Ooooh yes, it's been a while since the last notes. Let's get started. 1. Brand new Force Feedback (Target release = June 2021) We have been working hard on a complete rework of our FFB, which we are all looking forward to sharing with you. We know FFB is a sensitive subject for many simracers and we want to properly explain why we think this was the right way to go and how we went about building the new system up from scratch. We hope everyone will be as excited as we are about the major improvements we have made here! This project was initiated after taking stock of the current FFB situation. As car physics were being developed, the amount of details going in that work has been shadowed by our FFB, doing it no justice. A list of issues was put together, our FFB code was analyzed, flaws were discovered and the whole thing was redesigned and rebuilt. Old FFB: Steering geometry of the car is ignored, resulting in little differences in FFB between cars. It results in the steering torques being wrong, as they depend only on tyre forces rather than the actual characteristics of the car Lack of longitudinal tyre forces Lack of gyroscopic forces from the tyres The tyre forces from physics are badly translated into FFB torque Tyre contact patch movement is ignored, no dynamic change while driving Too many sliders in the FFB options menu, some of them making promises the FFB could not deliver New FFB: Steering geometry is used: much more uniqueness between cars, the resulting steering torques have real meaning All tyre forces are taken into account Gyroscopic effects of tyres taken into account 100% Physics-based calculations Tyre contact patch movement is used: much more dynamic feel while driving Only crucial settings to adjust the FFB to the various steering wheels Spoiler: Click here to read the full FFB story What is FFB? The core idea of FFB is to simulate the feeling of driving a car as closely as possible. There are two main aspects to this, the first being the steering forces that you would feel in the steering wheel of the real car, and the second aspect is to enhance the experience by emulating other sensations you would have while driving a real car. A good example of this is the feeling of the car bouncing up and down as you drive on the track, which is almost never perfectly smooth. In a real car you would feel this mostly through the seat, which is why you have motion platforms such as TrackTime, or "buttkickers" to get these forces to actually come through your simulator seat. However, those devices are still rare and not many simracers are equipped with them, so an alternative is to add these forces to the FFB in some way. It might not be the same as feeling it through the seat, but for a part of the simracing community, it can increase the feeling of immersion. Not everyone likes these extra effects, so it remains something optional for the simracers that do want it, while those that prefer pure steering forces can stick to that. Steering forces The steering wheel is obviously connected to the wheels to make them turn, so it makes sense that forces that are generated by the tyre can also be felt in some way in the steering. However, because the steering can only make the wheels rotate around the steering axis, what we actually feel through the wheel are the torques that the tyre forces generate around this steering axis. In general, the further away the force is from the axis of rotation, the larger the torque that is produced. This is also why having a larger steering wheel makes the steering lighter, as your hands are further away from the axis of rotation. The force you produce stays the same, but it generates a larger torque than it would with a smaller steering wheel. Another important factor is the direction of the force: when moving the steering wheel, you actually push upwards with one hand (when starting at neutral steering). If you were to push to the left instead, the wheel wouldn't rotate at all, even if you do it with the same force, no torque is created. If you were to draw a line in the direction you are pushing, the effective distance (moment arm) of the force is the smallest distance from the axis of rotation to this line: The torque acting on the axis of rotation is then simply the moment arm multiplied by the force. In 3D this gets a little more complicated, but the principle stays the same. Steering forces in the old FFB The current FFB system actually completely ignores this behaviour and simply uses the tyre forces directly as an FFB signal, scaled by a certain constant value. Furthermore, it even completely ignores the longitudinal forces that the tyres produce when braking (or on power for FWD). All it does is add up the vertical and lateral forces from the tyres with a separate scaling for each, which is what you can set in the FFB menu. However, this method does not necessarily mean that the FFB is completely unusable. What it effectively does, is to pick an imaginary moment arm that these forces have on the steering axis, and simply always use this value as a constant. This would be valid if the steering axis and the origin and orientation of the forces always stayed the same. Unfortunately this is absolutely not the case, even for a single car with one setup, because the tyres deform and move around. This can change where the forces originate from, which in turn can change the moment arm that these forces have on the steering axis. So even if the forces stay the same, the torques these forces are producing can change as the tyre deforms. It gets even worse when you factor in the changes that can be made to the steering