similar.But it will need some manpower to work on already released content to bring it up to the new standards.All in all its a matter of will and if S3S thinks that a conversion to DX11 will boost the game they can make it happen.
Let's assume we don't change anything of the visual outcome (no new fx, same lighting...). There is sort of two levels of migration: * port what you have to the new API: more or less keeps things a is, in principle, but may not give actually better performance despite the "better api", simply as you write some sort of translation code of old api to new, but your engine was in principle designed for the old one... * a design that leverages the new API well requires more work on changing the actual rendering engine to make use of new features in the api and also use the api well. You may want to move some of the CPU calculations or setups done for dx9 to GPU for dx11... the latter is of course the better option, and at this point in time, you'd likely want to design with dx12/vulkan in mind as well.
actually when ppl was asking JF about better triple screen support, he said he prefers VR over triple screen. and since it is not gonna happen on DX9, most probably they are already looking for the possibilities.... fingers crossed
yeah - I remember the answer from J-F in the Simracing-Mag and I hoping for VR to. Especially as VR isn't more expensive in comparison to a proper triple-Screen-setup - maybe even cheaper but with better immersion.
indeed , fixed it as for VR, VR is sexy and I also would love to avoid tripple, but in the end I don't think we will have that good of a VR in the first generation. The displays just don't have the resolution yet (1080×1200 per eye, to be "real" you would need something like 8k per eye afaik, of course much less will be sufficient for gaming, but let's say 2k-4k). We also need a lot of gfx power as you need even better anti-aliasing now. No doubt, it will be solved eventually and it's gonna be fun and exciting for sure when it comes out, but replacing a more "easy for the eye" system, not yet in my opinion. So tripples should stay relevant a good time. given the topic, there is also some background info about DX apis on the forum already
Well, it's a long one, a very, very loooooong pipe. Only threads with the topic of a new engine are longer.
If you take a gander of the @Alex Hodgkinson interview (most recent one) he does mention improvement of RR in the graphical department, be warned, I have just said as much as he did but it's definitely mentioned. So end of is it or isn't it, it is!