rFactor 2 Moving to DX11 - New Company Partnership Announced

Paul Jeffrey

Premium
rfactor 2 logo.png

With the build up to this weekend’s Sim Racing Expo, Image Space Incorporated have announced a strategic partnership with the Dutch software group Luminis. The new alliance has been created to help increase development in rFactor 2, including a number of exceptionally exciting announcements in relation to the sim.


The partnership will be brought together into a new company: Studio 397. Studio 397 will have its headquarters in Apeldoorn, the Netherlands and will operate under the leadership of Marcel Offermans, a fellow sim racer and the man responsible for bringing both rFactor and rFactor 2 to the Steam platform.

Gjon Camaj, Vice President of Image Space Incorporated:

“We at ISI are very pleased to have found a partner that shares our enthusiasm and vision for the continued growth of rFactor 2. We look forward to working with Luminis to expand the reach of rFactor 2 both in Europe and beyond. After having worked with Luminis for some time and seeing their passion for motorsport simulation, this partnership was a natural next step. By leveraging each of our company’s individual strengths, we will be better able to serve our current customers and further expand our reach into the world of motorsports”.

Marcel Offermans, Luminis Fellow and Managing Director of Studio 397:

“I’m really excited by the opportunities that our partnership with ISI offers; rFactor 2 represents the best of breed, stable simulator platform in the industry. Online simracing has been my longtime personal passion, so the opportunity to combine my passion with my business feels like a great privilege. By infusing the platform with some of our own technology we see great opportunities to accelerate development in general and specifically in the areas of competitions and training”.


The new partnership have released an ambitious roadmap for rFactor 2’s further development including the stated plans to move over to a DX11 compatible graphics engine, including VR support for both Oculus Rift and HTC Vive. With the move to DX11, Studio 397 have committed to bringing performance improvements for high end video card users as well as maintaining compatibility with current content.

Other highlights from the announcement include a proposed new HTML based front end interface and Matchmaker support, in game UI updates and a promise to make more frequent car and track releases, working on racing cars and series as well as "popular and challenging" tracks. The new partnership also promises to work more closely with the community to support development of free and paid mod releases...

Oh, and the game has a new logo too....


All in all, this is fantastic news for rF2 and the sim racing community. I can hardly wait.

Check our rFactor 2 sub forum for news, discussions, mods, club and league races and everything in between!

What do you think of the announcement? Excited about DX11, VR support, new content? Let us know in the comments section below!
 
Last edited:
Not really. Input lag is a serious issue and nobody who knows their stuff will not change any setting knowingly that increases it. Turning on vsync is horrible advice. I hope nobody does it. If tearing is issue then run you fps at 83fps limited for 60Hz monitor and it will look better with no lag. Also screen tearing is all about how the fps matches your monitor refresh rate. If monitor is 60Hz and screen is 60 then you will have tearing. If monitor is 60 and fps is 90, 120, 240, any multiplier of 60 or 30 you will have tearing.

Your post is so wrong on so many areas that it sounds more like self rationalization than valid advice.

Pretty much any game, any sim. You get best performance if you run either unlimited 100+fps or cap it at 83 (or any other number that is not multiplier of 30). If you turn on vsync you are hurting yourself period.

I've never suffered from input lag when I've locked to vsync, the game (RF2) is smooth as silk and runs perfectly with it on, my driving is not affected in the slightest (same with iRacing). When I turn vsync off I get annoying tearing which I simply cannot live with.
 
Not really. Input lag is a serious issue and nobody who knows their stuff will not change any setting knowingly that increases it. Turning on vsync is horrible advice. I hope nobody does it. If tearing is issue then run you fps at 83fps limited for 60Hz monitor and it will look better with no lag. Also screen tearing is all about how the fps matches your monitor refresh rate. If monitor is 60Hz and screen is 60 then you will have tearing. If monitor is 60 and fps is 90, 120, 240, any multiplier of 60 or 30 you will have tearing.

Your post is so wrong on so many areas that it sounds more like self rationalization than valid advice.

Pretty much any game, any sim. You get best performance if you run either unlimited 100+fps or cap it at 83 (or any other number that is not multiplier of 30). If you turn on vsync you are hurting yourself period.
I've been using 64fps cap on a lot of games to avoid screen tearing and too much work for the pc when you run unlimited fps. 83 is as valid as 64, maybe a bit more, but 64fps has been satisfying.
For example playing CS:GO I put a low resolution of 1024*768 because there I can achieve 75hz instead of just 60hz, so in this case I have to limit the fps above the refresh rate, to 83. (technically I limit it to 84 because in the Source engine with 83 it would drop to 82 and cause noticeable screen tearing).
 