geometry in the setup such as caster, which would rotate the steering axis around. This again changes the moment arm the forces have on the steering axis. And even worse when you take into account that different cars can have vastly different steering geometry. In conclusion, using one single approximation of the moment arm of the tyre forces is in no way realistic. Now the reason that this code was originally written like this makes sense, as it makes it so that the FFB is completely independent of the suspension and steering geometry. Which would only bring a lot of issues with it if you don't exactly know how to set up this geometry properly. However, this aspect of the car design is what @Alex Hodgkinson has been perfecting over the years. It is a huge shame that all this work is not making any difference in the FFB, besides the tyre forces changing as physics updates come in. Steering forces in the new FFB: With that in mind, we wanted to make changes to the FFB code, so that all these static and dynamic geometry changes can actually make a difference to the feeling of cars. It quickly became clear that the easiest way to do this would simply be to start from scratch and build the FFB in the most accurate way we can. Using the basic principle of calculating torques explained earlier, we can make a list of the necessary information to calculate the torque that the tyre forces produce on the steering wheel: - The position and orientation of the steering axis - The point from which the tyre forces originate - The direction the forces are working in - The magnitude of the forces - The steering ratio from the steering wheel to the steering axis As these factors are all used in the driving physics already, it was just a matter of collecting all the necessary data together and implement the calculations required to turn this data into a resulting steering torque. As these calculations are in 3D-space, there is some vector calculus required, about which I will not go into detail here. The bottom line is that we implemented a generalised calculation that can calculate the torque that one or multiple forces will create around any steering axis. The calculation is valid for any position and orientation of both the steering axis and forces. This means that we can do whatever we want with the car/tyre design and the resulting steering torque will still be mathematically accurate. Ofcourse we then have to make sure that the car/tyre design is as accurate as possible to actually achieve a realistic steering torque. The big advantage of a fully physics- and mathematics-based FFB system is that the numbers we get at every step actually have a true meaning. The torque we calculate to be around the steering axis is the actual torque that would be present in a real car with identical tyres and steering geometry. This makes tuning the different parameters we have available for the steering geometry and the tyres much easier, as we can easily see if a result makes sense or not compared to what we would expect to see in the real world. Gyroscopic effect of spinning tyres Besides the vertical, lateral and longitudinal forces that tyres produce to make the car move, there is another significant effect that wasn't taken into account at all in the old FFB. As tyres are spinning objects when driving and have quite some inertia, they behave as gyroscopes. This means that they will resist being rotated around an axis different than the axis around which they are spinning. When turning the wheels with the steering wheel, that is exactly what we are doing. However, this has another even more important effect, as gyroscopes want to stay in the same orientation compared to the rest of the world. So even when we are not turning the steering wheel, but the car is rotating around a corner, the wheels resist this rotation and want to keep going straight. You will be able to feel this effect in the steering wheel quite clearly. This might not seem all that exciting yet, but when we think of what kind of implications this has in different situations, it actually turns out to be extremely important to the feel and control of a car. What happens when you get a snap of oversteer, both on power or on the brakes, is that the car suddenly starts to rotate (yaw) very quickly. As the tyres resist this rotation due to the gyroscopic effect, you will actually feel the wheels pushing towards the countersteer direction. A similar self-aligning effect is produced by the lateral tyre forces because of the steering geometry. However, these tyre forces take some time to build up, sometimes making it too late to react to the slide. The gyroscopic effect on the other hand reacts instantly to any rotation and makes feeling the rotation of the car and therefore keeping in control of the car much easier! Maybe slightly less crucial for the driving feel, but equally as awesome is that this effect also massively reduces the oscillations you get when letting go of the wheel. This makes sense as the wheels simply want to continue in the direction they are facing, any deviations from this path create a force in the opposite direction, keeping the wheel much more stable in the center. 2. New Options Menus (Target release = June 2021) As you know we have been working on revamping our user interface, improving the look&feel but also adding functionalities to try and maximize the user experience. The options menu has received a new look and structure, but also two major quality of life improvements: Search, and FOV / Triple Screen set up tools. Search & Press key to jump to binding The amount of settings and control bindings in a simulator is often overwhelmingly high. In addition to a better sorting, we have added a search tool that lets you go straight to the setting you are looking for. Another good addition is the fact you will now be able to press the button and immediately jump to that assignment. FOV and Triple Screen tools You will now be able to set those up directly from within the game, with a useful preview illustrating the changes. For a single screen setup, pick your screen size, the distance from it, the eye offset and then tweak to your liking from the calculated FOV. (I always feel those are maybe natural but too narrow to be useful. Change my mind!) For triple monitors, simply fill in their size(s), angles, distance, bezels and eye offset. If you accessed the Options menu from your car, loaded on a track, you can even enter a live calibration mode allowing you to directly see the effects of your changes, as well as freely adjust the position of your eyepoint. 3. New Replay Interface (Target release = June 2021) One of the most annoying flaws identified in the current replay interface is the lack of controls immediately accessible to the user. Some of them do exist but are accessible through hardcoded keybinds that are nowhere explained. We also had limited free camera movements as well as not much freedom in how those movements are executed. With this Replay Interface update, we bring you keyframe camera animations, expanded free camera movements, full disclosure of all the useful keybinds as well as ability to bind some extra ones. A noteworthy addition is also the list of all the drivers and the ability to switch directly to any of them without cycling through the entire grid. Here's a quick preview of it all where I try and create a nice panning shot of cars passing by, and then a dolly cam shot. Forgive my lack of talent with those shots, I'm sure you will be using this for much better looking shots. 4. Sparks (Target release = When it's ready) 'nuf said. 5. Smoke and dirt puffs (Target release = When it's ready) While looking into our particle system for sparks, we also looked into improving on smoke and dirt clouds. We ended up tweaking their density, their points of origin as well as how they are affected by passing cars and by... the wind. 6. Thrustmaster SDK (Target release = When it's ready) We have been supporting Thrustmaster with their new suite of drivers and firmwares. Once that work is complete and everything works, it will be pushed in a RaceRoom update. 7. Physics updates (Target release = Often ) The focus has recently been on formula cars. You can already enjoy the reviewed Formula Junior and Tatuus F4 in the public version. Next in line are the Formula RaceRoom 3 (currently in beta), then the 2, and then the FRX-17. These single seater updates will be very good practice for our next FRX model! Which brings us to... 8. New content Some of you already know about the Opel Astra TCR coming soon to RaceRoom. We are constantly working on new content, but those are better announced differently. We are always busy on tracks and cars that will no doubt tick your interest. We also experiment with new categories. Stay tuned! 9. Ranked Multiplayer - Series & Championships The next step for Ranked MP is the creation of series / championships where each player will not only race for rating and reputation, but also score points in a race based on how highly rated the opponents were. The system will keep the best result scoring the most points and use that for the series rankings. This work involves further automation of server management as well as new backend features that take time to develop while also maintaining the existing infrastructure. I hope for this to be delivered this year.
The "New Replay Interface" is good. But it would be even better if the replays have the same overlays as on TV broadcasts. It is not always possible to conduct an online broadcast, and recording races from a replay is no longer beautiful and informative.
Note that OtterHud works in replays now. (see changelog here: https://forum.sector3studios.com/in...custom-webhud-with-additional-features.13152/)
Thank you devs for this wonderful updated road map. Brno is such a beautiful track. Nice to see it in a modern sim. And regarding sparks: was about time! They look awesome in those gifs. The new FFB gets me nervous. Just recently I finally got it dialed in for my wheel after a long time and having a blast with the sim since then. Personally I really like the FFB as it is right now. Fingers crossed that the new version will be as good as it sounds like.
I hope the canned effects will stay at least for the gearshifts. I really miss the little knocks in any other sims after getting used to it in RR.
I thought so too....but I've just spent the last 10 minutes trying to re-create the same view but I couldn't do it.
I have no idea what camera settings mightve been tweaked to get this exact shot (maybe even something only available through the new wip replay options?) but its definitely the norisring
You're right, definitely the Norisring. Although the banners are changed (might be a competition thing) and the lighting in the gif looks very diifferent.
That is definetely Norisring T3. I believe it was shot during a low sun angle, so the big stone grandstand puts his shadow on the back straight. Greetings
So it won't be possible to specify the screen width and the height for Triples anymore? This is an important part in the process to allign and correct proportions of mismatching triples (16:10,16:9,16:10).Without these parameters it will be almost impossible to get the correct view :-(