They can just add DX12 support in the future guys. Going DX11 now does not mean the door is closed, really get some common sense there is no reason to go DX12 now due to it being Win10 only, way too much people didn't move to it yet.

I wonder how many of those "didn't move to Win10" are really corporations rather than the gaming public. I know there's a vocal minority of gamers who always stay on the older Win version, but it's always been a minority.

Not disagreeing with your conclusion because DX11 is a no-brainer move whereas DX12 would be out on the bleeding edge of marketing.
 
rF2 however is CPU/loop limited in that it doesn't really put much stress on the render output at all, but a lot more is lost in what i believe to be some inefficient code at the very core of the simulation. Now, even if you switch your output to DX11, the core is still going to be the same, more or less and will still impact framerate severely.
rF2 is CPU bound mostly due to DX9. Due to the fact that it was created when single/dual cores were the standard and GPUs weren't really effective in anything but graphics rendering, it doesn't support many advanced technologies that could take load off the CPU and/or would allow the GPU to utilize multiple threads effectively. Just think about the current shadow implementation that relies hugely on CPU and think about the gain if it could be calculated completely on the GPU. iRacing is indeed a really good example of DX9 vs DX11, and the difference it can make in fps count and stability.

DX9 was the bells and whistles back in the day, but is one of the biggest bottlenecks in current games. RaceRoom? CPU bound. Automobilista? CPU bound. rFactor2? CPU bound... Assetto Corsa/iRacing (DX11)/pCARS? You can run your GPU on 100% percent without hitting a CPU limit. The move on to DX11 can only do good to rF2, and will definitely open up the possibilities to improve/implement features further. ;)
 
As for increasing fps, as long as you hit 60 fps, and can then lock it there (vsync) you can have the best experience. And unlike what the "self proclaimed alien elite" will tell you, the input lag of it is so small it is negligable, so trying to reach 200fps because you think it makes you a better simracer, while learning not to get sick and dizzy from the violent screen tearing and stutter is silly anyhow, imo.

To me vsync off made a big difference, even quantified: 0,6s per lap quicker than running vsync. But maybe I am alien elite and my opinion doesn't count :O_o:

Edit: Also I don't think you need alien elite eyes to spot the difference :)


 
Last edited:
What is up with the new RF2 logo? My god , my eyes started bleeding after looking at it. Terrible design.

On the DX11.....makes me real happy! I still run Win7 ultimate and don't plan on making the switch to win10 any time soon.
 
I wonder how many of those "didn't move to Win10" are really corporations rather than the gaming public. I know there's a vocal minority of gamers who always stay on the older Win version, but it's always been a minority.

Not disagreeing with your conclusion because DX11 is a no-brainer move whereas DX12 would be out on the bleeding edge of marketing.
It's easy, just check Steam data, 47% of people are on Win10 now, it is still too much people on other systems to drop them from your potential customer list
 
iRacing was rendering limited to a large degree, they had many effects that could make use of the DX11 post-ptocess pipeline, thus giving a massive increase in framerates on modern hw.

rF2 however is CPU/loop limited in that it doesn't really put much stress on the render output at all, but a lot more is lost in what i believe to be some inefficient code at the very core of the simulation. Now, even if you switch your output to DX11, the core is still going to be the same, more or less and will still impact framerate severely.

As for increasing fps, as long as you hit 60 fps, and can then lock it there (vsync) you can have the best experience. And unlike what the "self proclaimed alien elite" will tell you, the input lag of it is so small it is negligable, so trying to reach 200fps because you think it makes you a better simracer, while learning not to get sick and dizzy from the violent screen tearing and stutter is silly anyhow, imo.
Can't disagree more.
To me vsync off made a big difference, even quantified: 0,6s per lap quicker than running vsync. But maybe I am alien elite and my opinion doesn't count :O_o:
Can't agree more.
 
as my understanding of moving to an iRacing-like browser launcher.
I seriously hope rf2 is not moving towards the ac style of things where the launcher and the sim are two separate programs. One of the big strengths of having one program with launcher, menus and most of the rest is that it allows you far more freedom. Live for speed for example is done in this way (although skin viewer is separate program) and having this system allows people to change almost all of their settings while driving, change cars on server without having to leave server or to switch skins.

To specify a launcher is a program that you start first. You adjust your graphics, ffb settings and everything else there, you choose car, track and setup all your settings for the session you want to do. After that you then press go and load the simulation aspect which usually can not do any of that stuff. Which means you are stuck with what you chose earlier.

A launcher system like iracing works for them really well because they are 100% online based service. They have made all the selections for you so the average user only needs to select series (and maybe even a car) and then just choose what server to join. But I really hope rf2 is not going this direction. I'm kinda hoping we see more flexibility that really uses the benefits of the single program approach.

With the launcher and the sim portion being inside the same program you can do stuff like showing the 3d models for example. This means you can look at the model and switch between different upgrades (wings, tires etc..) and see the changes on the model. If you go for a 2d launcher you need to manually generate some 2d images to show off those upgrades. Same thing for every car and every skin. And every upgrade. Without any kind of automated screen generator program this is huge job alone. And needs to be done everytime you update a skin, car model or shaders.
 
I also have a question for Marcel. Now that you own rf2 does this mean that you could implement some of the automobilista stuff you have made for reiza into rf2 as well? What little detective work I've done it seems you guys also do some/all coding work for automobilista as well so is this something you think you could do? Or does reiza own the code you have created and changed for automobilista?
 
Going to DX12 probably is a bigger change than most think then to move to DX11, if im not completely wrong the onus of a series of things in DX12 moves from the API to the coder itself so probably it isnt as easy as going fom 9 to 11... i might be wrong though.
 
I also have a question for Marcel. Now that you own rf2 does this mean that you could implement some of the automobilista stuff you have made for reiza into rf2 as well? What little detective work I've done it seems you guys also do some/all coding work for automobilista as well so is this something you think you could do? Or does reiza own the code you have created and changed for automobilista?

This is the big question. With Luminis in the middle, can there be cross-pollenization between AMS (and Reiza 17) and rF2? If so, both ways or only one direction? And who profits from it?

I promoted a merger between ISI and Reiza for the purpose of rescuing rF2 and for the general advancement of racing sims a long time back. (Take ISI's technical know-how and combine it with a firm that understands marketing and customer relations and how to do much with little.) This new organization could achieve exactly what I was hoping, but through an even better structure, since Luminus appears to be highly competent at code development and implementation...and has people passionate about racing sims.

But the intricacies of what can be shared across the corporate relationships is very intriguing. Marcel may be able to shed some light. If Luminus can maintain and supply that intelligence to multiple bidders, all the better for the industry. A company that specializes in rapid development of racing sim technology and code components...wow, our hobby has now matured to a new level I think :)
 
Ok. Some nice discussions on the vsync thing as always :) I am happy to keep discussing it. Now, i see a claim that by some miracle 84 fps cap is supposed to keep stuff fluid. Now, i have never heard this claim in my 30 years of professional software development, part of the last 20 within 3d rendering.

Logically and strictly mathematical there is NO way to keep fluid motion on a monitor without flipping synced on vblank. None, nafa, zip. No matter how you cap rendering output it can not ever be fluid. Because you do not have sync between buffer and monitor the swap will be uneven.

With a fps cap it may be slightly regular, but still not on the blank so will always be janky when panning etc.

If you have mathematical proof that somehow you can get a switch exactly on a blank by using a cap, enlighten me.

Also, frame drift on say 84 fps cap on 60hz monitor equates to about 6.72 ms. Now, if you have reactions fast enough to have a difference then congrats. I don't. I think it is a case of what is studied in academia relating to the brain exaggerating small changes in visual stimuli and getting confused even on slight changes, so that you perceive it as very wrong. Thus, if you are used to the stutter and uneven motion of non synced video, you perceive vsync as "slower" in a much higher degree than it actually is.

Another note on this... Even if it made MY lap times 0.6s faster i wouldnt care for it. Why? Because simulation for me is about immersion. I run sims to immerse and experience, not for the reasons people play Quake 3 or DOTA, meaning i rarely care for 0.6s anyway, if it means making the sim less real and immersive. If you play for different reasons where it is important to just win, it is different. I ask you: if, by some mystery, turning off all sounds gained you another second on your laptime, would you?
 

What are you racing on?

  • Racing rig

    Votes: 528 35.2%
  • Motion rig

    Votes: 43 2.9%
  • Pull-out-rig

    Votes: 54 3.6%
  • Wheel stand

    Votes: 191 12.7%
  • My desktop

    Votes: 618 41.2%
  • Something else

    Votes: 66 4.4%
Back
Top